7"“; ‘p 3's; ‘7' J Lo) 'LL Li "I”: 121;.ng ”5.1:;517Mfi ". Fffir' n wigwmvfiw: ‘ 17'3" {35:1 1’" 3‘3? w“: ‘I’i‘é ."Vj' 1“ 1:5- ,5 *3: [Z i.“ E |éj1§lL L.LL'LL‘LLLL Huidhi‘a‘.\lifi'i 5:: w‘. nib»?! L rraoa'a r r . '3er” 5 1'? 4 g»; EVE-fig}: 11'? fl L‘L \‘J Eur Li L- LLU‘I é: ‘ - 633333 363333 ‘63? L33 333;: 33 3f ELLBL" , .LLI‘CWCELLL 3‘? ME 33. VERSW LMZHB BELSGR 1:676 Lima“! ‘ T I . m3]:L131311111213111»MM}: 1 a, To ABSTRACT MARITAL POWER: A THEORETICAL EXPLICATION AND EMPIRICAL EXPLORATION OF THREE PROCESS MODELS By Elizabeth E. Benson This report analyzes one major research program within family sociology, that focusing on the ”Resource Theory” of marital power. Two implicit cognitive processes by which resource possession is assumed to be ”translated” into marital power were identified, center— ing around the concepts of perceived competence and gratitude—and- respect. These models were contrasted with a third theoretical process suggested by critics of the Resource Theory, that of power-dependence theory. A case was made that the Resource Theory research program has reached an impassé, where each additional study only increases the number of contradictory and inconsistent findings. On the hunch that this impasse is at least partially due to the underlying assumptions of the theoretical reasoning, the job of explicating and testing the assumptions in this line of reasoning was undertaken. Hypotheses were tested on data from a sample of 90 upper-middle class, urban, newlywed Michigan couples ranging in age from 22 to 40 and including couples in both first and second marriages. Both wife and Aw.“ hi» «W .q. .U All til 1. r0 in“ N ./ Elizabeth E. Benson Mnband were interviewed, with gender of interviewers matched with that of respondents. Couples were interviewed at two points in time, by telephone shortly after their wedding and in their homes several months thereafter. One benefit of the longitudinal research design was that it allowed identification of the role preferences about task division with which respondents had begun the marriage. These preferences were compared with later reports about the actual division of tasks. The average degree to which a respondent's initial preferences were achieved was taken as the measure of marital power, using only those items on which husband and wife had begun marriage with different ideas. When previously implicit parts of the theoretical line of reasoning were directly tested, the results were rather mixed. This calls into question some of the premises on which these theories are based and suggests that theoretical refinement should be the first order of future research. Findings with respect to specific theoretical process—models indicated that each theory succeeded in describing marriages to some extent. In the perceived competence formulation, one‘s marital power is hypothesized to be directly related to one's competence. Gender differences were found in types of competence which were related to marital power. The competence of dependability was most consistently related to marital power. Husbands yielded to a ”dependable” spouse only in areas they considered unimportant, while wives did so on important issues as well. Wives, but not husbands, are likely to comply with the wishes of a spouse who is considered to be “intelligent.” Elizabeth E. Benson The gratitude formulation is the second model which was examined. Like perceived competence, it embodies a line of reasoning found in the Resource Theory work. Spouses who contribute greater ”resources" to the marriage are hypothesized to earn power because of the gratitude and respect of their mates. Only an indirect test of HMS formulation was possible, for which more support was found for husbands than for wives. Power-dependence theory was the final formulation tested. In this theory, the higher the dependence of a spouse upon the mate, thelower that spouse's power. Two related aspects of dependence are distinguished, ”motivational investment" in goals mediated by the spouse and the presence of ”alternatives” for obtaining valued goal— objects from a source other than the spouse. Relative marital power is posited to vary directly with level of alternatives and inversely with level of motivational investment. The data trends were in the predicted direction for some indicators of potential alternatives—to-the—present marriage. Very little support was found for power—dependence hypotheses involving motivational investment. An auxilliary set of hypotheses related to power—dependence was also tested. Based on the work of Dahlstrom, these hypotheses deal with the notion that the average amount of power exercised within a couple across multiple issues will increase as the interests of spouses become more divergent. That is, spouses will have an increased ‘Wncentive to comply” with the other's wishes if, in return, the other spouse will comply on another issue. Such ”trade—offs” were thought Elizabeth E. Benson more likely to occur when spouses felt strongly about different issues, rather than about the same issues. Tests of this hypothesis were generally positive. However, the presence of trade—offs was not directly measured. Finally, test situations were set up for which alternative theories would predict different results, as a technique for assessing the relative applicability of the several theories. In specific test— situations, the gratitude formulation described the findings better than the perceived competence formulation. Power-dependence theory “fit” the data better when the task area was one which had been tradi- tionally assigned to the spouse's gender, while the competence/gratitude if formulation provided a somewhat better “fit” for other task areas. In a different test—situation, power—dependence predictions were supported over those based on competence.and gratitude formulations. Yet another test, contrasting only competence theory with power-dependence theory, indicated slightly more support for the latter. The role of task importance, as a scope condition, was also explored. Power-dependence predictions held more strongly under high- importance conditions for wives and under low-importance conditions for husbands. Perceived competence predictions held more strongly for both genders when task importance was low. These assessments of the relative adequacy of different theories must be viewed as very tentative, inasmuch as the sample size was not large and the indicators require refinement. However, in general, it appears that power-dependence theory provided the best description of marital power processes, of the three theories which were examined in this study. MARITAL POWER: A THEORETICAL EXPLICATION AND EMPIRICAL EXPLORATION OF THREE PROCESS MODELS By qvkr' / I Elizabeth EI Benson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology I976 © Copyright by Elizabeth E. Benson 1976 ? Dedicated to my father and mother ACKNOWLEDGMENTS Although it is strange to say, any successful completion of a dissertation depends upon whether or not people other than the author are willing to go out of their way. I was indeed fortunate in having many such people "on my side” in this undertaking. For academic guidance and consistent support throughout the course of my graduate studies, I would like to extend appreciation to nw guidance committee members, Drs. Bo Anderson, Frederick Waisanen, Beatrice Paolucci, and Barrie Thorne. Particular thanks are due to nw chairman, Dr. Anderson, who combines the ability to see the crux of a complicated theoretical issue with the ability to strike a happy balance between guidance and letting a student explore his or her own interests and ideas. For assistantships and other indirect financial assistance which made it possible for me to continue my work, I am grateful to Dr. Ann Olmsted, the Department of Sociology, and to the Training Program in Comparative Sociology. Among the many people who have helped make possible this study, I wish to give special thanks to the following: M. Acquilina M. Lovall D. Bozyk D. Osborn M. Cook H. Perlstadt B. Finifter M. Radelet The l974 ”Grove Street group” A. Schroeder M. Kupsis H. Schwarzweller L. Lovall IN Perhaps my largest debt of gratitude goes to my parents, Dr. and Mrs. F. F. Benson, and to my husband, Werner von der Ohe, who have been unstinting in both their personal encouragement and practical assistance. Finally, I wish to acknowledge the contributions of Mrs. Elayne Ballance, a typist without equal. TABLE OF CONTENTS Page LIST OF TABLES ........................ x LIST OF FIGURES ....................... xiii Chapter I INTRODUCTION .................... l II THE RESOURCE THEORY OF MARITAL POWER: A LITERATURE REVIEW ................ 4 Importance of the Resource Theory ........ 6 Development of the Resource Theory ....... 6 Blood and Wolfe (l960) ............ 7 Resources ................. 7 Power ................... 9 The Link Between Resources and Power . . . l5 Empirical Findings ............ l6 Michel (l967) ................ 2l Safilios—Rothschild (1967) .......... 22 Burié and Zecevié (l967) ........... 24 Summary of Cross-Cultural Findings ...... 25 Rodman (l967) ................ 26 Burr, Ahern, and Knowles (l973) ....... 28 III ANALYSIS OF THE RESOURCE-THEORY WORK ........ 32 Critique: Marital Power ............ 32 The Criterion of Empirical Fit ........ 33 The Criterion of Empirical Independence . . . 35 Defining and Measuring Marital Power: The Approach of the Present Study ..... 37 The Concept of Power . . . . . . . . . . . 37 Meeting the Criterion of Empirical Fit . . 38 Meeting the Criterion of Conceptual Independence . . .. . . . . ...... 39 Meeting the Criterion of Empirical Independence .............. 40 The Indicator of Marital Power ...... 4O [napter Chapter Page Concluding Comments ............. 44 Individual versus Relational Power . . . . 44 Contested versus Uncontested Measures of Power ................ 45 Zero-sum versus Non—zero-sum Power . . . . 45 Identification of the Critical Element of Power ................ 47 Critique: Resources .............. 48 "Overdetermined“ Relationships ....... 48 Multiple Resources and Testability ..... 49 IV STRATEGIES OF INQUIRY ............... 52 Output as the Starting Point .......... 52 Resources as the Starting Point ........ 53 Throughput as the Starting Point ........ 54 V ”EQUAL-INPUT” THEORIES OF MARITAL POWER ...... 57 Introduction: Equal-Input vs. Power- Dependence Theories .............. 57 Perceived Competence .............. 59 Background of the Berger Program ...... 62 Development of Expectation State Theory ... . 67 Generalizing the Theory ........... 69 Relevance of Expectation State Theory to Marriage .............. 70 Hypotheses ................. 75 Gratitude and Respect ............. 80 VI A POWER—DEPENDENCE THEORY OF MARITAL POWER . . . . 84 Thibaut and Kelley (l960) ........... 86 Emerson (l962) ................ 92 Power-Dependence Hypotheses: Part I ...... 96 Power-Dependence Hypotheses: Part II ..... 99 VII CRITICAL TESTS .................. l06 Tests of Alternative Processes ......... l06 Competence versus Gratitude Theories . . . . l06 Equal—Input Theories versus Power- Dependence ................ lO8 Competence versus Power—Dependence ..... l 3 Tests Involving Specific Resources ....... ll8 Competence and Gratitude versus 119 Power—Dependence ............. 121 Competence versus Power-Dependence ..... vi mter IIII Chapter VIII METHODOLOGY .................. Overview of the Study ............. The Panel Design .............. Justification of the Design . . ....... Inclusion of Both Spouses ........ The Longitudinal Dimension ........ Sampling .................... Criteria for the Sampling Pool ....... Sample Stratification ............ Characteristics and Representativeness of the Sample . .............. Effect of Sampling Procedures Upon Representativeness ........ Response Rates .............. Characteristics of Respondents ...... Preparation for Data Gathering ......... VQuestionnaire Construction ......... Interviewer Recruitment, Training and Supervision .............. Recruitment ............... Training and Supervision ......... The Data—Gathering Process ........... Obtaining Access to Marriage Records Initial Contacts with Couples ........ Phase I Telephone Interviews ........ Phase II Personal Interviews ........ Preparation and Analysis of Data ........ IX TESTS OF THE “EQUALHINPUT“ THEORIES ........ Construction of Marital Power Indices ..... The Unweighted Marital Power Index ..... The Marital Power Index Weighted by Task Importance ............ Interpretation of Tables ............ Tests of Perceived Competence Hypotheses . . . . Hypothesis C: General-Competence + OVerall Marital Power ............... Hypothesis C-l: Dyadic General-Competence + Overall Marital Power ........... Hypothesis C—2: Task—specific Power + Use of "Competence” as a Rationale ...... vii Page 125 125 125 126 126 127 128 128 132 134 134 137 139 146 146 148 148 150 151 151 152 155 156 158 160 161 161 162 162 163 163 167 170 Rmof Hypot Mar 1 TESTS OF PG Part 1. Invest Hypol Mal Hypm 0w Part II Hypo i 4 RESULTS OI Compete Tes h les c Equal- les S les l Tes t C Compet Tes I Chapter Test of Gratitude Hypothesis . . . . . . Hypothesis G-l: Status Mobility + Overall Marital Power . . . . . . . . . . . . . X TESTS OF POWER-DEPENDENCE THEORY Part I. Alternatives and Motivational Investment . . . . . . . . . . Hypothesis D—l: Alternatives + Overall .Relative Alternatives + Marital Power Hypothesis D—2: Overall Marital Power Motivational I Hypothesis D-3: Overall and Task-specific Marital Power Hypothesis D—4: Relative Motivational Investment + Overall and Task—specific Marital Power . . . . . . . . . . . . . Hypothesis D-5: Relative Alternatives + Relative Motivational Investment . . . Part II. Power Use Across Multiple Issues Hypothesis D—7: Number of Areas of Disagreement + Amount of Power Exercised Hypothesis D—8: Average Discrepancy in Motivational Investment e Amount of Power Exercised ........... Hypothesis D-9: Average Discrepancy in Preferences + Amount of Power Exercised . nvestment + XI RESULTS OF CRITICAL TESTS BETWEEN THEORIES Competence versus Gratitude Theories ...... Test G-C l: Relative Competence and Relative Income + Overall Marital Power ...... Test C—G 2: Competence and Income ........... Equal-Input versus Power—Dependence Theories . . Test GC—D l: Relative Effort + Task—Specific ooooooooooooooo “Relative Effects” of Satisfaction Test GC-D 2: Task—Specific Marital Power + Task-Specific Satisfaction ........ Test D—GC 3: Relative Potential Alterna- tives and Relative Current Alternatives + Overall Marital Power ........... Competence versus Power—Dependence Theories . . Test D-C 1: Relative Competence and Relative Dependence + Overall Marital Power viii Page 175 175 178 178 178 182 187 197 207 214 216 217 219 223 224 224 228 229 230 233 234 239 239 Test Cm Test Tests I Compete Power-D Test Mc Ca Compete Tesi E 01 III SUMMARY AI HHNCB L LETTERS S t PHASE I. I. PHASE II. A LISTS ACC I, PHASE II. somemvl . . , Chapter Test DFC 2: “Relative Effects” of Competence and Dependence . . . . Test D—C 3: The Role of Task Importance Tests Involving Specific Resources . Competence and Gratitude versus Power-Dependence Theories . I Test D-CG 2: Relative Income and Wives} Motivational Investment in Having a Career — Wives‘ Overall Marital Power . Competence versus Power-Dependence Theories Test C—d 4: Relative Age, Marital Experience, and Having Children + Overall Marital Power . . . x11 SUMMARY AND CONCLUSIONS . . ; . . . APPENDICES A. .LETTERS SENT TO COUPLES ......... PHASE II. PERSONAL INTERVIEW (PARTS A AND C) . . . LISTS ACCOMPANYING PHASE II QUESTIONNAIRE ..... PHASE II. PERSONAL INTERVIEW (PART B) ...... m C ('3 US . o c o BIBLIOGRAPHY ................... I ..... PHASE I. TELEPHONE INTERVIEW ........... Page 247 248 252 253 253 265 265 271 287 291 318 349 356 '362 file I I Rank order of 1 according to o and husband's I Response patte telephone inte 3 Response patte personal inter 1 Characteristic type: Phase II S Hypothesis C: perception of E Hypothesis C-l perceived corny Hypothesis C-I within each I! marital power rationale for divided. . , Hypothesis G- OCCURational I father . , . I Hypothesis 0. alternatives ‘: HIpothesis D~ Relative [Me I HIPOthesis o- evel of moti I Hypothesis n. by IEVe] of ii 0Utcone , . ‘ Table LIST OF TABLES Rank order of wife's relative marital power according to comparative education of spouses and husband's occupational status . . . Response patterns by marriage type. Phase I telephone interviews . . . . Response patterns by marriage type: Phase II personal interviews . Characteristics of respondents by marriage type: Phase II personal interviews . Hypothesis C: Overall relative marital power by perception of the other spouse's general competence . . Hypothesis C-l: Overall relative marital power by perceived competence ratings for both spouses . . . . I Hypothesis C-2: Per cent of husbands and wives within each level of task-specific relative r marital power, who cited competence as a rationale for the way contested—tasks were divided . . ' .................. '. . . . Hypothesis G-l: Overall relative marital power, by occupational status of spouse relative to that of father ........................ Hypothesis D— l: Overall marital power by potential alternatives to the present marriage .......... Hypothesis D—2: Overall relative marital power by relative potential ”alternatives” . .......... Hypothesis D-3: Overall relative marital power by level of motivational investment in the marriage Hypothesis D-3: Task-specific relative marital power by level of motivational investment in task~division outcome ......................... Page 138 139 140 166 . . 169 . 172 195 'ule a Hypothesis 0-43 by relative moi married. . - . Hypothesis D-4 payer by relat‘ spouses in tas Hypothesis M by average rel spouses in tas Hypothesis 0-5 by relative al 7 Hypothesis D-i by number of a differed . . Hypothesis D-l by average di‘ task-division Hypothesis D-‘ by average si task-division Test C-G l: perceived con vhen relative Test GC-D l: (as reflectec relative effr Test GC-D 2; IRS reflectei relative powy lest Rec 3; IEIatIVe POt income of Sp lest he i: IEIfitIYe dep 0f130th spou Table 22 23 24 Page Hypothesis D—4: Overall relative marital power by relative motivational investment in being married . . . . . . . . . . . . . . . . . . . . . . . . . 199 Hypothesis D-4: Task-specific relative marital power by relative motivational investment of spouses in task-division . . . . . . . . . . . . . . . . 201 Hypothesis D—4: Overall relative marital power by average relative motivational investment of spouses in task—division . . . . . . . . . . . . . . . . 205 Hypothesis D—5: Relative motivational investment by relative alternatives . . . . . . . . . . . . . . . . 209 Hypothesis D—7: Overall amount of power exercised, by number of areas about which spouses' preferences differed . . . . . . . . . . . . . . . . . . . . . . . . 220 Hypothesis D—8: Overall amount of power exercised, by average difference between spouses in strength of task-division preferences . . . . . . . ......... 220 Hypothesis D—9: Overall amount of power exercised, by average size of discrepancy between spouses' task—division preferences . . ............ . . 220 Test C—G 1: Overall relative marital power, by perceived competence ratings of both spouses, when relative income is controlled ........... 225 Test GC-D l: Perceived fairness of task division (as reflected in satisfaction), by perceptions of relative effort invested in the marriage ........ 232 Test GC-D 2: Perceived fairness of task division (as reflected in satisfaction), by level of relative power in specific task areas .......... 235 Test D-GC 3: Overall relative marital power, by relative potential alternatives, when relative income of spouses is controlled ............. 237 Test D—C l: Overall relative marital power by relative dependence and by competence ratings of both spouses ..................... 24] xi ble 1; Overall relatii relative income TestD-CG 2: I ofvives, by I“ vives' motivat‘ career is cont Test C-D 4: 0 relative age a and previous e Table 25 26 Page Overall relative marital power of wives, by relative income of spouses . . . . . . . . . . . . . . . 256 Test D—CG 2: Overall relative marital power of wives, by relative income of spouses when wives' motivational investment in having a career is controlled . . . . . . . . . . . . . . . . . 257 Test C-D 4: Overall relative marital power, by relative age and previous marital experience and previous experience with children . . . . . . . . . 266 xii Tiyure ..r. C>o Blood and R as applied behavior . Blood and I Illustratit measuring 1 Relationsh' marital p01 Illustrati single res Calculatio for indivi Comparison of Blood a l972) . , Task 'ITTIDOI upon mari‘ Predicted and to p01 PI Rtevio ChIIdren Overview 3amPli ng Figure LIST OF FIGURES Blood and Wolfe's conceptualization of power, as applied to the explanation of Observed behavior . . . . . . . . ........... Blood and Wolfe's operationalization of power . . . Illustration of different meaning contexts in measuring power, as implied by Blood and Wolfe . . . . Relationship between selected ”resources" and mean marital power of husbands in four countries ..... Illustration of power-score calculation for a single responsibility-item .............. Calculation of weighted relative power scores for individual responsibility-items ......... Comparison of major variables contained in theories of Blood and Wolfe (1960) and Berger et a1. (1966, 1972) ........................ Task importance level and its predicted effect Page To 12 13 i I 27 1‘ 43 44 6T upon marital power ................... 117 Predicted marital power, according to competence and to power-dependence theory for the resources of previous marital experience, age, and having children from a previous marriage ........... 123 Overview of the Panel Design .............. 125 Sampling Design .................... 133 hue II. N lllustrati on 0 relative 11614le 3 Summary Of 59' Figure 12. 13. Illustration of the method used to measure relative marital power . . . Summary of selected findings ......... .. xiv Page 273 278 In the pres: elusive and metaphor enseinwhich It 15 olitics may be tran etaphoric sense. 1 :h’ierent from what Power of or Eymn-self-satisfy' tuned couples are :1 consider the wel in that power is or. The question spouses engage in e 1e partner benefit And, one 1 I “relative advani Niles like the ri 5N Is the questii III Patterns gets 1101' routines 10 av persist for ye CHAPTER I INTRODUCTION In the present study, I take a critical look at the perhaps ehwive and metaphorical concept of marital power. “Power“ in the sense in which it is used in the study of community and national politics may be transferred to the analysis of families only in a netaphoric senSe. That is, what we mean by marital power is somehow different from what we usually mean by power-in—politics. Power of one spouse in a marriage, in my view, is evidenced a by non-self-satisfying behavior on the part of the other spouse. But married couples are supposed to engage in non—self-satisfying behavior, to consider the welfare and desires of the other spouse. Does this meM1that power is a meaningless term for analyzing families? I think not. The question of relative advantage still remains, i.e., do spouses engage in equal amounts of non—self-satisfying behavior or does one partner benefit more than the other one? And, one might ask, in what areas of marpied life is the study of ”relative advantage“ useful and appropriate? For newly married couples like the respondents in my study, it seems that one critical area is the question of how the structuring of one's joint, everyday life patterns gets accomplished. Such patterns-~including the establish— ing of routines for dividing up the daily tasks involved in joint living—- may persist for years to come, once established. All couples must deal so these questions are groups of coupl imre ideosyncrati issues” between spou aybevhether to re another couple a cri tesacareer; identi adopted in the pre In a marita gee" seems to be it exercise of power. :runequaI, i.e., w relative advantages :eoseiul for the s NEW it's ex ateIi tries to redo The preser I the SWAT 01' mar «Hi the so-called Item: to me, has :eoretieai /exp1 an: contusion, e. 9., al :Nbiemg or to Sue :i female responds, ultural contexts, I Rover. with these questions; hence there is some sense in which one may com— pare groups of couples. This has certain advantages over the study ofnwre ideosyncratic--although doubtlessly important--”critical issues“ between spouses (e.g., a critical power issue for one couple may be whether to raise a child as Protestant or Catholic, while in another couple a critical issue may revolve around whether the wife has a career; identification and comparability of such issues was not attempted in the present study). In a marital context, the assumption that power is a ”zero—sum gamE'seems to be inappropriate. Both spouses may gain from the exercise of power. The question is whether or not this gain is equal or unequal, i.e., whether one has a relative advantage. Because such relative advantages are being gained, the metaphor of “power” can still be useful for the study of marriages. One only must be careful not to g§§pmg that it's exercise is intentional, i.e., that one spouse deliber- ately tries to reduce the benefits of the other. The present study begins with the examination of one tradition in the study of marital power, the research program which has focused upon the so-called ”Resource Theory of Marital Power.“ This program, it seems to me, has reached an impasse, a point at which no further theoretical/explanatory progress is evident. It is unclear whether the confusion, e.g., about contradictory findings, is due to theoretical problems or to such methodological shortcomings as the predominant use of female respondents, the inability to identify critical differences in cultural contexts, or the failure to find culturally equivalent measures of power. I suggest is Heretical unclarit Bout intervening co neliterature, sev .11th power proce uplicated, drawing setof hypotheses in ponmarital power entries as vehicle respondents. Hypotheses ing in mind. Where Iiiferent theories terelative explan tensabmitted to ; iiie pairs from a 5 hair first and sec of presently emplo iifferent populati and interpretation I suggest that a good part of the problem is indeed one of theoretical unclarity, a failure to explicitly identify assumptions about intervening cognitive processes. Upon careful examination of theliterature, several alternative and/or complementary theories of marital power processes are identified. These theories are then explicated, drawing upon related social psychological theories. The set of hypotheses which is developed in this exploratory study focuses upon marital power processes pg§_§g, as opposed to prematurely using theories as vehicles for explaining behavior in a particular sample of respondents. Hypotheses are constructed with this purpose of theory test- ing in mind. Wherever possible, test situations are developed in which different theOries would predict different results, in order to assess the relative explanatory poWer of several theories. Hypotheses are then submitted to preliminary testing, using data from 90 husband—and- wife pairs from a sample of upper-middle class Michigan newlyweds in their first and second marriages. Suggestions for future refinement of presently employed theories and for testing these theories with different population groups are incorporated during the presentation and interpretation of findings. The presen influential tradi ti Resource theory of peer differs from The The resou primarily in quanti HIIIEI‘ than in qual bed,e.g., as legi pier-bases (Ravenl Family ini HSsession and mar‘ iirectly studied. “Mo (1963), Lei iorvhom interacti Isthe"critical s arital power IS S Interaction scho CHAPTER II THE RESOURCE THEORY OF MARITAL POWER: A LITERATURE REVIEW The present study takes as its point of departure the most influential tradition in the study of marital power, that known as the 'Wesource theory of marital power." This approach to the study of power differs from other traditions in the field in several ways. The Resource—Theory Tradition in Context The resource theorists view marital power as having its base primarily in quantifiable resources such as income and education, rather than in qualitative ways in which resource—combinations are used, e.g., as legitimate, expert, reward, coercive, or referent power—baSes (Raven, Centers, and Rodrigues, 1975). Family interaction is assumed to intervene between resource possession and marital power, but is neither given special emphasis nor directly studied. This contrasts with such works as Kenkel (T957), Caputo (1963), Leik (1965), Farina (1960), and Mishler and Waxler (1968) for whom interaction processes are not only directly studied but are seen as the “critical stage“ (Turk, 1975) or dimension of power. That is, if marital power is seen as emerging through some choice-making process, the 'Wnteraction school“ assumes that power can best be identified by measuring interaction1 during need as not "reall results of power-in- sdmol defines the mice-making proce rim when spouses di boisions for the f bdecides who wil The resour inriew power as t ridial wants (altho individuals want, e iaily decisions). presiding in the recess, e.g., in t are the powerful 1 once-making and 1 Hoover, one may b Slitific suggestio ‘ 1For exam Anticipation duri fiber of discussi 111111.1960), or ‘Ithegroup (Bloo Wer-during-inte I-I-s some take “ AIRS do not. interaction1 during the decision-making. The decision-outcomes are viewed as not ”really“ being power but only the results of power, i.e., results of power-in-interaction. In contrast, the resource-theory school defines the critical stage as occurring near the end of the dwice-making process; power is measured in terms of the outcomes--who whw when spouses disagree or who is allocated power to make final decisions for the family as opposed to who decides who will decide or who decides who will win (Safilios-Rothschild, 1970). The resource—theory writers are among those social scientists who view power as the ability to attain outcomes congruent with indi- vidual wants (although they generally assume rather than measure what individuals want, e.g., people are assumed to want the power to make fley decisions). The major competing view in the field sees power as residing in the person who makes the critical contribution to choice process, e.g., in the person who participates most or who speaks first. Here the powerful person is the one who has an impact upon group choice-making and hence (presumably) upon the outcome of choice-making. However, one may be powerful in this sense without always having one's specific suggestions accepted by the group (Turk, T975). 1For example, power may be operationalized as total amount of participation during group decision-making discussions (Caputo, 1963), number of discussions in which individuals speak first and/or last (Farina, 1960), or who first suggests the idea which is finally adopted by the group (Blood and Hamblin, T958: 347). Researchers who measure power-during-interaction do not share a common theoretical orientation, e-9-,some take “resources” into account as an independent variable and others do not. Tape The resourc :ietiodological an starting point for pets the nature of pnpective which i snilable to men an nioible contributi ibis theory's va sociologists are iu their students a Second, th per field for its studies were replic aid to adapt and r tested to empirical aesnot share this researchers even us :1 knowledge and ti be A few majr imrized. Ihav decry itself. Th 11 the resource th Importance of the Resource Theory The resource theory literature, although beset with its share ofnmthodological and theoretical shortcomings, provides a useful starting point for the exploration of marital power. First, it inter— mets the nature of contemporary marital power from an ideological perspective which is widely accepted, namely that power is equally avaflable to men and women, and that whichever spouse makes the greater tangible contribution to the family ”earns” the greater power. Tests of this theory's validity would provide some guidance about whether sociologists are justified in continuing to offer this interpretation 1 to their students and readers. Second, the resource theory work is notable in the marital power field for its history of systematic theory development. Early studies were replicated with a variety of populations, the findings used to adapt and refine the theory, and the revised theory again sub- jected to empirical testing. (In contrast, the interaction—tradition does not share this history of continuing development. Seldom have two researchers even used similar measures of power inhibiting the accumulation of knowledge and the development of a theory.) Development of the Resource Theory A few major studies in this research program shall now be summarized. I have selected works most central in the evolution of the theory itself. The following discussion begins with the earliest work on the resource theory and follows the development of the theory over ye. Special atte inputs of this th this behreen them. pitIIIC resources phasis upon the c in of the theory. iIuodIIIoIfe (1960 Blood and past-faction expl an riosvith 909 wiv (mater Detroit or The resea usocial class in pets such socio-es status, income, on possession of whici her in the marit otained by spouse larger the advanta external-system re in uorital power. usources Resource: and education but the available to time. Special attention will be given to treatment of the central dements of this theory--resources and power—-and to the hypothesized links between them. Empirical findings about the association of specific resources with marital power will then be summarized, with emphasis upon the contradictory findings which necessitated a reformula- tion of the theory. My critique of this work follows in Chapter III. Blood & Wolfe (1960) Blood and Wolfe offer resource-theory or marital power as a post-factum explanation for certain unanticipated findings from inter— views with 909 wives, a random sample of all intact families in the greater Detroit area. ”The researchers found that dyadic marital power was related to social class indicators. Their ”resource theory of power” inter- prets such socio—economic variables as formal education, occupational status, income, and organizational participation as ”resources," possession of which forms a base for possession of marital power. Power in the marital dyad (”internal system”) has its base in resources obtained by spouses from the I‘external system“ (Blood, 1963: 475). The larger the advantage of one spouse over the other with respect to such external—system resources, the larger his or her corresponding advantage in marital power. Resources Resources in principle were not limited to factors like income and education but were thought to include ”anything that one partner may make available to the other, helping the latter satisfy his needs or pain his goals" (B .111 preschool chil p11 investment of is taken to mean t chars interpreted iiuternal as comp shstrain the wife trial resources a weris low at thi If resourc pier achieve goal- iluod and Holfe ass spouses seek in ma ‘—-—-—~_ Zln a late interpretation some Husband and wife ii arital power _i__r§_q_i pueming transact iidHolfe's decisii aspirational systei Hichases (car, in: be decision about this the external authors-all di rec 3isiooa a less explanatory p pntribute to need paternal resour bier theory devel hirer-dependence i attain his goals" (Blood & Wolfe, l960: l2). The finding that wives wiflipreschool children had the lowest marital power, at a time when their investment of effort and other ”internal“ resources is high, was not taken to mean that the theory excludes such resources. Rather, the authors interpreted this as evidence of the greater explanatory power of external as compared with internal resources.2 Because children constrain the wife from obtaining external resources, and because ex- ternal resOurces are more “potent” than internal ones, the wife's poweris low at this stage of the family life cycle.3 If resources are anything by which one spouse may help the other achieve goal-satisfaction, one might ask what goals are involved. Mood and Wolfe assumed that the primary personal goals which U.S. spouses seek in marriage are sexual fulfillment, emotional response, 2In a later article, Blood (l963: 475, 476) modified this interpretation somewhat. He noted that relative participation of the husband and wife in the external system will determine the balance of nwrital power insofar as marital power is measured in terms of decisions governing transactions of the family with the external system. Blood and Wolfe's decision—making scale included two transactions with the occupational system (husband's job and wife's employment status), five purchases (car, insurance, housing, food, and medical services) and one decision about where to go on a vacation. All may be said to deal with the external system to some degree and-—a point not noted by the authors--all directly or indirectly involve money. 3Blood & Wolfe did not specify why internal resources have less explanatory power than external ones. Such resources certainly contribute to need and goal satisfactions of the other spouse, just as do external resources. Such unanswered questions proved an impetus for later theory development (e.g., by Heer, l963; see my discussion of power-dependence theory in Chapter VI). nuanionship, the n: standard of living. indicator of power ( ions and types of p aerial resources a Power, for anility of one partn inherit) is a SubSI :ecnnse both agree Authority is thus :nln specified that areas that Blood 8. allowing types (se war (or authority :n‘ observed behavi ( \ ~ One w0ul. ifstrongly ass 0ci annotation was f0 MS“tlsfactions ill “UmPanlOnshjp nllesl resoupces w companionship, the new experience of parenthood, and some particular standard of living.4 They also assumed that decision-making is a good indicator of power (although presumably recognizing that other dimen- flons and types of power do exist), since decisions allocate future nmterial resources and hence affect outcomes and goal achievement (ibid.: 1]). nger Power, for Blood & Wolfe, is defined as “the potential ability of one partner to influence the other's behavior“ (ibid.: ll). Authority is a subset of power, namely that power held by one partner because both agree that it is proper for him or her to have it. (Authority is thus more than Weber's "legitimate power,“ in that Weber mfly specified that the less-powerful person must hold this view.) It seems that Blood & Wolfe would divide behavioral outcomes into the foHowing types (see Figure l below). Note that potential but EDEEQQ power (or authority) is excluded, since I am focusing on the explanation of observed behavior. ——-——-—-—,—————_—_—_ 4One would expect, therefore, that the husband's resources would be strongly associated with the wife's satisfaction in these areas. Some association was found between husband's education and income and certain goal—satisfactions of the wife, such as satisfaction with affection and with companionship received. Husbands' satisfaction and its relation to wives‘ resources was not studied because husbands were not interviewed. . . ill behavior drvndec the result of influence ”’7'— W here there is cons nvho has the rngh aeroise influence AUTHORITY W— 1‘.l outcomes due to PONER the eXP' How did 8' Trelrindicators wv weighted scores to nth the higher in ~33 considered to \ '6 autv EM“ were, in Instill) deiS for E nnhcompm-Sed ti RH behavior divided into that which is: the result not a result of influence of influence I I J per—4’ ¥—_W_—‘J Where there is consensus Implied to exist (e.g., where , on who has the right to behavior is due to considera— exercise influence tions of convenience, or to time constraints, to personal AUTHORITY preferences of the actor, etc.) AH outcomes due to influence POWER ngre 1. Blood & Wolfe's conceptualization of power, as applied to the explanation of observed behavior. How did Blood & Wolfe operationalize this concept of power? Their indicators were the following decisions (ibid.: 282), with un- weighted scores for each combined into one index score.5 The spouse with the higher index score, i.e., the one who made more decisions, was considered to be the more powerful of the two. ———-———.—_—_ 5The authors made no attempt to determine how important these decisions were, in the opinion of the wives they studied. Hence, there was no basis for assigning unequal “weights" to the several decisions which comprised their power index. lnevery family unfly will liV Ilrstn bUt the orthe wife. out a) what ca z:- ) whether some li what hc ment to r) o. wMtj should whethe should quit w (I) —O~. how mu family spend what d someon u: :- about vacati Answers tc p6 _ en hahng decis Ttwodecides mug 359that Person e . wbsepower) films include c ”‘ Wanlndex and ‘ “(“hpower as b ‘UiSCUSSlon of th ll In every family somebody has to decide such things as where the family will live and so on. Many couples talk such things over first, but the final decision often has to be made by the husband or the wife. For instance who usually makes the final decision about a) what car to get? b) whether or not to buy some life insurance? c) what house or apart- (answer categories) ment to take? (5) Husband always d) what job your husband (4) Husband more than wife should take? (3) Husband and wife exactly the same e) whether or not ygu_ (2) Wife more than husband should go to work or (l) Wife always quit work? f) how much money your family can afford to spend per week on food? 9) what doctor to have when someone is sick? h) about where to go on a vacation? Answers to these questions are assumed to represent power it- self6 (making decisions equals exercising POWEV-) Almost by definition, one who decides must be exercising power over the other's behavior, be- came that person allocates resources which affect goal-satisfactions. _-——-———_—._ 6The decision—index of marital power also implies that a power is assumed to be zero sum in nature (if one spouse gains power, the other mustlose power). In part, this is a methodological problem. Other such problems include criteria for determining which decisions to include in such an index and whether specific decisions can be assumed to indicate asnmch power as broader decisiOns. See Safilios-Rothschild (l963) for a discussion of these and other related problems. Sim iNoife furtht NMJMMHMA mira decision-m mmwcecontributi Their oper mmwtmi deiiniti fimfl man “abiiit mmMem variabie éflbmafior divide That fami' reson +——.__.__ W Decisions A i #5 POWER 3'5 behav j“--——-—..___ fiwez. Biood a If "Power ifivmuais, the q \\ 7They dic Wmar1t31yp0w6 pes(Where one ”he highesti 3t differ Slgmfi ‘3“?wa that up_ ”6de disapm 0%] Jbet it 12 Blood & Wolfe further assume that, among contemporary United States couples, traditional patterns of authority no longer determine who acquires decision-making power.7 Power must be “earned” by making resource contributions. Their operational definition of power thus differs from the conceptual definition (see Figure l). “Allocation of decision-making”, rather than ”ability to influence behavior of the other“, is now the dependent variable or thing—to—be-explained. All behavior divided into: That presumably due to family decisions about Other. (excluded resource-use from consideration) ‘VW ' W l' *1 Decisions made by Decisions made by A B t A's POWER over B's POWER over B's behavior A's behavior Figure 2. Blood and Wolfe's operationalization of power. If ”power”is operationalized in terms of decisions made by individuals, the question now becomes one of accounting for patterns 7They did not measure authority directly. However, they com- pared marital power patterns of older, farm, Catholic, and immigrant couples (where one would expect traditional authority of the husband to be the highest) with other couples. Since mean husband-power did not differ significantly among these groups of couples, Blood & Wolfe concluded that “patriarchy as a cultural norm is either already gone orrapidly disappearing“ (ibid.: 27). oldecision-making. slc'n decision-patte lens, i.e., resour tion (see footnote resources. However nomtic from nonr index measure of 'me :0 have power under ll; SHOULD THE DECISION BE; Spouses agree K‘— Spauses disagree 3'5 1deas accept ..lléileaégggggg ..l.llr9li§§_._ -------- W o A J defmeda 7. Bdeflned a X as ‘impl The Opera ranging from COmp' :erllete disagreel fit, 13 of decision—making. Blood and Wolfe saw only two possible bases of such decision-patterns, traditional norms and “pragmatic” considera- tions, i.e., resources.' They eliminated tradition as a viable explana— tion (see footnote 7) and hence assumed marital power must be based on resources. HoweVer, they provided no way to empirically distinguish pragmatic from normatively based decision—power. Hence their decision— index measure of marital power would consider the person-who-decides to have power under all of the 24 situations outlined in Figure 3. WHO SHOULD MAKE THE FINAL DECISION: (Is determined (Not determined by norms b Spouses disagree Spouses agree Spouses agree WHAT SHOULD THE A B A B DECISION BE: decides decides decides decides . /, /, Spouses agree ////l 5 , 9 l3 l Spouses disagree A's ideas accepted 2 6 l0 l4 B's ideas accepted 3 7 ll l5 Compromise 1// 4 8 l2 l6 7” A defined as having power I B defined as having power Figure 3. Illustration of different meaning contexts in measuring power, as implied by Blood and Wolfe. ' The operational measure of power, therefore, includes situations ranging from complete agreement of spouses (cells 9, l3, l7, and 2l) to complete disagreement (cells 2, 6, 3, 7, 4, and 8). In additio final decisions, ev ii, 3, 4) or when i]. In short, hanbe said to he . niiferently than B i, and l9). "Powc and narrower than ' n includes cases n influence and new ing about the u Iwill re ills Point, it is inner as the abili into to have th ‘iiillltu to afi \ 8 -n They din Loser to their C( :19 other's behav- if: h5°mething, . it eusuan Emmy: We aiec‘. fidw the ones T‘iiidwhether , am“ When. .—.— 14 In addition, A is considered to be powerful whenever A makes 'final decisions, even in cases where A does not want to do so (cells l, 2, 3, 4) or when A decides to do what B wanted (cells 3, ll, and 19L In short, if power is equated with who usually makes decisions, A can be said to have power even when A does not influence B to act differently than B would have in the absence of A (cells l, 3, 9, ll, l7, and 19). l‘Power” as operationally defined is thus both broader and narrower than the conceptual definition of power (broader because itincludes cases where behavior cannot be said to be directly due to influence and narrower because it is limited to allocation of decision- making about the use of resources). I will return to the problems of defining power later. At fifis point, it is sufficient to note that Blood and Wolfe do not treat power as the ability to influence another‘s behavior8 but simply as the ability to have the final say about resource—use (and hence presumably the ability to affect goal-outcomes of both spouses). 8They did include a second measure of marital power which is closer to their conceptual definition of power as the ability to influence the other's behavior. The question was: ”When you and your husband differ about something, do you usually give in and do it your husband's way, or does he usually come around to your point of view?” (ibid.: 23). Unfor- tunately, we are not told whether husbands who win such contested decisions are also the ones who score high on the decision-index measure. Nor are we told whether resources would correlate with this second measure of marital power. :‘m‘iy. One who resourceful and co niing as such. P in external resou cars, taxes, etc. , raking decisions ( associated with co arital decisions cowetence of a hi the other, who de Community peer. The wife p lusbaud in the eye lisuife. tonmuun‘ 33), which may fu‘ inefit by sharin aerial advantag ‘usuch a high-re Blood an rental power. ( neource SPOUSE ll The Link Between Resources- and Power Resources were thought to serve as bases of power in several ways. Such ”resources" as education, occupational status, income, and organizational participation allow one to make a contribution to the mely. One who makes such a contribution is, almost by definition, a resourceful and competent person, implying special skills in decision- nwking as such. Participation in the outside world, involved in obtain— ing external resources, may also yield specific knowledge about politics, cars, taxes, etc., as well as experience in taking responsibility and waking decisions (ibid.: l3). External system resources are thus associated with competence—~skill and knowledge-—which are relevant to nmrital decisions about transactions with the external system. The competence of a high—resource spouse is assumed to be recognized by the other, who defers to his or her greater knowledge. Community prestige standards can also influence domestic power. The wife presumably shares these standards and so a “successful“ husband in the eyes of the community receives also great respect from his wife. Community prestige gives the husband self—confidence (ibid.: 30), which may further increase the wife's respect for him. Wives benefit by sharing the prestige of their husbands, as well as the material advantages he provides. Wives are “indebted“ and "grateful“ to such a high—resource husband and willingly defer to him on decisions. Blood and Wolfe stress that coercion is not characteristic of marital power. Only once do they mention the possibility that a high- resource spouse might obtain power by withholding resources (ibid.: l3); eithe occurrence 0 new], resourcefu' their wives by tt goal-satisfacti ons . house satisfaction uneconomic aspect clear whether the s for both genders . u'uh resources. of irical Findings Blood and iuouith the relat her. The first 9High-resr in power by anotI ui'douinant wives' heseuives less on Imoetent that w uilih'es. In genern tier froma giVen the to failure hittional author list). At any rat tfwunt for mari ta “"4 inprovement 10The wor ihoses. If persc porter relative i has more resourr 16 mm the occurrence of physical violence in families is ignored; in 9 general, resourceful husbands are assumed to earn the willing deference oftheir wives by their contributions to the family and to the wives' goal-satisfactions. Exactly how material resources and competence in- uease satisfaction of sexual, emotional, companionship, and the nmreconomic aspects of parenthood goals is not specified. Nor is it dear whether the same resources are expected to "work” the same way forboth genders. Presumably, wives can also ”earn” power by obtaining high resources. Empirical Findings Blood and Wolfe tested two types of empirical hypotheses deal- lO ing with the relation between external resources and relative marital power. The first was of the type: the higher the income level of 9High— resource wives, on the other hand, apparently earn their Mgh power by another means. In view of the low marital satisfaction of"dominant wives”, Blood and Wolfe assume that husbands defer to flmse wives less out of respect and gratitude than because they are so incompetent that wives have no choice but to take over decision responsi- ilities. In general, Blood and Wolfe's wives appeared to accrue less to power from a given amount of resources than did husbands. This might be due to failure to measure all relevant resources or to lingering traditional authority (which Blood and Wolfe assume to be a thing of the past). At any rate, differences in the resource theory's ability to account for marital power of wives vs. husbands points to the need for theory improvement. 10The word “relative” indicates comparative positions of two spouses. If person A has more power than person B, A is said to have greater relative power. Likewise, A has greater relative resources if A has more resources than does B. tions between h lust of these h The hi (reasoned as hi helrigher thei prestige was no . than non-employ land not employ one than full- Nun-....- 1 1Ther Jenn power than is examined so tires. the same here the wife collar husband: ”A . the 413 emplo; tilled wives w basis for deci then as for m l7 husbands the higher the relative marital power of husbands. Second, and more in fit with their theory, were hypotheses in which resource levels of both spouses were considered, e.g., the greater the educa— tional advantage of husbands over their wives, the higher the relative marital power of husbands. Because their data was not gathered for the purpose of testing the resource theory, Blood and Wolfe apparently lacked complete informa— tion about wives' resources. At any rate, most of their hypothesis tests were of the first type, in particular those dealing with associa- tions between husbands' resources and husbands' relative marital power. Most of these hypotheses received support. The higher the husband's resource of occupational status Measured as high and low white collar and high and low blue collar), the higher their mean powerH (ibid.: 30, 3T). Wives' occupational prestige was not considered; however, employed wives had higher power than non-employed wives within each of the following categories: hus— band not employed, husband employed full—time, and husband employed l2 ( more than full-time ibid.: 40). An additional indicator of work 1]There was one exception: high blue-collar husbands had lower mean power than did low blue-collar husbands. When this relationship was examined separately for couples having employed and non-employed wives, the same pattern was observed for the non-employed wife cases. Where the wife was employed, however, low blue-collar and high blue— collar husbands had equally high mean power scores. 12Apparently, no husbands were employed only part-time. There were 4T3 employed husbands with non—employed wives but only three em— ployed wives whose husbands were not employed. Hence, one has little basis for deciding whether resource-advantages have the same effect for women as for men. garage, was intensely rel; Hush; mMMw host and hi houuue was on groups accord husbands was income alone, m was rith very loo lire power ti Imds' educa' nu” (no For eiuntion le education l8 participation, the number of years the wife had been employed since marriage, was directly related to wives' relative power (i.e., was inversely related to husbands' relative power) (ibid.: 4T). Husbands' jpggmg resources were directly related to husbands' mean relative power. This effect was especially pronounced in the lowest and highest income levels (ibid.: 31, 32). The effect of wives' income was only indirectly examined. When couples were divided into groups according to ggpal family income, the mean relative power of husbands was lower than when couples were divided according to husband's ( income alone, at least for high-income couples13 (ibid.: 32). Wives' {J education was related to marital power in a curvilinear fashion; wives ' with very low education and with very high education had higher rela- tive power than did wives with intermediate levels of schooling. Hus- bands' education, in contrast, was directly related to husbands' mean power'4 (ibid.: 28). n Fortunately, Blood and Wolfe provide data about comparative education levels of spouses. Wives' relative power was related to their education relative to their husbands' education. As predicted, wives “—— _ 13Perhaps this occurred at all income levels. Blood and Wolfe did not present the specific findings but only mentioned drastic decreases in husbands' power at the two highest income levels. I4Husbands with less than six years and with T6 or more years of schooling deviated somewhat from the trend. High—education husbands, al— though they ”lose“ a little power, never approach the lower power level oflow-education husbands. High-education wives, in contrast, have almost the same power scores as those with the lowest education. tithe highest me last five years an in their husband narrative educat paler (ibid.: 38). supational statu Table i. Rank or compare status . l hula order of wif arital power* by occupational stat High white-col Lou white-col in High blue-coll ‘l= highest wife ~-——————— The largest diffe roan power-score W 1Stars 1 iii. Ihave or her of cases. Tl lug deviations in advantage had Te: duration husbann ital-education 19 had the highest mean power when they had an education advantage (at least five years more schooling than their husbands), the lowest power when their husbands had an education advantage, and intermediate power scores when spouses had about the same education (ibid.: 37). Table l allows an examination of the combined effects of comparative education and husband's occupational status upon marital power (ibid.: 38). In general, husbands' mean power increases with occupational status within each education—combination as predicted. Table l. Rank order of wife's relative marital power according to comparative education of spouses and husband's occupational status. Comparative Education Wife More Egual Husband more Rank order of wife's mean marital power* by husband's occupational status High white-collar 6 7 9 Low white-collar 3' 4 <%-—-———9 8 High blue-collar l 3 5 *l = highest wife power The largest differences between adjacent cells in Table l (in terms of ' mean power-score differences) are marked with arrows. The “loss” of 15This table is adapted from Blood and Wolfe, l960, Table l0, p. 38. I have omitted low blue—collar couples because of the small num- ber of cases. This lowest-status group, however, produced some interest— ing deviations from the general trends. Husbands with an education advantage had less power than those with equal education. Also, equal- education husbands in this group had more, rather than less power than equal—education husbands in the next-highest occupational status group. men by wives, a hen wives worki In gener ledirectly assor the work parti c' hie-employed 0‘ nears the wit :uband's occupa :ipation of spou aritai power, a liith mi nanlicated in se inirance, of Sr bionic H957) - certain resouro rite Figure 4 f it broaden the inn, “55 subje \— l6I anti" in Amer“ ”it’ll“ f0)” eXE ten added Otiw nich seemed tr hid] inthe U { 9" SelEcted . :8th Vahd ti Wiles. Thu Anti ShoW the WW condj 20 power by wives, as husband's status increases, may be partly due to fewer wives working in higher status levels. In general, Blood and Wolfe found relative marital power to be directly associated with the following “resource“ variables: rela- 'Hve work participation (husband—-overtime, full-time, or not employed; wife--employed or not employed), relative education of spouses, number of years the wife had worked during the marriage, and husband's income. Husband's occupational status, husband's education, and relative parti— cipation of spouses in organizations were also directly related to marital power, as predicted, but to a lesser degree. With minor modifications,16 Blood and Wolfe's study has been replicated in several other countries. The studies of Michel (l967) in France, of Safilios-Rothschild (l967) in Greece, and of Burié and Zegevic (l967) in Yugoslavia revealed that the relationship between certain resources and marital decision—power varied among cultures (see Figure 4 for a summary of differences). Rodman (l967) attempted to broaden the theory to account for these differences. His work, in tumu was subjected to empirical testing by Burr et al. (l973). 16In particular, the specific decisions used to measure marital power in American couples were not always appropriate for other cultures. Michel, for example, used five of the original eight decision—items. She then added other decisions typical of French families, selecting those which seemed to provide the nearest cultural equivalents to the items used in the U.S. study. (The cultural grounds on which decision items were selected are discussed in Blood et al., l970). For this reason, it is not valid to compare the absolute value of power—scores between countries. The comparison focuses, rather, upon whether resources and power show the same patterns of relationship to each other under different cultural conditions. — As the rem Wm has apparer ianenot fully succ purer patterns amor oriusion about the he United States , (see Burr et al., l oiled for; only ti oi the nature and | lithe] (i967) Michel i n‘ iench families li generally consiste iehueen resources significant. As i epioyedwomen. an tountof. fonmal e hesame general 1 simple, low white empower as higl niuided into thre in mean power bet education of spou hair the husband} 21 As the remainder of this chapter will show, this research -program has apparently reached an impassé. Revisions of the theory have not fully succeeded in accounting for differences in marital power patterns among cultures. In my opinion, the present state of confusion about the relation of resources to marital power-~even in Um United States, where the theory has received the most support (see Burr et al., l973)--indicates that a new strategy of inquiry is caHed for; only then can we perhaps solve the complicated question of the nature and bases of marital power. Michel (l967) Michel interviewed 550 wives in a random sample of urban - French families living in Paris and Bordeaux. Her findings were generally consistent with those of Blood and Wolfe, but relationships between resources and power were weaker and often not statistically significant. As in Detroit, marital power of wives was higher for employed women, and husbands' power was directly related to their amount of formal education and income. Occupational status exhibited the'same general trend, although with some inconsistencies. For‘ example, low white-collar workers had the same, rather than greater, mean power as high blue-collar workers, and white-collar workers-~when divided into three levels instead of only two—~exhibited some decrease in mean power between the middle and the highest levels. Comparative education of spouses (whether the wife has more, equal, or less education than the husband) corresponded to the U.S. pattern in its relation with — 111th power. between income to advantage was di mooted. Howey artly until the out different i ones. Michel u did of income, lands that woulc iafilios-Rothsct Quite a mi decision-poo tployed wives : only in the trai education was in earital power. heir power sco ranges and high status did not least powerful rather the inde ind medium-l eve When i husbands' mean husband or the 22 marital power. Finally, Michel dealt with marital power and difference between income of spouses (which Blood and Wolfe had not done). Income advantage was directly related to possession of marital power, as expected. However, power of employed wives did not increase signifi- cantly until they had higher incomes than their husbands; power was not much different for wives with lower incomes and wives with equal in- comes. Michel notes, however, that wives are rarely able to obtain the find of income, education, and occupational superiority over their hus- bands that would yield them high marital power. Safilios-Rothschild (l967) Quite different relationship patterns between certain resources and decision-power were found in a T964 sample of T33 Athenian wives. Employed wives still had higher mean power than unemployed ones, but mfly in the traditionally ”feminine“ household decisions. Husbands' education was inversely, rather than directly, related to husbands' _marital power. Income of husbands showed a curvilinear relationship to 'their power scores; husbands' power was lowest in the middle income L ranges and high at both extremes of income. Husbands“ occupational status did not show a clear trend; however, unlike French families, the least powerful husbands were not the semi- and unskilled workers but rather the independent professionals and the group of lesser professionals and medium-level managers and administrators. When education of both spouses was considered simultaneously, husbands' mean power showed no differences between couples where the husband or the wife had an education-advantage. Equal-education couples — are the most husbc in mount of schor the power if they higher education It liberal ideas." : iviuuerican standan men than less-edr traditional feminin In conclu: I‘reory was at leas inekmen; wives w advantage or who w guai education or high resources, as both income and ed usbands with low “— . l7These l loosening of old 1 larger effect on i 13It is a andhuerican ones cTile. Americann children. For Fr the children were tilerhand, appea Iinst child and s hthschild sugges husband in Greece heconrnonly "ave neasibie option 1 hilt-related den liver in other an Course, the data ordinal; interprr notion.) 23 were the most husband—dominant of all, which was attributed to the very low amount of schooling of these couples. That is, husbands will exer- cise power if they are less-educated and adhere to traditional norms; at higher education levels, husbands were thought to have more exposure to 17 liberal ideas. Such “liberal” husbands may still be quite powerful by American standards, however; we only know that they have Jess decision- power than less—educated men in their own society--and then only in the traditional feminine domains. In conclusion, Safilios—Rothschild noted that the resource atheory was at least partially more applicable to Greek women than to Greek men; wives with high resources, i.e., those with an educational advantage or who were employed, “earned“ more power than wives with equal education or who did not work outside the home. Husbands who had high resources, as measured by a composite social status index, including both income and education, had less power (relative to their wives) than husbands with low resources.18 M 17These liberal ideas were not portrayed as new norms, but as a loosening of old norms which allowed resources and free cho1ce to have a larger effect on decision—making. 18It is also interesting that Greek families differed from French and American ones in patterns of power over the stages of the fam1ly l1fe Cycle. American wives had the lowest power when there were preschool dfildren. For French wives, the low point occurred somewhat later when Um children were in school but not yet adolescents. Greek w1ves, on the other hand, appear to gain rather than lose power with the b1rth of the first child and steadily increase their power throughout life. Saf1lios— Rothschild suggested that motherhood does not increase dependence upon the mmband in Greece, as it does in France and the U.S., because grandmothers are commonly ”available” for child care. This also makes employment a more feasible option for Greek mothers. Greek mothers ga1ned power only 1n dfild-related decisions; however, it is noteworthy that they d1d not lose Power in other areas, as compared to newlywed couples w1thout ch1ldren. (Of course, the data in all these studiesare cross—sectional rather than longi- tUdlnal; interpretations about ”gaining” power over t1me must be Viewed w1th caution.) iii and Ze‘c/evi The Yu ihilies, rando whim city. i'w to the ire m power than Serbian city, e pout three dec iw expensive w hheting, fami‘ little affected his inversely r hthschild had iith lower educ idly ideology ehcation also highest power w hfference betw i 1.5. and France to her power. ~--—-——_.__...\ 19Hust although power countries. Th‘ ‘heearly Detr 24 0/ }/ o/ Buric and ZeceVic (l967) The Yugoslavian study of marital power was done with ll7 fleies, randomly chosen from workers in a large car factory in a Serbian city. Findings were generally similar to the Greek, rather than to the French or American, studies. Employed wives had higher nman power than unemployed ones, as in all the countries. In this Serbian city, employed wives had more ”say" than unemployed wives about three decisions: purchasing expensive household items, purchas- ing expensive women's clothing, and installment loan decisions. Family budgeting, family planning, and decisions about leisure time were ' i little affected by wife's employment. As in Greece, husbands' authority wasinversely related to amount of formal education. As Safilios- Rothschild had suspected in the Athenian study, Yugoslavian husbands19 with lower education were found to subscribe more often to a traditional fmmly ideology than were better educated husbands. Wives with little education also were more traditional in this respect. Wives had the thighest power when both spouses were well educated. In addition, difference between spouses' education also affected power. As in the U.S. and France, the wife's education advantage was directly related to her power. “—— 19Husbands were also interviewed in Yugoslavia and Greece, although power scores reported are based on wives' responses in all countries. This was done to make the later studies comparable to the early Detroit research which had only female respondents. Unlike waslavian hus whitional st tiess pronounc lilis pwer; l we little dif lands. Part of tilled workers areas. In Gree as directly re‘ ilesser extent leery of Eros The mo that resource- wntries, and rental power 5 notified for us Eating power-me nersocres be resource had tl inaction of SI for tables l 1 25 Unlike the highly industrialized France and United States, Yugoslavian husbands “lose“ power (relative to their wives) as their occupational status rises. This trend also occurred in Greece but to aless pronounced degree. Husband's income was not directly related tolfis power; low-income husbands had the highest power, while there were little differences between power of medium and high-income hus— bands. Part of the traditionalism of low—salaried unskilled and skilled workers was attributed to their recent migration from rural areas. In Greece, but more strikingly in Yugoslavia, wives' power was directly related to length of their own urban experience and—-to alesser extent—~to their husband's urbanization. Summary of Cross-Cultural Findings ‘ The most succinct way to summarize findings is to select those resource-variables for which data was collected in all four countries, and examine the relationship of these resources to mean marital power scores. Since the original decision—index had to be modified for use in other cultures, and since techniques for calcu— lating power-means_differed, it is not meaningful to compare absolute power socres between countries. One can, however, see whether a given resource had the same pattern of relationship with marital power. The direction of such relationships is illustrated in Figure 4 (adapted fiom tables 1 through 5 in Burié and Zetevié, l967). than (I967 Rodi hgeneral, l hshnds' ed innoe. Inn hint to the “term tion, income inapiwers stricture“ ( lislileliho aritai nom larly with e was are ge is none oppc into playa with the em that even B tional stat hailies. tlich may 1 Cf'resourc land that den they lwerworki lines) the 26 Rodman (1967) Rodman summarized and interpreted the four preceding studies. higeneral, he argued, husbands' power scores are directly related to husbands' education, occupational status, and income in the U.S. and France. In Greece and Yugoslavia, husbands' power is inversely re- lated to these “resource“ variables. To account for this, he proposed a"heory of resources in cultural contex .“ He reasoned that educa— tion, income, and occupational status are "not merely resource variables in a power struggle, but are also positional variables in the social structure“ (Rodman, l967: 32l). In relatively traditional societies, thelikelihood of learning sentiments favorable toward equalitarian marital norms may increase with one's position in the society (particu— larly with educational level). In less traditional societies, where norms are generally equalitarian and more flexible to begin with, there is more opportunity for Situational variables such as resources to come into play and influence behavior. How does this revised theory compare with the empirical data available? Safilios-Rothschild (l970: 547) notes that even Blood and Wolfe's (1960) original data revealed that occupa- tional status and income are not consistently related to power in U.S. melies. Low blue-collar men had more say than high blue-collar men, which may indicate that traditional norms still restrict the operation of'Hesources“ in the lowest social classes. Similarly, Komarovsky (l967) found that-~among working class families--husbands had more decision—power when they were lower in education, income, and occupational status, i.e., lower working class husbands were more powerful (in relation to their wives) than were higher working class husbands. ’4 l: _ v _ o\ _ . _ I. \ LMEOQ Sm WIAIIIIIZOJ LDrIAIIIIIZOJ mme ..mJUm. GLOZ SmeAlITIISOJ UGXOHQEN UQNOFQEN Emmi 0$w3 Wsz “OZ .WUCMOWJI 0rr—00CH WJQGFW COPHQUUUN— COTUQUIUU {DrquerULml CF Midunvivaxrlunr.‘ -.AA W.~UEWQW3T~ v.13} 1.3Km3 M.UCfiDmJI POCOrUGEJUUO h.i£53tl: .mmwgpczoo Lack Cw mficmnmzc ....o LwSOQ FmprmE cme vcm ...mwoLzome: Uwpumpmm :wwzwwn awcmcowum_mm moz<¢m saw: 304 mommmw saw: szom cam: .mucwnmzz inninnnininnnininnnnln .. mFIAiI-ai304 umAOFQEm uonFQEm iaiuzo mvzkinn 304 mmw4 szcm mgoz ; . gmwzmi 4 ; mew: mew; poz all In mHqucma m :owpmozvm cowpmozum cowum . . mwmmHWf: chwflwmwmooo cw mmocmxwwewo m.v:mnm:: xcoz m_www3 _ m.vcmnm:: fi - .¢ stme <4mowz> Puzzli ”Japan, incon hile occuPatl‘ industrialized and one would I equalitarian nu use. In Belg support (Lepl e urian ideolog association of hnnarl and Sw lure, Ahern, a Rodme pi hypotheses college studei income of spou iltained: \~_ 20 . a lusnon may be onllllives' p lilenore deg reports, howe penile. Th melltl'ons a 28 Puzzling results, however, have been found by other researchers. In Japan, income and education were directly related to marital power, while occupational status was inversely related (Blood, 1967). Although industrialized, Japan is still relatively traditional in many respects and one would have expected higher education to be associated with more equalitarian norms and hence with lower husband—power as in the Greek case. In Belgium, on the other hand, the resource theory received support (Leplee, 1968) while it did not in countries with more equali— tarian ideologies; Kandel and Lesser (1972) and Trost (1969) report no association of marital power with education or occupational status in Denmark and Sweden, respectively.20 Burr, Ahern, and Knowles (1973) Rodman's post-facto explanations were formalized into a series of hypotheses by Burr et al., and tested using 1,956 Brigham Young college students as informants about their parents.21 Using relative income of spouses as the resource indicator, the following results were obtained: 20Safilios-Rothschild (1970: 548) suggests that part of the con— fusion may be due to methodological problems, in particular the use of only wives' perceptions of power. Greek working wives think that they have more decision-making say than do housewives; according to husbands' reports, however, marital power did not differ by employment status of the wife. Thus it seems that the resource theory applies only to women's perceptions and not to men's. ’ 2'Results of these tests must be viewed with caution, because of the use of students as informants and because strongly equalitarian norms may not have been present in this sample of Mormons (90 per cent were members of the Church of Jesus Christ of Latter—Day Saints). Also, only four of the decisions included in the decision—power index corres- ponded with those in Blood and Wolfe's original index. Data analysis techniques also differed from earlier studies. Hypothesis 1 Hypothesis 1 Hypothesis Hliothesis HYPOthesis in short, norms 29 Hypothesis 1. The greater the relative resources of a spouse, the greater is his/her relative marital power. Supported when correlation techniques were used but not when multiple regression was employed. When controlled for type of norms and flexibility of norms, the relationship of income to power was curvilinear, very low in magnitude, and changed from a positive to a negative relationship as norms changed from patriarchal to equalitarian. Hypothesis 2. The amount of power normatively prescribed for a role is directly related to the relative marital power of the incumbent of that role (e.g., the more patriarchal the normative beliefs, the higher the husband's marital power). Strongly supported. This relationship re— mained strong even when resources (relative income) were controlled. Hypothesis 3. Resources will be more strongly related to marital power when norms are equalitarian than when norms are patriarchal. Not supported. In fact, there was a slight trend in the opposite direction; resources influenced power somewhat more under patri- archal norms than under equalitarian ones. Hypothesis 4. The greater the flexibility of norms (range of tolerable behavior), the stronger the relation- ship between resources and marital power. Supported when t—tests were used but not when multiple regression techniques were applied. Norm flexibility had the predicted effect only when norms were in the patriarchal range. Hypothesis 5. The greater the flexibility of norms, the weaker the relationship between norms and marital power. Not supported. In short, norms had much more effect on power than did relative income resources in this Mormon population. Resources had the predicted effect only within opposite ef Bu findings: Ihe all with i that l patrie equal' it my be, their bel II of course, for exampl [rather th countries. occupation reasonably in Greece DECt 30 only within the range of patriarchal norms and, if anything, had the opposite effect under more equalitarian norms.22 Burr et a1. conclude with a new hypothesis, based upon their findings: The amount of power normatively prescribed for a role interacts with the relationship between resources and power in such a way that resources are positively related to power when norms are patriarchal and negatively related to power when norms are equalitarian (ibid.: 16). It may be, then, that Detroit families were much more patriarchal in their beliefs than Blood and Wolfe assumed. This hypothesis does not, of course, fit all of the findings from the studies I have mentioned. For example, it cannot explain why resources show no relation to power (rather than a negative relation) in equalitarian-ideology Scandinavian countries. It also cannot explain why income and education but not occupational status “act as resources” in Japan. It does, however, fit reasonably well with the data about husbands' resources and ideologies in Greece and Yugoslavia. M 22Burr et a1. (1973: 17) note the similarity of their results to those of Hoffman. Hoffman suggested that employed Wives who are ambivalent about marital ideology-—who had only partially reJected patriarchal norms--may actually become less dominant than before they WOFKEd, "in order to compensate for the threat offered by the sheer fact of her employment“ (Hoffman, 1960: 34). Wives who accepted male dominance, she reasoned, might gain in power when employed for three reasons: U) These women are more likely to work out of economic neceSSity and might, therefore, consider their husbands to be failures. (2) They may be more likely to feel that the economic role 15 the legitimate basis of power, since this is an idea present in male dominance ideolo . (3) Women who endorsgymale dominance ideology may have such an exaggerated ideal of masculinity that their husbands can't live up to these ex- pectations. The r appears to me of the confusi ' analyze thisb the shortcomin theory itself. 31 The research program on the resource theory of marital power appears to me to have reached something of an impasse. As a way out of the confusion of contradictory findings, it is time to critically analyze this body of research and to find new approaches to circumvent the shortcomings in both the methodology and in the formulation of the ‘theory itself. None mtg the Res is that they 5 sonething whit about the oils The ( associated wi‘ ratters (see I it decision-to one "right" decisions inv pier benefits he inportanc ill (necessar be old quest Mlc‘ if the endles iflower. Ag milldnce’ at CHAPTER III ANALYSIS OF THE RESOURCE-THEORY WORK Critique: Marital Power None of these previous studies, in my opinion, has really A primary shortcoming tested the Resource Theory of Marital Power. is that they simply did not study marital power at all but, rather, something which might be called resource-control authority (deciding about the allocation of family resources)“ The one consistent finding is that a spouse's income is associated with the making of marital decisions about financial nwtters (see Hoffman and Nye, 1974, for a summary of findings according It seems that earning money somehow gives one- to decision-topic). more ”right“ to have a say about money matters. Making income—related decisions involves control over family resources, implying influence Without in any way discounting over benefits of other family members. the importance of such resource-control, I would maintain that it is not (necessarily) identical with marital power. Of course, this raises the old question: What is power? My distinction between resource-control and power evokes images of the endless literature devoted to ”pinning down" the elusive concept Attempts to conceptually distinguish marital power from of power. dominance, authority, influence, and control have not proved very 32 fruitful (W alization of avoid the she The Criterion Ther of empirical conceptual 1e empirical lei Raven's (I96l coercive p0wi open). The‘ use in resea 1115) becausr several powe to identify liter-bagfi 33 fruitful (Turk, 1975). Yet another definition and/or operation— alization of power will serve no useful purpose unless it can avoid the shortcomings of previous efforts. The Criterion of Empirical Fit There are two aspects of what I call the ”criterion of empirical fit.“ The first is that distinctions at the conceptual level are not very useful unless they refer to phenomena which can be distipgpished from each other on the empirical level. A good case in point is that of French and Raven's (1960) typology of the bases of power (expert power, coercive power, reward power, legitimate power, and referent power). Their typology has been little used and of little use in research (however, see Ravens, Centers, and Rodrigues, 1975) because almost any "empirical“ instance of power involves several power bases simultaneously. It is almost impossible to identify or measure the relative contributions of each power—base. Th one conce t distinguish is diificul that consti 11 clearly fa‘ amng situr tonsider a and makes am) iithin suc following ii 3 ouse I) to of ti ti o 34 The second aspect of the criterion of empirical fit is that empirically distinct phenomena should not be "lumped together” under one conceptual rubric. That is, one's concepts should allow one to This criterion distinguish among significantly different phenomena. is difficult to put into practice, because it involves judgments about what constitutes such a “significant difference.“ The definition of marital power as “resource control“ quite clearly fails in the latter respect, that is, it fails to distinguish among situations having a wide variety of “causes” and “consequences” Consider a group of married couples, all of whom report that the hus- band makes more money than the wife (i.e., the husband has higher resources) and that he usually has more say about money-related decisions Within such a group of couples, one might find instances of all of the following situations: a) Spouses agree about who should make financial decisions 1) Couples subscribing to the norm that this is the proper role of a husband, that it is his legitimate authority to make these decisions. Such couples would probably not change their decision—making patterns even if the wife were to start a high-paying job (gain more resources). 2) Couples subscribing to the norm that the earning of money entitles one to the major say about the use of that money. If the wife begins earning money, she automatically would be given the major say about the allocation of her own earnings. 3) Couples subscribing to neither of the above norms, where the husband happens to take on more of the financial responsibilities If a difference of for reasons of preference or convenience. opinion would arise about the use of money, the husband's ideas would not necessarily prevail. 11W 4) Cor ber wii dei as: m c—) 9 lecall the any (hat (have gree resource p the Resouo fESOUi‘CG-r It titles 1, ltbld hay senses (t liteement be trite c Elllrical 35 b) Spouses disagree about who should make financial decisions 4) Couples in which the husband makes financial decisions only because his wife refuses to do so. He feels overburdened with other responsibilities and/or dislikes making these decisions but cannot convince his wife to give him some assistance. Couples who disagree about use of income, in which the At least the wife (and perhaps 5) husband's ideas revail. also the husband; does not subscribe to the norms of types (1) and (2) above. She feels that her contributions to the family, although not of a monetary nature, entitle her to a say about all kinds of family decisions (including financial ones). However, she does not succeed in convincing her hus- band to accept her opinions about how money should be used. (Note that in this case, the disagreement about who should make financial decisions is only a reflection of a disagree— ment about the outcome of the particular decisions in question, not a disagreement about basic marital role patterns). Recall that in all couples under consideration, the husbands earn more money (have greater resources) and also make more financial decisions (have greater resource-control). This apparent ”correlation“ between resource possession and resource control, often taken as support of the Resource Theory, fails to make important distinctions: I. The distinction between couples where the husband's resource-control would not change, even if his resources were lowered (types 1, 3, 4, and perhaps 5) and those in which a resource change would have an effect (type 2 and perhaps 5). II. The distinction between resource control based on con— sensus (types 1, 2, and 3) versus that occurring in situations of dis— agreement (types 4 and 5). The Criterion of Empirical Independence Distinction(1) above is complicated by the difficulty 0f empirically isolating (or “partialling out”) the effect of normative influences, i an independen confounding t be eliminated There are nor pork and woma ppeprp of f rents exist). If n rarital roles hope of isole rarital power should, there the other ma, Ihi: but to plead Separated fr lites in thi possession w fllld betwee that is, hug llliituniti e “it cases i 'easure Dowe indicatop g, 36 influences, in order to identify the actual strength of resources as an independent variable. There are two types of normative influences confounding the relationships observed, at least one of which could be eliminated by a better operational definition of marital power. There are norms about the content of marital roles (what is man's work and woman's work), and there are norms about the overall authority patterns of families (who should have the final say, where disagree- ments exist). If marital power continues to be measured by the content of marital roles (who makes what decisions, for example), there is no hope of isolating the relative effects of resources and norms upon marital power. An indicator known to be strongly normatively patterned should, therefore, not be used to measure marital power, when one of the other major variables in the theory is that of norms. This is not to deny the influence of norms upon marital power, but to plead for a procedure where the influence of norms could be separated from that of resources. The theoretically most interesting cases in this respect would be those in which norms and resource possession would predict different results. (In the past, relationships found between resources and marital power have been “over-determined”; that is, husbands were favored both by the norms and by the structural opportunities for obtaining resources. The ”deviant” cases provide good test cases for getting around this problem, but only if we cease to measure power in terms of the content of marital roles but can find some indicator less affected by norms.) The Wautho is difficult by this norm. tyneans of ( share similar ioneasure tl ieiining and thm As i resource conr Theory of Mac conceptual 11 possession; rariable of r such vastly consensus ve In cation of ma however, 1 w Aid for my Y big that of Bl 0( 37 The second type of normative influence, that specifying Séflfifiil authority patterns, is more difficult to handle. That is, it is difficult to find a measure of marital power which is not confounded by this norm. The best we can do, perhaps, is to control this variable by means of (a) sample selection, i.e., choosing a sample likely to share similar norms for general marital authority and/or (b) attempt to measure this norm directly, using it as a control variable. Defining and Measuring Marital Power: The Approach of the Present Study As has been discussed, the definition of marital power as resource control is inadequate for purposes of testing the Resource Theory of Marital Power. As a dependent variable, it: (a) lacks conceptual independence from the independent variable of resource possession; (b) lacks empirical independence from the other major variable of norms; and (c) fails in general to differentiate among such vastly different situations as resource-control in a context of consensus versus conflict. In the following discussion, I will present an operationali- zation of marital power which avoids most of these shortcomings. First, however, I will briefly discuss the conceptual definition of "power" used for my study. The Concept of Power My conceptual definition of power is somewhat different from' that of Blood and Wolfe. I define power to be present in a relationship men there occur in t eliminates and wants. the influe is exercis decision-n and ii the not low ir heating tl Inpi ri cal l and llolfe cote powe rho shoul responsib different 0' 0linir d‘Silree llhosing '0 it the 38 when there is (a) non—self-satisfying behavior23 which (b) would not occur in the absence of the other party to the relationship. This eliminates behavior which is aimed at attaining the individual's needs and wants, since such behavior is self-generated rather than due to the influence of another. This definition cannot handle power which is exercised via socialization; if wives are socialized to want little decision-making responsibility and much housekeeping responsibility, and if their marriage corresponds to these desired patterns, they are not low in power by my definition. Meeting the Criterion of Empirical Fit My treatment of power differs considerably from that of Blood and Wolfe. I do not consider the making of marital decisions to indi— cate power unless (a) spouses differ in their preferences (wants) about who should decide,24 and (b) the spouse who decides also ypppg that responsibility. Blood and Wolfe simply assumed that everyone wants to 23This term is borrowed from Martin (1971). He uses it in a different context, however. 241 do pg; require that spouses communicate their differences of opinion or that overt resistance be present. The failure to openly disagree does not necessarily mean that one doesn't hold a strong opposing view; it may simply be a reaction to great powerlessness. Also, I use preference measures obtained at an earlier time point than were the outcome—measures. Undoubtedly, preferences measured later would be more congruent with reports of the “actual” outcome. Such shifts in preferences over time may themselves be the result of influence and hence would camouflage the amount of power which was exercised. ' male decis fined as r distinct p rubric of heating tl ionceptual l me less tl power-ind lhaue fun and effor -—-— missed th insal to purer. l tions of ll dominant fiiltal 39 make decisions.25 By narrowing the range of behaviors which are de- fined as reflecting power, I have reduced the number of empirically distinct phenomena which are grouped together under the conceptual rubric of ”power”. Meeting the Criterion of Conceptual Independence Possession of resources and resource—control power seem to me less than totally independent concepts. Rather than limiting my power-indicators to decisions about allocation of financial resources, I have focused upon allocation of such non—monetary resources as time and effort. 25Their treatment of ”resource control“ as power altogether missed the kind of power described as type (4) on p. 35, namely re— fusal to take responsibility which can be a very potent sort of veto- power. This failure, I think, accounts for some serious misinterpreta- tions of findings. Take the example of Blood and Wolfe's (1960: 258) claim that "dominant wives” have low marital satisfaction. Yet they tell us else- where (p. 67) that these wives, although often employed full-time, do “far more household tasks than any other type of wife . . . This counters the common assumption that powerful spouses compel their partners to carry out their bidding, and is further evidence of the non—coercive nature of marital power." I would suggest that this provides evidence, not of the non—coercive nature of marital power, but of the lack of power of these wives. Can a woman who can't convince her husband to help a little really be ”powerful”? The husbands, rather than being weak and incompetent as Blood and Wolfe assume, appear to exercise a great deal of ”refusal power”. I suspect that these wives really belong to the “husband dominant“ category, a category also characterized by lowesatisfaction wives. Combining these two types of couples would create a group homogenous with respect to wives' low marital satisfaction, including wives with both low and very high ”resources”, having in common a low level of marital power. Thus it may be that very low marital power (rather than low resources or low resource—control) accounts for low marital satisfaction. mpiri cal of normat confidenc variable about who cultural decision- betveen v decision, who shot eliminate from my : tlti, rev ll quest compare vvestion 40 Meeting the Criterion of Empirical Independence The problem of finding power measures which are independent of normative prescriptions is critical; otherwise, one cannot have confidence that power variations correlated with some independent variable are not actually due to normative influences. Questions about who makes what decisions clearly are strongly patterned by cultural norms and hence are not satisfactory. I also include decision-making questions but, by measuring power as the discrepancy between who makes a decision and who the spouse prefers to make that decision, I reduce the "contamination” of results by norms. The influence of norms about general authority patterns (e.g., who should prevail if there is disagreement) has not been satisfactorily eliminated, although the elimination of working and lower class couples from my sample may have reduced variations in such norms. I did, how- ever, reduce response bias from this source by the indirect nature of my questions about power; few respondents would guess that I would Icompare their preference—answers in the first interview with outcome questions asked months later. The Indicator of Marital Power The major indicator for marital power involved a series of eight responsibilities. The specific items included were selected according to the following criteria: (a) that both ”male” and "female” responsibilities be included; (b) that the responsibilities have conse— quences for both spouses, i.e., that their performance affect the joint life of ti couples; have diff ln neetin to have b not claim couple, c who were breakdom selectec lllll. Iits a r iii assc lithougl wife, density ill help 41 life of the couple; (c) that they be common to almost all married couples; and (d) that they be areas where spouses would be likely to have different preferences about who should take on the responsibility. In meeting the last criterion, I utilized several items which proved 26 to have high “role discrepancy“ scores in a previous study. I can- not claim that these items reflect the total balance of power of a couple, of course. The responsibility items were as follows:27 ) Grocery shopping 2) Repairing things around the house (furniture, toys, appliances) ) Making complaints, if necessary, to salesmen or landlord Keeping in touch with relatives Deciding what people you will invite to the house ) ) 6) Taking care of the car ) Preparing supper ) Doing the supper dishes (or loading the dishwasher) 26This study was done by Rogers (1972) with Michigan couples who were somewhat older than my respondents. She provided me with a breakdown of items according to mean role-discrepancy scores, and I selected those having the highest scores (Rogers, personal communication, 1973). A role-discrepancy is the difference between how a spouse pre- fers a responsibility to be divided and how it actually is divided. One may assume that preferences of spouses differ for role—discrepant items, alphough Rogers did not actually compare preferences of both husband and wa e. ' 27Couples with children were asked about two additional re- sponsibilities: getting up at night, if necessary, with the children; and helping the children with their homework. Re5| about their 1 During ‘ have to bi 115. would p regardl way or llho wou shoppin Son hone, respor and asked tr answer cater Fir which they r the discrep. iii Time 2 husband's w discrepancy Sill relati ll Flgure 5 ithltving h 42 Respondents were asked in the earlier telephone interview about their preferences for these items: During the next few months, there will be many jobs which have to be done--everything from cooking to paying the bills. I'll mention a few of these jobs and ask how you would prefer to divide them up between the two of you—- regardless of whether they probably will be divided that way or not. Who would you prefer to (name of item, e.g., "do the grocery shopping”)in the next few months? Would you say: Husband almost always Husband more than wife Both about the same Wife more than husband Wife almost always Neither one Some months later, during a personal interview in the couple's home, respondents were presented with a list of the same responsibilities and asked to check “Who usually does this?", indicating one of the above answer categories for each item. First, I compared preferences of husband and wife. Items on which they agreed were eliminated; for the remaining items I calculated the discrepancy (absolute difference) between a spouse's Time 1 preference and Time 2 report of who does each responsibility. Then I compared the husbandfs with the wife's discrepancy scores; the spouse with the higher discrepaney of the two was considered to have the lesser relative power. Such relative power scores could vary from +4 to -4. In the illustration in Figure 5, the wife has the greater power because she came closer to achieving her wants than did her husband. life‘s Pt I Husban Almost llways (1) Figure 5. This and vidual 5; average strongly reason, ing to t lieferen to achie i asked ligardl itcisior ”litter 43 Wife's Preference Husband's Preference \L Actual Division ‘L Husband Both Wife Almost Husband About Wife Almost Always More Same More Always (l) (2) (3) (4) (5) Her discrepancy score = 11 - 2| = 1 His discrepancy = l - 2] = 3 Her relative power = 3 - l = +2 His relative power = 1 — 3 = -2 Figure 5. Illustration of power-score calculation for a single responsibility-item. For some hypothesis tests, a power-index was constructed. This index was a simple average of all discrepancy scores of an indi- vidual spouse. Relative power was then determined by comparing the average discrepancies of husband and wife. The use of unweighted power-indices has been criticized strongly in the past (see Safilios-Rothschild, 1970). For this reason, I also constructed an index in which items were weighted accord- ing to their perceived importance. I am assuming that achieving one's preferences in important areas reflects more power than the ability to achieve preferences in unimportant areas. In the second interview, I asked respondents to indicate the importance of each responsibility: "Regardless of who does it, do you think this is an important job or decision". Relative power scores for each responsibility were then weighted according to importance, as illustrated in Figure 6. FORMULA: Figure 6 Such we weighter spouse': inporta index 3 dealing issues. 44 FORMULA: Importance factor times relative power score = Weighted relative power score. And the relative Then the weighted power score was relative power If importance was: -2: score was: Very Important .20 x -2 = -.40 Fairly important .17 x —2 = -.34 Slightly important .13 x —2 = —.26 Not at all important .10 x -2 = —.20 Figure 6. Calculation of weighted relative power scores for individual responsibility items. ’ Such weighted relative power scores could vary from +.8O to -.80. A weighted power index was then constructed by averaging an individual spouse's weighted scores for responsibility items. Although the importance-weight coefficients were chosen arbitrarily, this weighted index still represents an improvement over an unweighted one. Concluding Comments In summary, I should like to identify several key issues dealing with the nature of "power“ and indicate where I stand on these issues. Individual versus relational power: The Weberian concept of power as the ability to achieve one's will despite resistance implies that power is an attribute of the individual. Emerson (1962) and, to some extent, Thibaut and Kelley (1960) view it as a relational concept. That is, an individual has power in the context of a particular relation- ship. He or she may have the power advantage in one relationship but not in another. B study of n notion of t is an exar agrees th.‘ sensus. l authority (the llebe I have so interest. decisions had been hill" on M» ti measure 1 45 Relationa1_power seems to me to be the proper focus for a study of marriage. This does not mean, however, that the Weberian notion of resistance need be ignored. Contested versus uncontested measures of;power: Authority is an example of uncontested power. Because the less powerful party agrees that the other party should have power, power rests upon con- sensus. Blood and Wolfe's decision-index comes close to measuring authority in this sense. Power to achieve one's will despite resistance (the Weberian definition) is power within a contested situation. As I have said before, this is the aspect of power in which I have more interest. My choice of ”contested power” meant that the outcomes of decisions can be said to indicate power only when prior disagreement had been present. That is, one must somehow measure what the individual's “will" was-~and ascertain that resistance was present to some degree (e.g., that the other party "willed” something different)--in order to measure power as the achievement of one's will despite resistance. Zero-sum versus non-zero-sum power: Any measure of contested power in a relational context must come to grips with the question of whether that power is to be viewed as zero-sum in nature. A question which asks “who wins more often, you or your spouse” treats power pg 1: it were zero-sum; the higher his power, the lower hers must be. The unweighted version of my power indicators--discrepancy be— tween preferred and attained division of responsibility--also implies a zero—sum assumption. One way to circumvent this questionable assumption is to as husband use of “ motion 0 levels. ferences preferer preferer tween h' 11105511 her Pre l y—— 5”those The per me 31 his in 46 is to use some weighted power-index, where the weighting factors of husband and wife need not be the same. I have already mentioned the use of “perceived importance“ to create weighted power scores. Another very similar way to get beyond a strict zero—sum notion of power is to take into account ”motivational investment“ levels. Take the basically zero—sum measure of “who wins” when pre- ferences differ. There is a five—point scale, upon which his and her preferences are located. If the final solution corresponds with his preference, his power is higher. Operationally, the discrepancy be- tween his preference and the outcome is much lower than hers, and her 'Woss" is directly proportional to his “gain”: Her Preference His Preference L 2 3 4 54 OUTCOME same as his preference his discrepancy her discrepancy II ll #0 Suppose that we now introduce the notion of strength of preference. The person who feels more strongly about his/her preference would have more subjective ”gain” from winning, while the person who cares little has little to gain or lose. Suppose for the above example that her motivational invest— ment (preference strength) is very high, while he is almost indifferent; he has very little at stake whether he wins or loses, while she would have more subjective “gain” from winning and more subjective ”loss” from losing. He prevails in this example. She loses a great deal, but he does not really gain. game, the pig some example c power as their in‘ iul indi thing to geggay it encom does not ilected ilhilar outcome. Choice A data 1 and mp ‘ 1 than i 47 Without the requirement that power be considered a zero-sum game, the implicit assumption that the powerful person is trying to ggip something for him/herself is no longer necessary. The particular example discussed comes closer to Habermas' (1971: 254) definition of power as the ability to hinder other individuals or groups to pursue their interests (as contrasted to Weber's definition where the power- ful individual is trying to attain his own will, i.e., to gain some— thing for himself). Under a non—zero—sum assumption, then, it is not necessary to assume that power is motivated by hope of personal gain; it encompasses both situations where this occurs and those where this does not. Eggptification of the critical elgmgpp of power: As Turk (1975) pointed out, oneds choice of a power measure reflects an assumption about where power “really” is located, i.e., the stage of’ choice-making where the exercise of influence is most consequential. I will be using power—measures dealing with influence as re- flected in outcomes of ”decisions.” In this respect my measures are similar to Blood and Wolfe's. I have no basis for assuming such outcome-power to be more consequential than interaction-power.28 My choice to select this measure of power was made solely on the basis of data quality. 28One could compare interaction and outcome measures of power and empirically determine which is the more crucial dimension, e. g. , by examining which can better predict satisfaction with outcomes. Tl ing improve The treatmr problem is the basic I A Theory had then const not think W l ship betwe [however i of the res that is, i cannot ea: effects 0 idlncide example, tion and reSUUi‘ces ion), a l ———ai v 48 The measurement of marital power is not the only area requir- ing improvement before the resource theory can be adequately tested. The treatment of ”resources” is also open to question. Here the problem is more than one of definition and measurement; it involves the basic overall strategy of research to be employed. Critique: Resources Assuming for a moment that previous studies of the Resource Theory had adequately defined and measured marital power, would they then constitute adequate tests of that theory? For two reasons, I do not think so. “Overdetermined” Relationships Within a given culture or subculture, any statistical relation— ship between possession of particular resources and marital power (however the latter is defined and measured) constitutes a poor test of the resource theory because such relationships are ”overdetermined.” That is, many factors are operating in the same direction, and one cannot easily isolate the effect of resource possession from the effects of other variables. It is not just that patterns of resource possession tend to coincide with normative patterns, although this is also true (for example, husbands are expected to--and usually do--have higher educaf tion and incomes than their wives). Structural opportunities to obtain resources are unequally distributed by gender as well (see Gillespie, 1971). Gender-role socialization patterns further strengthen the male advantag to striv testing couples Then one to part and out still s vhich r culture finding except- its ah tiered tional norm), liver income 49 advantage in resource possession by putting less pressure on women to strive for the resources which gyg potentially available to them. All of this would present a less serious problem for theory testing if study samples included relatively sizeable numbers of couples in which the wife has a resource-advantage over the husband. Then one could begin to isolate the relative contribution of resources to marital power; e.g., when patterns of relative resource—possession and cultural norms "predict” different results, is resource—possession still strongly associated with marital power? However, finding sufficient numbers of such “deviant” couples for study has been (and still is) extremely difficult. Multiple Resources and Testability The researchers in Chapter II have focused on identifying ypipp resources are associated with marital power in particular cultures. Although this approach has resulted in some interesting findings, the theory itself has never been put to a rigorous test. This point deserves some elaboration. With the possible exception of a well-designed laboratory experiment, multiple resources are always present in any social situation. Any negative findings, therefore, can always be explained away as being due to some addi— tional unmeasured resource (or, for that matter, to some assumed norm). Take Blood and Wolfe's (1960: 27, 42) finding that average power of husbands declines with age, despite the fact that husbands' incomes and occupational status tend to increase as they get older. Froponen vhich d6 cancel 0 is almos theory i one can marital situatii the the even be lesourc rrgorou 50 Proponents of the resource theory can always find some other resource which decreases with age,29 whose influence can be claimed to partially cancel out the effect of income and status resources. In short, it is almost impossible to find conclusive negative evidence against the theory itself by correlating resources with marital power (although one can establish that a particular resource is not associated with marital power under particular circumstances).30 One criterion of an adequate theory is that it should be falsifiable. That is, it should be possible to identify potential situations which would constitute conclusive negative evidence against the theory. If no possibility of conclusive negative evidence can even be envisioned, then the theory is not really testable. l; the Resource Theory hopelessly unfalsifiable and hence not amenable to rigorous testing? Certainly the approach of examining relationships 29Or, of course, they can point to some resource of the wives which tends to increase with age. 30This does not constitute negative evidence against the theory, since the theory does not pretend to predict which resources will lead to marital power under particular situations. Even exist— ing explanations about why a certain resource "works” in some cultures and not in others have been largely of a post—hoc nature (e.g., Rodman, 1967). Any attempt to predict anything more than very gross differences between cultures would run head—on into the problem of multiple resources. An additional theory would be required which could predict power outcomes from particular combinations of resources of one spouse in relation to those of the other spouse. Judging by the difficulties posed when Only two ”resources” are studied in labora— tory experiments, as in Berger and Fisek's study of ”specific status characteristics” (1970), the simultaneous assessment of dozens of resources would seem to be a hopeless task. hetveen res stringent t can test th question of To do this and operati gem 1 result in r 51 between resources and marital power does not provide this kind of stringent test. But perhaps another strategy can be used, one which can test the theory itself more or less independently from the question of which resources are important in a particular situation. To do this, I believe, the best strategy is to more directly identify and operationalize the elements in the hypothesized translation mechanism (i.e., the process by which pay resource is thought to result in marital power). This is the task of the following chapters. an "inpr refers interve into po are cen interac output, discus iPPl‘Od button liver lltrir theor till 11 Strand CHAPTER IV STRATEGIES 0F INQUIRY A resource theory of power contains three basic elements, an "input," a ”throughput," and an ”output.” The input, of course, refers to resources. The throughput phase refers to a postulated intervening process or mechanism by which resources are translated into power, a process in which mental' processes such as perception 1.3] The distribution of power (whether measured during are centra interaction or in terms of outcomes of contested decisions) is the output, the last element in the series. Depending upon one's research aim, i.e., upon what one wishes to describe or explain, any of the three elements may be taken as a ”starting point” in the strategy of inquiry. I will briefly discuss several approaches which might be taken, concluding with the approach chosen for the present study. Output as the Starting Point Here the objective is to describe or explain how given distri- butions of power came to be. An appropriate strategy might be to hold power "constant,” e.g., to select for study a group of couples having _—-~—-———_—._—_.—— 3'An exception would be a theory meant to deal with cases of outright coercion, including the use of physical force. Most resource theories of marital power have not attempted to deal with coercion in any explicit way. The work on violence in families (e.g., Steinmetz and Straus, 1975) has not generally employed a resource formulation. 52 similar p01 couples who they manage input-thror and the pro explanatio of non-equ to the equ crucial pr I point or t c come to mi distribut" tend to h lillespie account fr If the st Cvlar fes iiiects g hiatus C under why“ indium similar power structures. Say that one is interested in those rare couples who are extremely equalitarian and one wants to discern how they manage to achieve equalitarianism. Working backwards in the input—throughput—output chain, one examines their resource situations and the processes by which they arrived at a power division. If explanation rather than description is the main goal, a control group of non-equalitarian couples who are othenwise demographically similar to the equalitarian couples would facilitate the identification of crucial prerequisites for equalitarianism. In all of these examples, "output” is taken as the starting point or thing to be accounted for. Resources as the Starting Point Several approaches which begin inquiry with the input stage come to mind. One may want to account for the pattern of resource distribution, e.g., why husbands or wives in a particular society tend to have the advantage with respect to certain resources (e;g., - Gillespie, 1971). One may ask ygyppl resources can best predict or account for power distributions; this is the major objective of many of the studies reviewed in Chapter II. Or one may begin with a parti- cular resource and explore the conditions under which its distribution affects the distribution of power. The work on multiple resources (status characteristics and on relevance of resources [the conditions under which they are "activated'I and affect powerl) falls in this tradition (see, e.g., Berger and Fisek, 1970; Freese, 1969; Kruse, 1967). Re norms and i could atter tions accov post-hoc e; identifica' S stratified want to co oi resourc: higher the involves t core extre resources) l Point of c | )0 be rel. llhequal lvint“) m, It resouv the role 54 Researchers interested in the interrelation between cultural norms and resources would also begin inquiry at the input stage. One could attempt to predict power-outcomes of particular resource situa- tions according to type of culture (in contrast to Rodman's first post—hoc explanations in his 1967 article). This would require prior identification of theoretically significant cultural parameters. Studies focusing upon resource—input might utilize a sample stratified according to resource levels. For example, one might want to compare couples where both spouses are high on a given set of resources with couples where both are low, where the husband is higher than the wife, and vice—versa. A variation of this strategy involves the elimination of ”intermediate“ situations in favor of more extreme differences in resource levels (very high or very low resources). All of the above are strategies which take “input“ as their point of departure. Throughput as the Starting Point Here the concern is with the process by which resources come to be related to marital power. One might compare equalitarian and non—equalitarian couples as suggested above (“output as the starting point“) with the additional specification that both groups be similar in resources. Controlling in this way for resources, one focuses upon the role of process in accounting for differences in power-outcomes. according role of 1 those m from the cognitivl situatio be “acti require one can spouses strateg) tive thr err'dencr tire thr T__7 55 A similar strategy could be used with a sample stratified according to resources (both spouses high, both low, etc.). Here the v i role of throughput could be studied by examining the “deviant” cases, those which do not exhibit the power outcomes which would be expected from their resource situation. One could search for differences in cognitive processes or for conditions lacking in these couples' situations, the presence of which may be necessary for resources to be “activated." All of the above strategies, while potentially useful, require sampling criteria which are difficult to meet. For example, one can measure perceptions about relative overall contributions of spouses without having to identify whj£h_resources are involved. My strategy to increase testability also involves the comparing of alterna- tive theoretical processes. The possibility of finding negative evidence is provided by specifying test situations where the alterna- ‘ tive theories would predict different results. i The more direct operationalization of the theoretical process itself serves to minimize the sampling criteria problem discussed above. Unless one can avoid the need to deal with specific resources, it is necessary to find couples who are similar with respect to a large number of demographic and resource-related characteristics.32 Such 32My sample is relatively homogeneous with respect to general social class level, length of marriage, place of residence, age, country of birth, and race. It contains four subgroups according to previous marital experience (both spouses in their first marriage, both in second marriages, husband in first marriage/wife in second, and wife in first marriage/husband in second). But it was not possible to control for specific resources in a precise way or for such extraneous variables as re igion. control prooticz option I M an expe than-id best ac this or oost to that I resouro limitai (self-i W0 pl) will d 56 controlling for extraneous variables by sample selection is a practical impossibility for a study of limited finances. The other option of obtaining an extremely large sample, large enough to allow simultaneous statistical controls for a number of variables, is also an expensive procedure. My strategy does not entirely overcome the limitations of less- than-ideal sampling criteria. If I find that one theoretical process .best accounts for power distributions in my data, I run the risk that this process may only operate under certain resource conditions (those most common in my upper-middle class Michigan couples). To the extent that I do not deal with specific resources, I cannot determine the resource-related scope conditions of a theoretical process. A second limitation is that I necessarily must rely on indirect information (self—reports) rather than upon more ”objective“ observational data. With this introduction, I will now proceed to introduce the two process—theories which serve as the focal point of my study. I will discuss each in turn and specify the major hypotheses. Then I will identify the test situations I have devised for the assessment of their relative explanatory power, including the statement of alternative hypotheses. relevani or_o_c_es_s differei :arital assumes less-po origina equal-i CHAPTER V "EQUAL-INPUT“ THEORIES 0F MARITAL POWER Introduction: Equal Input vs. Power-Dependence Theories Two exchange-theoretical formulations seem appropriate and relevant for the present study. Both deal with the "throughput" ppppg§§_by which resources are translated into marital power. A major difference between the two is their assumption about the nature of marital power.‘ The first type, which I call "equal-input" theory, assumes that marital power rests upon the willing compliance of the less—powerful spouse. This view is consistent with Blood and Wolfe's original resource theory. I have identified two specific versions of equal-input theories. In the first, competence (derived from experiences associated with attaining such resources as income and education) is exchanged for compliance. When one spouse is perceived as having greater com- petence, the other spouse yields willingly to the other's greater ability. This is similar to French and Raven's (l960) ”expert power.” In the second version of the equal-input theory, resources are directly exchanged for compliance. If one spouse contributes more resources (especially those from which the other acquires benefits such as prestige in the community or a high standard of living), the other spouse has 'Wespect and gratitude.“ 57 {or defe is as if equal th he or sh auailabl even mil other 5; butions. w free of of the a resourca oar dif A house resourc rum 1‘ 50h)? so easily 58 In both the competence and gratitude versions, compliance (or deference) seems to become an exchange commodity in itself. It is as if there were some norm of equal exchange; when a spouse cannot equal the contributions of resources or competence of the other spouse, he or she feels obliged to repay the debt incurred with the only ”coin” available: deference. Marital power is thus not coercive; it is not even mildly repressive, in that the less powerful partner views the other spouse's power as the legitimate return earned for greater contri- butions. The second theory relevant for the present study is that of pgwer-dependence. Here marital power may be, but is not necessarily, free of all repressive qualities. Power is based upon the dependence of the other spouse, rather than directly upon contributions of resources. Two couples with identical patterns of resource-contributions nmy differ in their marital power structures, if they differ in dependence. A housewife with a good education is less dependent upon the husband's resources than is one without such qualifications, even though neither woman is employed.' The first woman gpplg_obtain financial resources from some source other than her husband. The second woman could not as easily because of her greater dependence upon the husband for resources; she is not as much in a position to insist on her own way if he feels strongly about something. She simply has more to lose, and may comply out of a conscious or unconscious fear of losing her husband. Having introduced the theories with which I will deal, I will now proceed to discuss each in some detail and to identify central Mahms equal-inpu mmeI l Muiss mwu,h Mower thmc anmdu hu"UE muons Momma Memm uumd a mega mmmm \- whua mama COhiribut that high omen 33$ 9 ma partitipa hill] to dc 59 hypotheses to be tested. The present chapter examines two types of equal—input theories, while power-dependence theory is discussed in Chapter VI. Perceived Competence The competence version of the Resource Theory of Marital Power is suggested by Blood and Wolfe. White-collar husbands, they suggest, have higher power than blue—collar ones because their job— related experiences make them "extra—equipped with the knowledge and skills required for decision—making, and their wives correspondingly are inclined to recognize their husband's competence along these lines“ (l960: 3l). The lower power of black husbands is attributed to job discrimination and poor schooling, i.e., to low educational and occupational status resources, which undermine their morale and make them “undependable” (ibid., 35). Dependability can be seen as an aspect of competence to handle marital responsibilities. Although Blood and Wolfe assume that high—power spouses have more actual competence,33 I believe that this is not a necessary assumption. All that should be required is that such spouses be 33They quite directly assume that possession of “resources" indicates that the possessor must be a resourceful person: ”This dis- cussion of the ways in which the husband's position in the community contributes to his power in the family may be concluded with a reminder that high-status husbands do not appear to be coercing their wives into submission. Such husbands have more to offer to their marriages because of their unusual competence and resourcefulness: it is this contributing to the marriage which brings with it greater freedom for the husband to participate in decision—making and greater willingness by the wife for him to do so" (italics added: Blood and Wolfe, l960: 36). V Lin—3.2.1. M. ii ,. ousted ‘ uell and kl ceived corn however, a tion forms I (a) Spouse they do no power is d rare-coupe views mari he or she l oasis of r reticians not VECOg‘ succeed i hOUid not hives ac Visionsib 1“ an eff her strat lend“ to mco"lihéte will re 60 perceived as having more competence. When two people know each other well and know each other's strengths and weaknesses, actual and per- ceived competence may often coincide. Sometimes they will not,34 however, and in these cases it would seem that the perceived situa— tion forms the basis upon which people act. Implicit in this formulation are the following assumptions: (a) Spouses agree in their evaluations of relative competence (if they do not, then the effect of perceived competence upon marital power is diminished accordingly); (b) The spouse perceived as the more-competent one likes and accepts that “label“; and (c) Any spouse i! views marital power as desirable, at least in those situations where he or she feels competent to exercise it well.35 This theory, in which resource-related competence forms the basis of marital power, coincides with some ideas developed by theo- reticians in the status characteristics tradition (e.g., Berger et al., 34For example, a competent wife may have a husband who does not recognize her competence, and a less-competent husband may still succeed in convincing his wife that he is the more competent. 35One can think of situations where assumptions (b) and (c) would not be met. Blood and Wolfe‘s concept of marital power often in- volves accepting responsibility. If a spouse does pp; want a particular responsibility, that person may deny or underrate his or her competence in an effort to justify why the other spouse should take on the burden; her strategy of feigning incompetence to avoid housework could be ex— tended to avoidance of decision—making responsibilities as well. Hence, incompetence can be a basis for power to avoid undesired decision (and other) responsibilities. expecta with ti terger‘ coupetr deveio plicat the he one-re 61 36 What Blood and Wolfe call "resources" (income, educa— l966, I972). tion, occupational status), Berger et al. term ”status characteristics.” Perceived “level of competence” becomes, for Berger, “performance expectations." Since the dependent variables in both formulations deal with the distribution of power, a remarkable similarity exists between Berger's work on Status Characteristics and Expectation States and the competence version of the Resource Theory of Marital Power. The work on expectation states, being at a higher stage of development and theoretical refinement, may therefore be useful in ex— plicating hypotheses for marital power. The close correspondence of l the two formulations is illustrated in Figure 7 below, for a simple one-resource situation. (i) (2) (3) BLOOD & WOLFE (Perception of) (Perception of) Relative Marital Resources Competence Power BERGER et al. (Perception of)37 Performance Power & Prestige Status Character— Expectations Order istics Figure 7. Comparison of major variables contained in theories of Blood and Wolfe (l960) and Berger et al. (l966, l972). 36Neither Berger et al., nor Blood and Wolfe, it should be noted, attempt to deal explicitly with the creation of actual competence. Specific competences could be viewed as varying along two dimensions: how difficult and time-consuming they are to learn and how desirable the possession of a competence is (for a particular individual or for men or women in general). During marriage, the probability that one learns a pew_competence may be directly related to its perceived desirability and inversely related to the difficulty of learning. Com— petences, after all, can be learned—-perhaps from a more experienced spouse. Thus differences in initial resource-related skills can be re- duced over time (without the requirement that learning experiences be linked to changes in external resources). 37In research situations, perception of status characteristics is often controlled or manipulated by the experimenter. In effect, this mani- pulation is a substitute for the evaluation of status characteristics as this would occur in normal interaction situations. 62 Once one goes into situations where multiple resources or status characteristics are involved, an additional step is required. This step would occur between stages (l) and (2) in Figure 7, and would involve an additional mental calculus on the part of partici- pants. Berger et al. (l966: 32 ff) explicitly identify the need for this intervening step. Each participant evaluates the several resources (status characteristics) of both self and other, arriving at overall competence evaluations or ”aggregated (general) expectation states” for both parties. In this calculation, resources or status characteristics presumably are weighted according to criteria such as their degree of relevance to the task at hand. The behavior of parti— cipants (e.g., whether or not they allow themselves to be influenced , by others) is a function of their evaluations;38 those that conclude Other to be higher than themselves in overall competence or aggregate expected performance ability will allow themselves to be influenced. Background of the Berger Program39 Some discussion of the history of Expectation State Theory may be useful at this point, because it may show to what extent substantively diverse theoretical ”strings“ share very similar ideas analytically. It emerged, of course, not from the study of marriage or even of ongoing . _ 38Other influences, e.g., intimidation, are not encompassed Within this model. 39 have drawn upon Freese (l969) for this historical introduction to the Berger Program of research and theory development. social and Sla laborat hihitec (obseri eualual rather interar collec‘ lifter eualua stabil group status seruec pusttt 63 social groups, but from the work of Bales and associates (e.g., Bales and Slater, l955) on ad—hoc groups. Bales found that his task—oriented laboratory groups of apparent status-equals, initially strangers, ex— hibited a cluster of correlated inequalities. Inequalities in behavior (observed frequencies of initiation and receipt of activity) and in evaluation (member—ratings of ”best ideas“ and ”productivity”) emerged rather early and tended to persist over time. These components of group interaction are conceptualized by Berger and associates as constituting, collectively, the ”observed power and prestige order” of the group. So-called equal—status groups did not remain equals for long. » Differential performance expectations were thought to arise from evaluations of early performance outputs. These expectations soon stabilize and proceed to determine the power and prestige order. Where group members are differentiated at the outset by some socially valued status characteristic (“resources“), positions of members in the ob- served power and prestige order tend to be ordered according to their positions with respect to the external status characteristic.40 The main generalization which can be made with respect to un— equal—status groups is the following: If task groups are differentiated with respect to some status characteristic external to the task situation, the observed power and prestige order of the group tends to coincide with the distribution of the status characteristic—-whether or not that characteristic is related to the group task. ~————___~ M 40This effect has been documented with respect to such status characteristics as occupation and sex (Strodtbeck, James and Hawkins, 1957), race (Katz and Benjamin, l960), and organizational position (Torrance, l954). in ad-ho program The 1966 with d1] as race and unit 9% "resour he diff 64 The line of research concerned with explaining this phenomenon in ad-hoc groups of initial status-unequals is that part of the Berger program which he labels “status characteristics and expectation states." The l966 Formulation Berger, Cohen, and Zelditch (l966) proposed a theory dealing with diffuse status characteristics, i.e., those characteristics such as race and gender which are surrounded by many beliefs and stereotypes and which are associated with expectations of superior or inferior general performance ability. (I believe that Blood and Wolfe's ”resources“ of occupational status, education, and income would also be diffuse status characteristics.) If one particular diffuse status characteristic were the gflly basis of discrimination between two actors in task—focused groups, and if actors were forced to decide who was right and wrong on an ambiguous Atask, then they would decide on the basis of who ranked high or low on the diffuse status characteristic. Lacking any other basis for their performance evaluations (and being unable to evaluate quality of per- formance directly due to the ambiguous nature of the task), actors will base evaluations on the diffuse status characteristic even if this has no direct relation to competence at the specific task. Empirical.support for the l966 formulation includes Moore's (l968) study of the diffuse status characteristic of prestige of school attended and Cohen, Berger, and Zelditch's forthcoming work on Air Force rank. Bi the manner behavior 0 pouer and Fisek (l9l (ll idliCh Bey Observed coutrollt aCtivitp, ill coula correspo, 65 Before relating this discussion to families, let me summarize the manner in which expectations are thought to affect the observable behavior of individuals--that is, the four components of the observed power and prestige order--based on my interpretation of Berger and Fisek (l970). (l) Individuals either give or do not give action opportunities to others, and they either receive or do not receive action opportunities from others. An action opportunity such as being asked for an opinion is a chance to perform. Indivi— duals expected by others to perform well will be given more action opportunities. (2) Once given an action opportunity, an individual either does or does not take this chance to contribute a performance output to the group interaction. (I would also suppose that acceptance of an action opportunity is more likely if the person ex ects himself/herself to perform better than other members.) (3) Given a performance output, others will evaluate that out— put either positively or negatively. Such evaluations appear to be colored by initial performance expectations (based on diffuse status characteristics, for instance); the same per— formance output by a high—status member would receive higher evaluations than if it came from a low—status member. Such evaluations are reflected in observable positive and negative reactions directed at the performer. (4) During the interaction, differences of opinion become apparent. Sometimes an individual is influenced to change opinions and sometimes he or she is not. Again, an individual who expects that another has more competence will be more easily influenced to change his or her mind. ——————-.-_—.—_____ 411 should note that the standardized experimental situation with which Berger tested his theory operationalizes only 92g component of the observed power and prestige order—~that of influence--while experimentally controlling for the remaining three (initiation of activity, receipt of activity, and judgments of task ability.) In principle, any other compon— ent could be chosen instead of influence. Of the four components of power and prestige orders, influence corresponds most closely to Blood and Wolfe's view of marital power, in lo the can be and pr status output expect gender to pet hoveva have ' total (abou chara istic DUNE)” opp he ot 66 To the extent that either equal-status or unequal—status ad hoc groups can be compared to marital dyads, it appears to me that marital power and prestige orders should be shaped by pgpfl initially present diffuse status characteristics (resources) and by evaluations of performance outputs during early stages of the relationship. Initial performance expectations based upon diffuse status characteristics (including gender42) may be powerful in that they determine who gets an opportunity to perform a particular task. Once a spouse gets this opportunity, however, his/her performance may be so poor that initial expectations have to be revised. For marriages do not usually exhibit either the total uncertainty (about task—relevant competence) or the ambiguity (about whether a performance output is correct or incorrect) that are characteristic of Berger's experimental groups. Hence, status character— istics (external resources) should affect the distribution of marital power but not to the degree that this occurs in such special ad hoc groups. that influence and final—decision power both occur toward the end of an interaction process. The two views of power differ, however, in whether 0 osition must be overcome. My own study takes Berger's position that influence lpower) can only be said to exist when some opposition has to be overcome. 42The ”strength” of gender as a diffuse status characteristic in premarital or marital dyads could vary greatly. If it is not seen as a status characteristic at all, one would expect that: (a) The relationship between other status characteristics (resources) and marital power will not differ between men and women. For example, a wife would gain as much power from a given income-advantage as does a husband. (b) Power and prestige orders will be more affected by direct evaluations of early performance outputs (than in cases where gender is associated with preconceived performance expectations). never 1 charac evalua detern situat subiec equal charac Answer situai exist 67 Development of Expectation State Theory In task—focused laboratory groups, composed of strangers who never see one another and know only how they rank on one diffuse status characteristic, the theory finds support; having no other basis for evaluating relative competence, the status characteristic distribution determines the distribution of influence. But this is an unusual situation, not at all like “real life“. Would the theory also hold if _subjects were differentiated on one status characteristic while being equal on another? Would it hold for specific as well as diffuse status characteristics (resources), some of which are inconsistently evaluated? Answers to these questions would move the theory closer to marital situations, where many specific and diffuse status characteristics exist in combination. As would be expected, a discriminating diffuse status char- acteristic proved to be maximally “effective” in determing influence in laboratory situations when there was no other activated status characteristic on which the subjects were equal. ”Activating“ a status characteristic in a laboratory situation is done by the experimenter. When spouses are left to choose from multiple sources of status informa— tion, they might or might not ”activate” diffuse status characteristics on which they are equal.43 M 43Take the example of two couples in which the wife has more education and in which spouses are equal in income. Both must solve the same “task“, that of deciding how much life insurance to buy. In he; s with b) tiatioi expects "abili' that it to the in whi being two cl were were that 68 The question of whether specific status characteristics, i.e., specific performance abilities, also affect influence was dealt with by Berger, Fisek, and Freese (l970). They found that differen- tiation of subjects on one specific status characteristic had the expected effect on influence distribution even when the specific “ability” was unrelated to the task. However, if subjects believed that they were high on one irrelevant ability and low on another one, they “cancelled” each other out (Freese, l969). If the specific characteristic is dissociated from (rather than merely being independent of) ability at the task, however, its distribution does not determine the distribution of influence (Kruse, l967). What about specific status characteristics which a£g_relevant to_the task? Berger and Fisek (l970) and Tress (197l) did experiments in which two specific status characteristics were ”activated“, each being egually relevant to the group task. In the condition where the two characteristics were inconsistent, resulting influence measures awere intermediate between the conditions where both characteristics were high and where both were low. These and other findings suggest that subjects "combine" or "average“ inconsistent status information. both couples, the husband and wife disagree on this. The first couple considers both education and income as relevant to the issue at hand. That is, both status characteristics are "activated”; competence at the task is assumed to increase with both education and income. The wife gains somewhat more influence than the husband. The second couple may act only on the basis of education and ignore income, Giving the wife a great deal more influence. with diffs sinultanev ing behav to a nari fonoulati about eac be define task he a to SUCCEE tion to 1 develop _a this by ' exlectat‘ ' form a g. exlectat tion (p9 to the t istics, iiitinuo if one p: lltstige 69 Generalizing the Theory The l966 Expectation States Theory was extended to encompass ppph diffuse and specific status characteristics and to include the 'simultaneous operation of multiple status characteristics in organiz- ing behavior (Berger and Fisek, l970). This moved the theory closer to a marital situation, although the scope conditions of the revised formulation still required that the gply information subjects have about each other be about status characteristics, that task outcomes be defined in terms of either "success" or ”failure”, and that the task be a unitary and collective one in which members are motivated to succeed (they are not only engaged in the task but are task—oriented). For purposes of the present study, the most important addi- tion to the 1966 formulation was the assumption that actors will develop aggregated expectation states for self and other. ‘They do this by "combining" multiple status characteristics. Performance expectations based on separate status characteristics are combined to form a general or overall judgment of competence level (performance expectation level). In addition, each component performance expecta— tion (perceived ability) is weighted according to degree of relevance to the task. Unlike expectations accompanying single status character— istics, aggregated expectation states are not dichotomous but are continuous variables. The greater the aggregate expectation advantage of one partner over the other, the greater that person's power and prestige advantage. (More precisely, power and prestige position is posite expect acceie 70 posited to vary as the sguare of the expectation advantage. Small expectation differences will have negligible effects but effects will accelerate with increasing expectation differences.44 Relevance of expectation state theory to mgrriage: In its present form, Berger's theory, of course, was not intended to apply to ongoing social groups. It is aimed at answering the general question: what do people do when they are in a situation of uncertainty? What kind of information do they use to define a situation--reducing some of the uncertainty—~50 that they have some basis for action? In the absence of any clear information (about competence, for example), they use information about status characteristics as clues. Families-- even when faced with ambiguous tasks——should not need such status clues in order to ”get by”, since they know each other intimately. Indeed, Leik (l965) found that status characteristics lose their "effect" when group members know a great deal about each other. He compared three-person family groups with three-person pseudo—families (the 44A mathematical equation based upon this notion was remark- ably successful in fitting data from two studies (Berger and Fisek, l970; Berger, Fisek, and Crosbie, l970). These studies both dealt with two equally—weighted status characteristics in five combinations of high and low, consistent and inconsistent characteristics. The Berger, Fisek and Freese study (l970) discussed earlier also provides some support for the assumption that expectations are weighted by degree of task-relevance of a status characteristic. Al- though specific status characteristics “determined” the influence distribution whether or not it was made relevant to the task, the effect upon influence was more marked in the “relevant" condition. However, the above mentioned mathematical equation did not "fit“ their data as closely as in other studies. latter husban disapp istics with it Under studie relati beconx inilu give : (resou foma lower 7l latter including the same individuals as the former groups). While husbands in pseudo—family groups were more dominant, such differences disappeared in actual family groups.45 Yet the correlations between such diffuse status character- istics (resources) as income, education, and occupational prestige with marital decision—power cannot be so easily explained away. Under certain conditions, e.g., in the permanent aircraft crews studied by Torrance (1954), ongoing social groups gp_exhibit a direct relationship between status characteristics and influence. The question becomes: under what conditions will status characteristics affect influence or power in families? In speculating about what these conditions might be, I will give separate consideration to the link between status characteristics (resources) and performance expectations and the link between per— formance expectations (perceived competence) and the distribution of power. (a) Conditiopg under which status characteristics may determine performance expectations in families: Status characteristics ought to coincide with performance expectations when, as Blood 45One could argue that this is the result of an "averaging out" of status characteristics as posited in the revised expectation states theory. I think that this explanation is highly unlikely. If all status characteristics of each family member were aggregated, the father would surely exceed the mother who would exceed the daughter in “aggregate status." Any selection (activation) of some subset of such characteristics would, more often than not, also tend to favor the father. 72 and Wolfe suggest, status characteristics are associated with experiences leading to actual competence (providing that this competence is recognized and is seen as a com- petence which has some bearing on the issue at hand.) This approach requires some way of assessing actual competence, which is not possible in the present study. In addition, it ignores the more interesting part of the theor -—that status characteristics can shape performance expectations even when they have gp_bearing upon the task (when there is no reason to assume that they would be associated with com— petence at this particular task). When spouses have access to more direct information about their relative competence (as they usually do), this information will sometimes be inconsistent with the distri— bution of status characteristics. Such an ”inconsistent” situation is the most theoretically interesting, for purposes of testing Expectation State Theory in marital dyads. The major question is: Under what conditions will performance expectations be based upon status characteristics rather than upon evaluation of past performance outputs, when the two sources of ”information" would predict different expecta— tion patterns? Status characteristics may override contradictory sources of information, I would suggest, under the following condition: In the pi propositi coupetenc (b) 73 When the task is ”new”, i.e., when there is no basis for assuming that past performance outputs (on other tasks) indicate ability relevant to this task. In other words, expectation state theory should apply to families when the scope condition of uncertainty about task-relevant competence of spouses is met. In the present study, I do not have sufficient data to test this proposition. In addition, instances of high uncertainty about competence would seem to be relatively infrequent in marriages. (b) Conditions undgr which performance expectations may determine the distribution of influence in families: This seems to me to be the more interesting area of inquiry: under what con- ditions do performance expectations (regardless of their base) pattern the distribution of power and prestige? I can think of instances where perceptions of competence would ppt determine the distribution of power in marriages. A wife who believed herself to be more competent might will— ingly yield influence, knowing full well that the quality of task—performance would suffer as a result. This might be the result of strong norms that the husband ”should” have the say in a particular area. (Note that this is not the same thing as saying that gender is a ”powerful” status characteristic in shaping performance expectations. The case under considera- tion is one in which the wife's performance expectations are 391 consistent with her gender status). 74 The same wife might also yield influence, believing that the loss in task—performance quality would be made up of gains in non—task areas. For example, this may benefit the affective quality of their relationship by increasing the husband's satisfaction and sense of participation. Or she may view the task-performance as a learning situation for the husband in which making “mistakes“ may even be a valuable part of the learning process. In all of the above examples, ”success“ at the task is not the primary motivation: Berger's scope condition that the group must be strongly task—oriented is not met. I would, therefore, hypothesize that performance ex— pectations will be more strongly related to the distribution of marital power and prestige when doing the task well is perceived as important, i.e., when the consequences of poor task-performance are seen as serious. Even strongly held norms may be violated if there is enough at stake, enough to be gained by distributing influence according to perceived competence rather than in accordance with normative gender—role expectations. The importance of the task thus becomes an additional variable. Expectation State Theory will most closely apply to family behavior when families are the most task—oriented, i.e., when they meet this scope condition of the theory. —'~_ .... A. . .. ”...“.me - M ‘ hypothese expectati "lrauslat Hypo Several x terns thx about re‘ Hm 75 Hypotheses My first hypothesis is that aggregated (overall) performance expectations will pattern the distribution of power and prestige. “Translated” into resource-theory terminology, this hypothesis is: 46 The higher the perceived overall competence of Hypothesis C: a spouse, the greater his or her overall marital power. Several conditions should be added to this hypothesis. The first con- cerns the fact that there are two parties to a marriage, whose perceptions about relative competence may or may not coincide. Hypothesis C-l: The greater the agreement between spouses in perceptions of relative overall competence, the stronger the relations: a. between the wife's perceptions of relative competence and her relative marital power; b. between the husband's perceptions of rela- tive competence and his relative marital power. M 46 _ Hypotheses are labeled with ”C” for competence, “G“ for those relating to gratitude as a theoretical mechanism, and ”D" for power— dependence theory. 76 Agreement about competence, of course, ‘mplies that resulting marital power patterns are also based upon consensus.47 This is what Blood and Wolfe assume about the nature of marital power. They also assume that competence is seen as relevant to marital decision-making. In the present study, it was not possible to measure relative "weights“ of specific status characteristics (specific competences) or to determine which ones are “activated” in a given situation. I can, however, determine whether a particular power—outcome is perceived as resulting from considerations of competence (regardless of the source of that competence-evaluation). Blood and Wolfe suggest that a spouse often complies or yields power because he or she recognizes the greater competence of the other. If that is the case, “competence“ should be offered as a rationale for power outcomes at least as often by the less-powerful spouse as by the more—powerful one. If anything, the less—powerful persons should men— tion competence considerations mppg often. Hypothesis C—2: The frequency with which ”competence” is given as a rationale for specific power—outcomes either: a. will not differ between spouses having high and low power in that specific area, 93 b. will be higher for low—power than for high— power spouses. 47An additional problem with “agreement about competence” is that such agreement may be the outcome rather than the "cause“ of power. A powerful spouse may influence the other to change competence perceptions. Dissonance theory might predict that such shifting of competence perceptions Dy the weaker spouse might result (even in the absence of influence attempts) in order to decrease cognitive inconsistency. In this case, perceptions of competence would shift over time, with the less-powerful spouse shifting per— ceptions more than the other. 77 If highrpower spouses mention competence more often, I would suggest that competence is not the “real" basis of the power exercised. Talk about competence may serve mainly as a justification for power arrangements, which actually have some base other than competence. Or such arguments may be a strategy48 for obtaining compliance from the other spouse, e.g., by devaluing the other's competence in order to gain decision—making power or by feigning incompetence in order to avoid an unwanted responsibility. In either case, differences in com— petence are not the major basis of power. Strong support for this hypothesis would require the measurement of actual competence or, at least, measures of perceived competence prior to marriage (neither of which are available from the present data). Under these circumstances, I would not consider hypothesis C—Z to be either supported or not supported unless the findings are highly significant. Let us now turn to expectation state theory. I suggested that perceived competence should be more strongly related to the dis— tribution of marital power when successful task performance--doing the task well--is seen as important (i.e., when Berger's scope condition of task—focused groups is most nearly met). Hypothesis C-3: The relationship between perceived competence and marital power will be stronger for ”tasks” 48With my data, I have no way to determine whether arguments about competence are deliberately used as a power strategy. Probably the use of power in marriages is usually not as ”cold blooded” or con— scaously motivated as the word “strategy” would imply. 78 judged to be important than for those judged to be less important. This hypothesis, in effect, would involve testing hypotheses C and C—l separately for high—importance and low—importance tasks and then comparing the results. What effect would the agreement or disagreement about task importance have? One would anticipate that each spouse rates his or her areas as especially important, for example. Hypothesis C—3 ought to receive the clearest and strongest support when both spouses agree on task importance. Hypothesis C—4: The greater the agreement between spouses in judgments of relative importance of “tasks,“ the stronger the support for hypothesis C—3 will be.49 Spouses may choose to g§§_what competence they have, more often for tasks which they view as being important. Blood and Wolfe suggest that high-status husbands take over more financial decisions because these decisions involve larger expenditures than in low—status families (l960: 33). The consequences of poor decisions are potentially greater, so the husband recognizes the importance of these decisions and takes an interest in their execution. Since high—status husbands are also assumed to be more competent at such things, both competence and importance of the decision (to the husband) operate in the same direction. M. 49This hypothesis was not directly tested in the present research. For related hypotheses, see Chapter VII. it wouli task in ceived a "task Hr itself 1 power ' to yie' issue a insistr cowpli in are 79 It would not be inconsistent with this approach to hypothesize that task importance by itself should affect marital power, even when per- ceived competence is held constant. The spouse who ”cares more” about a "task“ will take more initiative and will presumably want more say. Hypothesis C—5: When perceptions of relative competence are con— trolled: the greater the discrepancy between spouses in perceived importance of a “task,” the greater the relative power in that task—area of the spouse with the higher importance evaluation. This hypothesis is not central for the competence formulation itself, but it shares with that theory the assumption that most marital power is based upon consent. For example, husbands are thought likely to yield to wives' preferences when she cares more than he about the issue at hand, and vice-versa. Or it may be that one simply is more insistent about having a say when the issue is important. Either way, compliance is directed at allowing one's spouse to increase satisfaction in areas where he or she cares a great deal. I have now presented six hypotheses, the most central of which is the first. All are consistent with resource-theory notions about competence and marital power, although I have taken additional ideas from expectation state theory (in particular, the hypotheses about task importance). The major change I have made is the substitution of per- ceived competence for “actual” competence, which again was suggested by the work on status characteristics and expectation states. 80 This concludes my discussion of perceived competence as a theoretical mechanism by which resources are translated into marital power. The second type of ”equal—input” theory, in which gratitude is the central element, is closely related to the competence formula— tion. In many cases, it would predict the same power outcomes. Separate treatment is justified because it provides a clearer case of the ”equal—input" assumption-—an assumption which is in clear contradiction to power—dependence theory, as will be seen in Chapter VII. Gratitude and Respect j Another explanation for the greater power of high—status husbands, as compared with low-status husbands, is that they earn their wives' ”respect" by their success in the community (Blood and Wolfe, l960: 30). This success, of course, benefits the wife by giving her a high standard of living. She also shares the community prestige which accompanies her husband's occupational status. She is grateful for these benefits he provides and feels indebted to him.50 The exchange here is one of socially—valued external resources in return for compliance (as contrasted with the competence formulation, where competence based upon external resources--rather than the resources themselves--is the exchange commodity). To the extent that competence '——-————.—-—_—_—_ 50For example, the finding that suburban Detroit husbands exceed City husbands in average power receives this explanation: “Perhaps sub— urban wives feel indebted to their husbands for providing them with a place to live which is more attractive than the industrial city of Detroit“ (italics added; Blood & Wolfe, l960: 36). 81 is indeed rooted in experiences of external-resource procurement (or is perceived as such), the competence and gratitude hypotheses would predict identical distributions of marital power.51 The gratitude explanation implies, I think, the presence of some norm of equal exchange. Spouses feel obligated to contribute equal “inputs" or contributions. In a similar manner, Blau (1955) analyzed the exchange among office workers. Job—related assistance, when it could not be returned in kind, was repaid with "deference.“ Deference, which includes the yielding of power, becomes an exchange j commodity in itself. The greater the gap between the "resource“ con- tributions of two spouses, the more deference is required by the lower contributor to equalize the exchange.52 For example, take the finding that the balance of power shifts somewhat in the wife's direction, when wives are employed (Blood and Wolfe, 1960: 32). Presumably the “working” wife, by obtaining more resources of her own, has now narrowed the gap between what he and she have to offer in their exchange relationship. She needs to give less ”deference“ than the non—employed wife to fully repay her debt for re- sources received from the husband. 5IOne may, of course, also be ”grateful“ for a spouse’s contri— butions of competence, as well as for direct contributions of resources such as income or prestige. This interpretation of the competence formu- lation makes the “equal input“ assumption more explicit. 52Note that the presumed norm specifies that it is “inputs,“ 22: (Power) ”outputs,“ which should be equal. In fact, one sacrifices outputs in order to equalize inputs, provided that one has no other means of equalizing them. by whi cowpet follow H More s betwee equali becaus are or syndrr more x the s. econox those 82 If “gratitude and respect" are presumed to be a major mechanism by which external resources (or, for that matter, contributions of competence) are translated into marital power, one would predict the following: Hypothesis G: The greater the gratitude to a spouse, the lower is the marital power (of the grateful person). More specifically, one could state that: the greater the inequality between spouses in their gratitude to the other, the greater the in- equality of marital power (to the advantage of the l§g§_grateful spouse). In the present study I cannot test this hypothesis directly, d because I have no attitude measures of “gratitude.“ Indirect tests are possible, however. One deals with what I call the “Cinderella syndrome." Cinderella, once married to the prince, ought to be much more grateful for benefits received than would a princess married to the same prince. In like manner, wives who grew up in lower socio- economic homes ought to be more grateful to “successful" husbands than those who came from more advantaged backgrounds. Hypothesis G—l: The greater the external-resource advantage of one's spouse over one's parents, the lower one's marital power. This hypothesis, I should note, would also be consistent with power-dependence theory; the greater the ”motivational investment” in goal-objects (valued resources) mediated by the spouse, the greater one's dependence upon that spouse, and the lower one's relative marital power. Motivational investment-~the value placed upon the spouse‘s resources—-is 83 arrived at by a social comparison process. In this case, the compari- son is against one's family of origin. Power~dependence theory, how- ever, would focus upon comparisons with current alternative sources of goal-objects, such as potential alternative spouses or employment. The research of Pearlin (1975) on cross-class marriages is relevant to Hypothesis G-l. He found that class-of—origin operates as an exchange commodity in itself, i.e., that the effect of having a lower class-of—origin than one's spouse is not “cancelled out“ by up- ward mobility (high present resources). Lower—origin spouses were apparently “grateful" to higher-origin partners, reporting higher levels of marital satisfaction than persons who had "married down.” This effect, however, occurred only for persons with strong desires for success (in power-dependence terms, with high motivational invest- ment in getting ahead). Among such social climbers, marital satis- faction was increased by having a higher-origin spouse and decreased =by having a lower-origin spouse. The "gratitude“ formulation diverges from that of power? adependence theory in the role played by alternative sources of goal-objects. lOne cannot be "grateful“ for resources which the other spouse potentially :could acquire (but does not); yet in power-dependence theory, such Epotential resources are postulated to affect the balance of power. It is now appropriate to discuss power-dependence theory in some detail and, in the discussion of critical tests in Chapter VII, to return to the question of potential alternative sources of resources. D an. M u ..s r CHAPTER VI A POWER-DEPENDENCE THEORY OF MARITAL POWER David Heer (1963) suggested a theory which he thought could better account for empirical findings of marital power than did Blood and Wolfe's explanations. Blood and Wolfe, he said, could not explain why certain external resources such as income led to marital power, while internal resource contributions did not. For example, wives generally increase their contributions of time and labor upon the ad- 53 vent of a baby. Yet this increased contribution, although it pre— sumably benefits the husband, apparently wins her no “gratitude”; at least, it wins her no power. Her power sharply drops at this time (or appears to do so, judging by cross-sectional data). It certainly drops faster than she would exhibit any loss of “competence.“ Thus it would seem that neither competence54 (knowledge or skill) nor gratitude can fully account for marital power distributions. 53Power of wives also varies inversely with the number of children (Heer, 1958: 341-347; Blood & Wolfe, 1960: 134), although the mother of many children undoubtedly contributes more “internal" resources than the mother in smaller families. 54Heer admits that relative competence may indeed have some role in determining outcomes of uncontested decisions. He feels, as I do, that power is better measured by outcomes of contested decisions (that there must be some disagreement or ”resistance“ present before one can say that power is operating). When spouses disagree, Heer suggests that relative competence is probably an insufficient explanation of who pre- vails. Blood and Wolfe's decision-index of power did not distinguish be- tween contested and uncontested decisions; most likely, the majority of the component decisions were of the uncontested variety. 84 85 Heer's ”revised theory” has its roots in the work of Homans (1961: 55-56), Waller's principle of least interest55 (Waller and Hill, 1951: 190-192), and—-although not acknowledged—-probably also in Thibaut and Kelley's (1960) concepts of power and dependence. In Heer's words: . . . Here the focus is not on the value to each spouse of the resources contributed by the other but on the value placed on these resources outside the marriage. According to the revised theory, the greater the difference between the value to the wife of the resources contributed by the husband and the value to the wife of the resources which she might earn outside the existing marriage, the greater the power of her husband; and vice—versa. This theory ex— plicitly states that each partner to the marriage conceives of the possibility of separation, divorce, and subsequent remarriage. (Heer, 1963: 138). Thus, the mother of pre—school children has low power because her potential returns, outside the marriage, have lowered. Perhaps partly because she likely would have custody of the children, her "bargaining power in the remarriage market is probably not very high“ (ibid.). Until the children are in school, working for pay is a problematic alternative. In short, the returns she might expect under some alternative to her present marriage are rather meager. Similarly, the woman married to a successful man realizes she has made a relatively good "catch", probably better than any other husband she might find. She is wary of contradicting him if he really feels strongly about some- thing, for fear of jeopardizing the relationship. 55This principle states: the partner with the least interest in the relationship is the one most apt to exploit the other. Waller, according to Heer, does not explain gpy_this is so. 86 So the spouse with more to lose (should the marriage end) is at a disadvantage. Such a spouse is more dependent upon the partner than vice-versa, although the dependence need not be financial.56 Dependence is thus the basis of marital power: the greater one's dependence upon the spouse, the lower one's relative marital power. As Heer does not state explicit hypotheses, I have drawn upon the work of Thibaut and Kelley (1960), Emerson (especially his 1962 paper) and Dahlstrdm (1966) to formulate hypotheses that may help to overcome the theoretical impassé referred to at the end of Chapter IV. Thibaut and Kelley (1960) Power, rather than being the ability to affect psychological and behavioral changes (e.g., compliance) in the recipient, is defined as the ability to affect the ”outcomes“ of another (in the sense of causing changes in Other's cost/benefit balance). Such outcome 56Safilios-Rothschild (1970: 548f) suggests that emotional dependence operates in a similar way: if he loves her more than she loves him, then he is the more dependent. He has more to lose (not only in case of divorce, but, I would suppose, from temporary with— drawal of affection during disagreements). The greater social isolation of suburban working—class housewives, as compared with those in older working—class neighbor— hoods in the city, is noted by Tallman (1969). It could be that Blood and Wolfe's suburban wives had lower power than city wives not, as they suggest, because they were ”indebted“ to their husbands for providing them with more pleasant living arrangements (see Blood & Wolfe, 1960: 36). Rather, they may have had fewer alternative sources of emotional support and assistance. The network of relatives and neighbors in stable working- -class neighborhoods may make the wife less dependent upon her husband, increasing her marital power. 87 changes, of course, are assumed to affect behavior. Thibaut and Kelley's ”power“ is still defined as an ability of individuals (as in Weber's definition of power), but now both individuals (in a dyadic situation) must be considered. Power and also cost/benefit outcomes of both actors are treated, each actor from his or her own perceptual point of view. A's power is the extent that Actor A, by varying his or her behavior, can affect the quality of B's outcomes. B's counter— power is the extent that B can affect the quality of A's outcomes. The key to ”calculating” power lies in the calculation of outcomes. Outcomes are subjective, based on a “scale“ where the size of the increments appears to be determined by the value of the (experienced Or imagined) outcome received. The ”zero point” of the scale, which ferms the comparison base, is that level of outcomes which the actor might attain in his or her best available alternative situation. This zero point is his or her Comparison Level in the next best alternative, or ”CLaltu' Interacting persons affect each other's reward—cost posi— tion as measured on this scale and thereby can influence or control each other. Furthermore, the average level of outcomes of a relationship :annot fall below the zero point of CL for very long or else the alt relationship would be terminated in favor of the next best alternative. The central process here is one of social comparison, somewhat dmilar to reference group theory. It is not enough for A to give B ome output (“resource”) which B values; in order to reap power, A must ive B more of that output than he or she could receive elsewhere. 88 If B has pp_counterpower, A's power equals the distance between B's average outcome level and B's CL This is called "fate control”. alt' A's power in this case is exactly equal to 8'5 dependence, dependence being defined also as B's average level of outcomes minus B's CLalt’ The degree of A's power, under fate control conditions, rests on his ability to supply high rewards at a low cost to himself. Supplying rewards and even “using” power can be costly to A. Potential power hence is greater than usable power. Since marriages are seldom “fate control“ situations, let us examine Thibaut and Kelley's ”behavior control“ case. B now has some control over the variations in his or her outcomes by adjusting be— havior to “synchronize” with A's desired behavior alternatives. A's power depends upon A's ability to make it desirable for B to synchronize. By not synchronizing, B has some counterpower with which to decrease A's benefits and/or increase A's costs. Before A decides to use potential power, A must calculate not only the direct power costs of power use (see above) but also the indirect costs which are possible because of B's counterpower. A and B usually have several behavior choices which are possible. Gains of each are a function of the interaction of the choices of each. Some of B's behavior choices will yield higher gains for A than others. A must make it sufficiently desirable for B to choose that coordinated behavior which leads to higher outcomes for A. Power of A is no longer strictly equal to 8'5 dependence upon A. Because of B's counterpower, A must increase the rewards aVailable to B 89 (to induce B to select a behavior alternative which is beneficial for A). A does this at some cost to himself or herself, a cost which enters the ”equation”. A may increase power over B by increasing B's dependence, i.e., by raising B's outcome level (relative to 8'5 CLalt)' In marriage, there is seldom only one ”desired object“ such that one party must gain at the expense of the other. Often both can gain. If so, Thibaut and Kelley suggest that goals tend to be coopera- tive and the mutual rewards lead to identification. Identification, in turn, may make the act of delivering rewards to the other a source of satisfaction in itself. Outcomes increase even more57, and norms may develop which stabilize the relationship. (Competitive relation— ships found in the win-loss pattern are less stable and may lead to a strategy of alternating who gets the good outcomes.) In short, high levels of interdependence——where each partner has high ability to in— crease the outcomes of the other—-should lead to high satisfaction.58 57Dahlstr'om (1966: 246) calls attention to the underlying assumption in such arguments, namely that rewards are assumed to be stronger “reinforcers” than are punishments, and hence have more im- pact upon the learning process. 58Outcome levels are increased by deletion of costly conflicts and by the addition of new satisfactions. Satisfactions are increased by the convergence of values and attitudes of the parties over time. Thibaut and Kelley specify four mechanisms by which convergence occurs, based upon learning theory. Two are “associative processes“. Aspects of A's behavior associated with B's rewards come to acquire reward value for B in themselves, increasing the reward value of A's behavior to B. B's own behaviors may come to have increased value to him or her in and of themselves, after being consistently rewarded by A. ”Motivational processes” include increases in output as a re— sult of conflict avoidance. Such cost-reduction can occur during : ;W’L‘.‘.x --t-—-—-——-—-_.__~_L 90 Groups which are highly dependent on each other have been compared to slightly dependent groups by previous researchers. Back (1951) found that, regardless of the type of attractiveness upon which group cohesiveness (interdependence) was based, highly cohesive groups reached more agreement and also showed more overt resistance to others' suggestions. Although both power and counterpower are high, the development of high mutual outcomes over a long time is posited to reduce the likelihood of severe conflict (elimination of costly con- flicts raises the outcome level, of course). Conflict—avoidance strategies under conditions of high counterpower include the defining Vof limited zones of conformity and allowing the slightly more powerful member to be more influential. Perceptions are central for Thibaut and Kelley.~ Reward-cost outcomes are perceptual. Synchronization of behavior requires that B be aware that his or her outcomes could be improved by such synchroni— zation, etc. So it is not surprising that most of the strategies cited for increasing power are also perceptual. A less—powerful husband (Actor B) can increase his power by five strategies: (a) increasing the value of B's product to A (or reducing the value of A's product to himself); (b) limiting opportunities for A to acquire new skills or inhibiting A's performance level; .m— decision—making, when relative evaluation of alternatives are changed. After the decision, dissonance reduction can be attained by changing relative evaluations of objects of choice. The two associative mechanisms clearly lead to greater attitude and value consensus and the two motiva— tional mechanisms may lead to greater consensus also, although this is less clear. 91 (c) delivering rewards which make A's counterpower less usable (because now she has more rewards to lose); (d) raising his own CL 1t or lowering A's CLalt (perhaps by manipulating alternatives available); (e) secretly lengthening his time perspective, thus temporarily lowering his own CLalt, which increases the value of rewards received from A. Thibaut and Kelley's concept of power is only partly a pro— . perty of individuals, for the ability to affect outcomes of the other depends upon the relative outcome position of A and B. The power of A and B are said to be conceptually independent of each other. In part, this appears to mean that power is not zero-sum in the sense that B's power must decrease as A's power increases. Bppp_may have high power, as in the situation where both have low CLalt levels. They do not distinguish positive from negative (e.g., “veto”) power, nor is “resistance“ as central a concept as it is for such power theorists as French and Raven (1960). “Power" and “dependence” are 1323 conceptually independent, as both are defined in terms of CLalt’ Part of the value of Thibaut and Kelley's work lies in the delineation of determipggts of the exchange rate between A and B. Another contribution, incorporating some of French and Raven's ideas, is the linking of an interdependence level with the probability of attitude and value convergence. Thibaut and Kelley assume that people are basically rational, in the sense that they will act so as to maximize outcomes. Blood and Wolfe, in contrast, seem to assume that people are rational in the sense that they will allocate decision—power to the person most quali— fied to make good decisions—~that is, that couples will act so as to 92 maximize group (rather than individual) outcomes. Whether such “rational" models correspond with family behavior should be ascertained empirically, if only to find out the extent to which a rational calculus is, in fact, employed. Emerson (1962) Richard Emerson apparently explicated and simplified Thibaut and Kelley's ideas about power and dependence (although he does not acknowledge them). Noting that it is meaningless to state that A has power unless one also states “over whom“, Emerson stressed the idea that power is the property of a social relation rather than an attri- bute of an actor. Social relations are said to be characterized by ties of mutual dependence. Power to control or influence the other [I “resides in control over the things he values . . . , which corres— ponds to Thibaut and Kelley's definition of power as the ability to affect the outcomes of another. Dependency is the basis of power. The dependence of A upon B is (1) directly proportional to A's motiva— tional investment in goals mediated by B and (2) inversely proportional 59 to the availability of these goals to A outside of the A-B relation. This is similar to Thibaut and Kelley's (1) value of outcomes and 59Emerson supplies an economic analogy for these terms: motiva— tional investment = demand, and availability = supply. Supply-demand relations thus determine the exchange rate or ”Price”. Thus baseball shortstops are less “valuable” than pitchers, pp; because they are less essential to the team but because good pitchers are harder to find. In the same manner, good housewives are less ”valuable” than good bread— winners because good breadwinners are scarcer. 93 The ppggp Of A over B is defined as the amount of resistance on the part of B which can be potentially overcome by A. The stress on resistance is a point of difference from Thibaut and Kelley. In Emerson's formulation, there must be some resistance of B to A's de— mands before the change in B's conduct can be attributable to demands made by A.60 As with Thibaut and Kelley, "power“ and "dependence" are not conceptually independent because, by definition, the power of A over B equals the dependence of B over A(and vice—versa). Also consistent with Thibaut and Kelley is Emerson's observation that the power of A and of B (over the other) need not be equal. If A's power and B's power are not equal, their relations are said to be imbalanced (one party has a power advantage, defined as PAB - PBA). The power of each party (PAB and PBA) is not viewed as neutralizing or cancelling out each other. For example, there can be balanced (equal—power) rela- stions where both have high power or where both have low power. ‘ Thibaut and Kelley's level of interdependence becomes, for Emerson, "cohesion“. It may be crudely defined as the average of DAB 60This stress on resistance is also stronger than in my defini- tion of power, as reflected in my operational measures. I do not limit “power" to situations where overt resistance occurs, but to situations where the other wishes for a different outcome than the one which occurs. This disagreement about preferred outcomes may be llilPHClt; indeed, A need not even know that B disagrees. I use this broader definition be- cause absence of overt resistance may itself be a VGIIGCtIO” 0f the pgpgr situation; low-power spouses may not express disagreement because they have little chance of gaining anything that way (and perhaps be- cause they have much to lose). ~ expei tive 94 and DBA’ which Emerson thinks is superior to Festinger's (1950) definition of cohesion because it takes into account considerations of availability of alternative goal—sources, in addition to “motiva- tional investment“. Emerson does not elaborate on the role of cohesion or interdependence, however. Thibaut and Kelley thought that highly interdependent groups experience convergence in attitudes and values over time, via associa— tive and motivational processes (see footnote 58). One result is that the cost/benefit balance of each party is increased, both by increasing benefits and by reducing costs (e.g., by avoiding major :onflicts). Emerson, in contrast, hypothesizes that imbalanced relations (regardless 0f level of cohesion or interdependence) will :end to change. They are unstable, for they encourage the use of mwer which, in turn, sets in motion processes of (a) cost reduction md (b) balancing operations (similar to Thibaut and Kelley's trategies). I . Cost reduction61 is a process involving value—change. Inter- alization during socialization and identification with the aggressor re given as examples. B lowers his amount of resistance and hence )wers his costs. However, unlike Thibaut and Kelley, cost reduction es not necessarily alter the fundamental imbalance of power. It ——————————_—._— 61Unlike Homans (1961, Ch. 3—5; 1966), Emerson and Thibaut & iley do not consider values foregone to be ”costs”. Rather, fore— ie benefits are dealt with in the variables of ”alternatives" erson) and "CLalt" (Thibaut & Kelley). For these theorists, ”costs" restricted to the present exchange relationship between A and B. lid): [with oreas inves decre 95 y, however, function to deepen and stabilize social relations ithin either imbalanced or balanced situations): Balancing operations necessarily involve increasing or de— easing dependence. Since dependence has components of motivational vestment and availability, there are four logical possibilities for creasing imbalance. If DBA< DAB’ this imbalance may be decreased by: l) B reduces motivational investment in goals mediated by A (in extreme form, this involves withdrawal of B); 2) B cultivates alternative sources for gratification of those goals (e.g., ”network extension"); 3) A increases motivational investment in goals mediated by B (exemplified by Emerson in a strategy of “status giving”); 4) A is denied alternative sources for achieving goals (as by coalition formation of B with other goal-sources). Two major differences exist between Emerson and Thibaut and ley. First, Emerson's second strategy (above) requires that B ually go to an alternative partner for some valued resources. In baut and Kelley's formulation, the key factor is whether B EQELB so (whether B could implicitly or explicitly threaten to do so). ential but “unused” alternatives thus can gain one power. I will iefining “alternatives” in the latter and broader sense in the ent study. Second, Emerson hypothesizes that imbalanced relations will toward balance or increasing eguality of power. Thibaut and :y make no such claim. Emerson seems to assume that an imbalanced 96 ate necessarily creates ”tensions“ in participants, leading to range--a perhaps uncalled-for, overly restrictive assumption of ilance. I will now summarize the major hypotheses I have drawn from nerson and Thibaut and Kelley, after which I will introduce refine— ents and extensions of these hypotheses based upon the work of imund Dahlstrdm (l966). )wer-Dependence Hypotheses: art I For both Emerson (1962) and Thibaut and Kelley (1960), the [SIS of A's power over B and of B's power over A is dependence. apendence is the degree to which one's outcomes (benefit/cost ratio) e controlled by the other. Hypothesis D: The greater the dependence of one spouse upon the other, the lower that spouse's marital power. 62 Dependence level, hence, is determined by an outcome-calculation acess. Outcomes available from the spouse are evaluated by comparing am with a mental “scale”, on which the zero-point (Thibaut and Kelley's 1t) is the highest perceived outcome level available from a source er than the spouse. A given outcome level available from the spouse more subjective value as the zero—point declines. Although 62One need not accept Emerson's assumption that the power of A equals dependence of B, in order to accept the above hypothesis. I will be 9 a definition of power which is conceptually independent of dependence. 97 easurement of precise CLalt levels is not possible here, one can )ughly distinguish between situations of greater and lesser alterna- ives. Hypothesis 0-1: The greater the value of goal—objects available or potentially available from a source other than the spouse, the higher will be the indivi— dual's marital power. Within a dyadic context, this hypothesis becomes: Hypothesis D-2: The greater the inequality between spouses in the value of goal-objects available or potentially available outside the marriage relation, the greater the inequality of marital power (to the advantage of the spouse with the higher alterna- tives). Under conditions of equal motivational investment in goal- )jects differences in availability reflect differences in dependence. )wever, motivational investment may also vary. The role of motiva- ional investment is to determine the increments on the outcome—evaluation ale. If one values a given exchange commodity highly (i.e., if one 5 high motivational investment in it), each ”unit“ received would ise the scale-value greatly. On the other hand, receiving a ”unit” a less-valued commodity would result in little increase in one's sub- :tive value gained. Hypothesis D-3: The greater the motivational investment in goal-objects mediated by the spouse, the lower the individual's marital power. When both spouses are considered: Hypothesis D-4: The greater the inequality between spouses in moti- vational investment in goal-objects mediated by the other, the greater the inequality of marital power (to the advantage of the spouse with the lesser motivational investment). 98 Note that alternatives and motivational investment operate together to determine dependence (and hence to determine power). Either an increase in alternatives (raising the zero—point) and/or a decrease in motivational investment (reducing the increments and hence the valuation of a given commodity received) can decrease dependence. However, motivational investment is even more intertwined with the level of alternatives than indicated above. Subjective values of ”commodities” are affected by previous experiences, i.e., by ppgp exchanges with other people (some of whom may be perceived as present “alternatives” to the spouse). A ”commodity” which has not been scarce or difficult to obtain in one's past experience would not have great value. For example, a ”dependable” spouse is highly valued' by someone whose experiences indicate that this is a rare 'phenomenon'. It takes on less value to someone who expects dependability as a 63 The ”Cinderella Syndrome“ discussed under the natter of course. iratitude hypothesis is another case in point. A wife should have weater motivational investment in financial resources contributed by "successful” husband if her past experiences indicate that such men re rare, than if she has known many such men. To the extent that the past exchange relations which provide ie basis of motivational investment level are similar to the present ternatives perceived to exist, the following hypothesis should hold: ~.—-———_—______. 63On the other hand, if dependability were expected but not esent, such a person might reevaluate the importance of this trait. I 1 imagine a situation where the other spouse '5 dependability fell below ‘ CLalt level. Dependence upon such a spouse might be extremely low, potential alternative spouses would offer higher dependability and the ue. of dependability is now seen as very high. 99 Hypothesis D-5: The greater the value of goal—objects available or potentially available from a source other than the spouse, the lower will be the motiva- tional investment in goals mediated by the spouse. Motivational investment and level of alternatives, hence, are not conceptually independent. For analytical purposes, however, they can be treated as separate variables. Power-Depgpdence Hypotheses: Part II Dahlstrdm (1966) presents a much broader analysis of power. I will not attempt to summarize his lengthy and detailed analysis, but will draw upon certain relevant parts of it. He makes the point that dependence upon another (for one's outcomes) is gpp conceptually equivalent to influence—power,64 con— trary to Emerson's assumption, because potential power—resources are 64Nor, says Dahlstrdm, is the degree to which preferences are *ealized the same as one's power. Achievement of preferences reflects bower only when this is due to influencing others. I take this into iccount by operationalizing power as achievement of preferences under he condition of initial disagreement between spouses (i.e., they must mve different preferences at Time 1 I do pp: consider the relation of Time 2 preferences to decision— utcomes to be an adequate measure of power. Spouses who initially differ- d in preferences at Time 1 may often agree at Time 2; the relation be- ween opinions or preferences before and after “decisions“ can itself be egarded as a measure of attitude conformity, an outcome of power (see arch, 1955: 431-451 and Hopkins, 1964: 79). A third point relevant to the measurement of power is Harsanyi's 1962) assumption that power—negotiations center on the relative fre- aency with which B is to emit a particular compliant behavior. Hence Jouses are concerned with how often the husband helps with the dishes rer a given time period, rather than with whether or not he helps on particular evening. 100 always used. The probability that A will use his/her power 65 A expects to gain from exer- ources depends upon the “utility” ing power over B, including the expected utility of B's compliance. 5 is A's power incentive. The compliance incentive of B is the 66 from A's power over B including the utility lity B expects to gain yielding to A's influence. The act of compliance, in itself, may costly from B's point of view. The exercise of power, in itself, 'be a source of benefits to A (that is, power may be a goal—object ch is intrinsically valued in itself, rather than only a means to ier ends). In short, power presupposes a ”double incentive“ (pp. 256—7) ationship; both must expect some gain. Hence I would hypothesize that dependence (motivational in- .tment and alternatives) will be more strongly related to power when alues being a powerful person and will be less strongly related n B values being a powerful person. The former increases A's in- tive to use potential power (because A has more benefits to gain) the latter decreases B's incentive to comply (because B's costs now lude the loss of intrinsically—valued power). 65"Utility“ is an economic term used by Homans (1961). It "esponds to Thibaut and Kelley's outcomes, that is, to the cost/ afit ratio of an individual. 661f B is in a situation where he or she will get low utility outcomes regardless of which behavior—choice is made, B has no in— :ive to comply with A's wishes. A must make it a little more worth— e for B to choose a particular behavior in order to obtain compliance. 101 Hypothesis D—6: The relationship between dependence and marital power (in Hypotheses D-l through D-4): a) will be stronger when the less-dependent spouse considers power as intrinsically valuable, and; b) will be weaker when the more—dependent spouse considers power as intrinsically valuable.57 :or example, a husband who values being the ”head of the house” as a ioal-object in itself (beyond any benefits he may obtain from ”winning” particular disagreements) has more to lose by complying than a husband who does not. (Whether or not this “ego-investment“ in being a power— iul person is due to normative prescriptions or to some other factor is not the point here. However, this may be one mechanism by which iorms are translated into marital power.) A major contribution of Dahlstrdm is his consideration of iultiple issues. Although his discussion of multiple issues deals with - arger groups, it can appropriately be applied to dyads. Multiple ssues increase the probable “exchange of power resources" (p. 259), hich I interpret to include the probable use of power. Trade—offs re now possible. Person A, who ordinarily would not give in on one ssue, may be willing to comply in return for B's compliance on another ssue. Hypothesis D—7: The greater the number of areas of disagreement between spouses, the greater the average amount of marital power exercised (by either Spouse or by both spouses jOintly). 67This hypothesis is not tested in the present research. 102 heater exercise of power means, among other things, that trade-offs ill outnumber half-way compromises. Where compromises do not occur, :ade-offs can be inferred from the fact that each spouse "wins" part i the time.68 This effect ought to be the most apparent when the interests éspouses vary, that is, when their motivational investment in'a rticular area is different. If A and B feel strongly about the ggmg puted issue, trade-offs are less likely. But if A feels more ongly about issue 1 and B feels more strongly about issue 2, then h has something to gain by a trade-off. A is allowed to prevail issue 1 and B in issue 2; each "wins“ in an area which reaps high vjective benefits and each ”gives in" in an area where the loss of refits is minimal. Hypothesis D-8: Within a couple's areas of disagreement, the greater the average difference between their motivational investment in the solution of the disagreement, the greater the average power exercised in these areas. solution of a disagreement is now considered, in Emerson's terms, )al in itself. To have a strong motivational investment in a solu— 1 means, in this case, to have strongly held preferences. Say that 68In my study I may not have detected many existing trade-offs use I only measure a few specific issues. Trade-offs can exist out being consciously viewed as such. However, I did ask couples they arrived at a given allocation of responsibility, with one of 13 answer options dealing with an explicit trade-off agreement. 103 .nd B differ in their preference about how to divide the responsi— ities of keeping in touch with relatives and of deciding what ple to invite to the house. A trade-off is most likely when the e only has strong preferences about one issue and the husband only strong preferences about the other issue. Each is allowed to rail in the area where he/she cares the most. In Dahlstrdm's us, the more the interest of each participant varies, the more exchange of power.69 Finally, of course, there is the variable of how far apart differing preferences of spouses are. Dahlstrdm hypothesizes the smaller the extent to which constituent parts of participants' rests coincide in different issues, the greater the exchange of r resources. Hypothesis D—7 deals with the number of issues where rests of spouses differ. To that, I should add the following: Hypothesis D-9: Within a couple's areas of disagreement, the ‘ greater the average discrepancy between their preferences, the greater the average power exercised in these areas. hypothesis I find somewhat less plausible than the two preceding It implies that compromises will occur more often when spouses slightly differing preferences, and that trade—offs will occur often when their preferences are widely discrepant. The hypothesis ausible, perhaps, in larger groups or in situations where there is 691t should be noted that hypotheses D-7 and D-8 should not be nued as a tautological undertaking. Rather, both the number of con— 1 areas and the differpptial preferepce strength of both spouses in areas are hypothesized to vary with average marital power. Whether or not Dahlstrbm's concept of power exchange corres- to my usage of ”power” will be discussed in Chapter IX. It may be i, for example, that more or less explicit trade-offs decrease or 'wipe out“ each spouse's initial disagreement about how any parti- issue should be handled. Thus, one defining characteristic of would be ”dissolved" by way of a trade—off. 104 a option of pp: resolving the difference. Here the lowest level of aer—use is not the compromise but the stand-off. Such stalemates rally are not an option in marriages, at least not for the issue h which I am concerned. Someone p33 to do the grocery shopping; isagreement about who this should be will inevitably get solved t in the sense that both are happy with the solution, but in the se that someone gill end up doing the shopping). Dahlstrom also deals with the relation of power resources ilable and the costs incurred in the gain of influence. A person ing much higher power resources, he argues, does not actually i to “use up“ his or her resources; vague promises and threats are active without having to carry through on either (without having ive the promised resources, thus reducing the costs of power use , and without having to carry out threats, which might increase costs so much that his/her incentive to comply goes down). The er assumes that power can be ”used up“ if exercised too often; )utcomes may be so low that B even withdraws from the relationship. Hence, I would suggest that the greater the dependence arential between spouses, the more likely is A to use his/her tial power. Hypothesis D—lO: The relationship between dependence and marital power (in hypotheses D-l through 0-4) will be weak under conditions of small dependence— differences between spouses and will become stronger at an increasing rate as dependence- differences widen. 105 This hypothesis posits some kind of exponential function in e relation of dependence to marital power. Unfortunately, the data se in the present study does not allow testing of this hypothesis. This concludes my discussion of the power—dependence theory marital power. I will now proceed to the identification of test tuations, for the assessment of the relative adequacy of alternative eories. 'CHAPTER VII CRITICAL TESTS In the event that each of the theories of marital power :ceives some support, the assessment of their relative explanatory vwer becomes interesting. I propose to compare the adequacy of ese theories by constructing test situations in which the different eories would predict different results. These test situations ex- icitly provide for the possibility of obtaining negative evidence; e can have greater confidence in findings which support a theory der such conditions. Tests pertaining to theoretical processes or “translation :hanisms" are presented first. These are followed by a few tests rolving specific resource-variables, where a close connection of )ourcezpossession with translation processes may be inferred. Tests bf Alternative Processes etence Versus titude Theories q_ The competence and gratitude versions are closely related. fine it is perceived competence (presumably obtained via experiences fciated with external resources such as income or education) which %xchanged for compliance, while in the latter it is "resources" hselves which are directly exchanged for compliance. If spouses with a resources are also perceived to be resourceful (i.e., competent) per— :. then both theories predict identical patterns of marital power. 106 107 Although I do not have direct measures of “gratitude“, it 5 still possible to identify situations in which different pre- iCtions would follow from these two theories. Income is a resource 11Ch, at least in western cultures, has been shown to relate to writal power. This relationship presumably involves gratitude-for- icome. If perceived competence does affect marital power, then xis effect should be evident even when income (gratitude) is held hstant. Test 0-6 1: (Competence prediction) Controlling for relative income of spouses, the greater the perceived com- petence, the greater the marital power. an both spouses are considered, this hypothesis predicts the follow- 9 rank-order, from highest to lowest marital power: 1) Spouses agree in perceptions: high-earning spouse seen as more competent. ‘ 2) Spouses differ in perceptions of who is more competent, or spouses agree that both are equally competent. 3) Spouses agree in perceptions: low—earning spouse seen as more competent. ithis prediction is pp£_supported, one would have negative evidence inst the competence theory. One would not, however, have strong itive evidence in favor of the gratitude theory, because a direct icator of gratitude is lacking. A somewhat different strategy is to compare those couples ge income and "competence" are inconsistent, i.e., where predictions ‘ =ower based on resources differ from predictions based on perceived S 3L“- petence. This is Rosenberg's (1968) strategy for assessing relative 108 ffects of two variables. If resource-possession is more “potent" ian "competence", marital power of high-income/low—competence spouses would exceed that of those with high-competence/low-income. The >posite results would be expected, if competence is the more potent octor. Competence assumption Gratitude assumption predicts A’s power to predicts A's power CONDITIONS be: to be: A > B in competence . A < B in income Higher :Ewer vs vs A < B in competence l . A > B in income Lower Higher Test C-G 2: (Competence prediction) Where resource-advantage and competence—advantage do not coincide, spouses having (1) higher competence and lower resources will have higher marital power than those with (2) lower competence and higher resources. It may be, of course, that competence considerations are mary when the task outcome is important (Hypothesis C-4) and that atitude" mechanisms operate when task outcome is not so important. this is the case, one would expect: Test C-G 3: Hypotheses C-G l and C-G 2 will be supported under ,conditions of high task-importance but not under conditions of low—task importance. al-Input Theories Versus er-Dependence The gratitude theory and, to some extent at least, the competence pry can be said to assume a norm of equal exchange (i.e., that spouses obligated to contribute equal exchange "commodities," and that 109 xchange “debt“ will be repaid in deference or compliance, if no r coin is available). This exchange rule may be summarized as )ws: Unequal resource inputs-9 B raises inputs——>Equal total inputs A = B A> B by giving more Unequal output compliance (power) A > B Equal resource inputs Equal outputs (power) A = a A = B In contrast, a power-dependence formulation need not assume a of equal inputs (in the sense of equal contributions of exchange dities). Neither unequal inputs nor, for that matter, outputs70' sarily must shift toward equality. Although norms of "fairness" equality of exchange may exist, they are not presumed to override ffect of relative dependence upon relative power. The exchange for power—dependence theory deals not with inputs but with depend- 7'] ' . 70It is true that one power-dependence theorist, Emerson, adds an onal proposition that imbalanced relations (where dependence is un- will move over time toward a balanced state (where dependence is and hence, for Emerson, where power—outputs must also be equal). nequal dependence——;Balancing Operations-9Equal dependence A = B A> >B Equal outputs (power) A = B 5, he assumes that equal outputs will be the final result. He does owever, assume that inputs will eventually become equal (see foot- 1). 71Equal dependence is not the same thing as equal inputs, since 1 input of exchange commodities by A may ”yield” different amounts andence from B, depending upon B's alternative sources of these 'ties outside of the A-B relation. 110 F: Unequal dependence Unequal power A > B 5} A > B F: Equal dependence Equal power A=B >A=B Marital exchanges involve many different “commodities“ of ex— 1ange, and it is almost impossible to determine what amount of com— )dity X “should” be exchanged for a given amount of commodity Y :cording to ggy exchange rule. I cannot, for example, empirically :sess all of the exchanges of a couple and——weighting each commodity :cording to its perceived valuation by each spouse-—measure whether at couple is engaged in equal exchange. But I can determine whether spouse perceives the overall marital exchange of inputs to be equal. Since relinquishing power serves to equalize spouses' total puts, the equal-input theories would predict that situations of aqual—resource—input and a corresponding unequal—power—output would viewed as fair by both spouses. Power—dependence theory would differ its prediction; spouses will sometimes pg§_view the exchange as iitable. 72 The indicator used for perception of exchange equality does explicitly tap inputs of external resources (such as income) or 72The question used also appears to tap inputs other than pliance. I first asked respondents: “Some people think that married a is easier on women, while others say it is easier on men. In your experience, does it seem that married life is easier on men or women?" s was followed by asking: "What about your pwp_marriage? (Would you the same thing? Right now, who probably has it a little easier, you rour spouse?)“. 111 arnal resources (of the competence variety); rather it seems to tap ats of internal resources such as amount of time and effort. Be— ;e some exchange commodities are not included in this assessment equality, there is no reason to predict that inequality of effort- ts should result in corresponding inequalities of pewer. Extra ts of effort, in fact, may serve to equalize an otherwise unequal ange of inputs (in much the same manner that compliance is posited qualize inputs in the equal-input theories): aal inputs-—————4> B contributes more effort ——————9 Equal inputs A > B A = B Assuming that this also is consistent with the “norm of equal xnge“ posited by equal-input theories, one would hypothesize that: Test GC—D l: (Equal—input assumption) Spouses perceiving that they contribute (a) more effort and (b) less effort than the other spouse, will not differ in perceptions of the fairness of exchange—outcomes. pouse contributing more effort presumably does so in order to the contributions of the other spouse in other resources, thus lling the norm of equal exchange—inputs. Both should view this iitable and “fair”. The outcomes of the overall exchange (i.e., asulting power structure) should also be viewed as fair. Although lot have measures of ”fairness” pgp_§p3 I do have data on satis— .n with power—outcomes; if these are seen as fair, they should perceived as satisfactory. In like manner, if power-outcomes emerge via a norm of equal ge, there should be no difference in perceived fairness of 112 change results (satisfaction—with—outcomes) between more and less werful spouses. Test GC—D 2: (Equal-input assumption) There will be no difference in perceptions of the fairness of exchange—outcomes between high-power and low-power spouses. oer-dependence theory, on the other hand, would not assume that all uples act according to a norm of equal exchange. I would expect It spouses who contribute more effort (Test D-GC 1) and who also e low power (Test D-GC 2) would have lower satisfaction—with-power- comes than other spouses. In addition, I would expect that high er depresses satisfaction somewhat (as compared with equal—power uses), but not as much as does low—power. This corresponds to ans"(l96l) notion that inequality may bother both parties but it 1 bother the disadvantaged party a great deal more. In like manner, )uld expect spouses who contribute less effort to be somewhat lower 1 equal-effort spouses in satisfaction and spouses who contribute a effort to be much lower. The alternative hypotheses may now be :ed explicitly: Test D—GC l: (Power-dependence assumption) Perceptions of the fairness of exchange outcomes will vary according to perceived contributions of effort of spouses, such that the following rank order will obtain (from highest to lowest satisfaction): 1) Equal effort , 2) Self less effort than spouse 3) Self more effort than spouse Test D—GC 2: (Power-dependence assumption) Perception of the fairness of exchange outcomes will vary according to perceived marital power, such that the following rank order will obtain (from highest to lowest satisfaction): 1) Equal power 2) Self more power than spouse 3) Self less power than spouse 113 :h of the above rest upon the (untested) assumption that satisfaction- h—power-outcomes is a reflection of the perceived fairness or equity those outcomes. Another test between these alternatives theories, which does require this assumption, deals with the notion of “alternatives“ _ :h is so central in power-dependence theory. In a gratitude or com- ance explanation, only the actual (but not the potential, but not 1) resources should affect marital power. No one is ”grateful“ to souse for resources which could be but are not contributed, and no use gains "competence“ from the experience of foregoing resource— inment efforts. In short, if the competence/gratitude theories , I would expect: Test GC-D 3: (Equal-input assumption) There will be no difference in marital power distributions be- tween situations of equal and unequal potential alternative sources of resources. Power—dependence theory, of course, would predict that the se with higher alternatives will have higher power (controlling xlternative sources of resources which gyg_currently being utilized). Test D—GC 3: (Power-dependence assumption) The greater the inequality between spouses in potential alterna— tives, the greater the inequality in marital power (to the advantage of the spouse having higher potential alternatives). tence Versus -Dependence If both the competence and power-dependence theories receive support, an assessment of their relative adequacy would be in order. 114 st, one would want to control for one in order to see if the effect the other is reduced or eliminated. Test C-D l: (Competence assumption) When perceptions of relative competence are held constant, relative dependence will not be related to marital power. Test D—C l: (Power-dependence assumption) When relative dependence of spouses is held constant, relative competence perceptions will not be related to marital power. above hypotheses are stated rather strongly. A reduction in the ation between dependence (or competence) and power, due to such :rols, would also constitute some support for the hypothesis in tion. A related strategy, particularly useful if both of the above theses receive some support, is the assessment of “relative effects“ enberg, 1968: 169—182). Here the inconsistent cases are examined, a high competence and low dependence do not occur in combination. impetence is the more ”potent“ independent variable, then one i expect: Test C-D 2: (Competence assumption) Spouses with higher per— ceived competence combined with high relative dependence will have more marital power than spouses with lower competence combined with low relative dependence. pendence is the stronger influence on marital power, the opposite ts would be expected (Test D—C 2). Also interesting would be an examination of ”task importance” variable. Hypothesis C—3 predicts that competence and marital should be most strongly related under conditions of high perceived 115 aortance of the task (where it is important that the task be done 1). Power-dependence theory might be the better explanation under ditions of low task—importance and competence theory under condi- ns of high task-importance. Test C—D 3: Perceived relative competence will be more strongly related to marital power under conditions of high perceived task—importance, and relative dependence will be more strongly related to marital power under conditions of low perceived task—importance. A different prediction results, however, if task—importance nterpreted as an aspect of motivational investment, with power- :ndence theory. From that point of view, power-dependence theory 1d hold even more strongly under high task importance than under importance conditions (because in the former the power incentive is increased, due to the greater benefits at stake). Test D-C 4: (Power dependence assumption: task importance = motivational investment level) Relative dependence will be more strongly related to marital power under conditions of high task importance (than under conditions of low task-importance). 3 ire this hypothesis with Hypothesis C—3, previously presented. Hypothesis C-3: (Competence assumption: task importance = rationale for subordinating individual goals in the interests of group goals) The relation- ship between perceived competence and marital power will be stronger for “tasks" judged to be important than for those judged to be less important. 73“Dependence” here does not include the effect of task importance. (prior) dependence is presumably increased by the high motivational tment in the task. 116 There are reasons, then, for speculating that each theory ought to iold most strongly under the high-task-importance condition. This :ondition is also a critical test situation from another perspective: :afilios—Rothschild (1970) notes that "winning" in issues which are mportant to one reflects more power than “winning“ on unimportant ssues.‘ Any theory which only accounts for power distributions within nimportant areas cannot be said to have high explanatory power. The heory which receives the strongest support under the high-importance 3ndition could be said to have more explanatory power. This brings up the question of underlying assumptions. Does igh task-importance indicate that the individual has high individual vals at stake (as Safilios—Rothschild may assume and which is assumed rTest 0-6 4) or that there are high gypgp goals at stake (as the mpetence theory seems to assume; less competent B yields to A, crificing short—range individual gains for the sake of achieving oup gains). Compare the predictions in Figure 8. The power-dependence predictions in Figure 8 are based on an terpretation of task—importance judgments as motivational investment.74 74One should ask whether my particular measure of task- iortance tapped the importance of group-goals or individual—goals, ice this gets to the heart of the underlying assumptions. If the oer—dependence predictions in Figure l are not supported, this may due to the importance—indicator measuring motivational investment individual—goals (rather than in group-goals). The hypothesis in question (D—4) was originally written with ference-strength (not task-importance) in mind as the measure of ivational investment. These two indicators may yield different re- ts (e.g., a person may think a task is very important but not care \ALOUILI @UCflUCwn—QOIL9301 \Aatzomr‘ih WUEWU WQEHOU 117 xiiaiaiiiiiiiaiaiiiaiaaiiiiiiiir aaaiaaaaiaiaiaaiaaaaiiiiiaaaaaaaiaiaivvr .LwZOQ Fmpwxma cog: bummed uweovumcq mew use Fw>wF mocmpcoqaw meF .m mczmwd )lllllnlllnlnl‘lll‘u .wmsoam “coucdamv mmm_ u < .cowmcm> mocmucmamuacmzmm “W .mmzoam pcwpmaaoo mLoE u < .covmcm> mocmpm E * Auwmwm; mw cwzoa m_< vcmv vwmwm; m? mucwccmgmw m_m aim memwsponzz uwngo_ my Logoa m.< ucmv vmmww; we mocwvcwqu m_< aaiaaiiaaaiiiiiaiaiiiiaiiaiaiiiaaiiiiiiir.ti bmmmmcomv Lo .awmcmgo oz v one pmmh . nmmmmcocH nxwzoa use wocwvcmamu semapwa,cowpm_wm .aiaiaaiaaaaaaaaaaaaaaaavaiaaaiaiaaaaaaaaaiaaaa Amxmpm pm on on uwsammm pcwsw>mwsoaa_com .firrvr As_ooaaLeCL area acowmewL esp do 2“? -_m:c mommmgocW mwspv wees mmcmo Lospo asp mUCWm .mumex mmzoam mocmpsoaaw azo_ wcp xFanoca mgmw_o:: covpowumcm ummwmc Logoq m_< umcmzo_ Lozoq m_< mucmpcoqaw :w < A m wocwpcoaaw Cw m A <¥ ”mmcmmmwu mmmzomm vwmmmcooo m-o mwmmzpoa>:\\\\ / nwmmmcocH “cmzoa use wocopmasoo cowapwa cowue_wm wocmpcoaaw xmwk cm_ mucapcoaew xmmp ..C "macaw mmmzoqm Fmauw>Fucqv \DKUJO J3 J..I!..II>’. ..).I 118 The more that a spouse views an issue as important, the greater his or her motivational investment in a goal mediated by the other. This in- creases dependence (since more rewards are now mediated by the other spouse) and tends to decrease power. Tests Involving Specific Resources As discussed previously, one cannot hope to subject a theory if marital power to stringent testing by correlating specific resource— ossession with marital power distribution. Still, there are a few pecific “resources“ which intuitively would seem to be “potent" (in heir probable effect on power) and for which competing theories would redict distinctly different power-outcomes. Such tests necessarily ivolve additional assumptions, and so require more caution in the iterpretation of findings, but they may provide fruitful clues for irther theoretical development. ch who does it, or a person may care very much who does a relatively important task). The consequences of poor task performance (at stake the task—performance questions) may differ from the consequences of ing tasks one dislikes or which “cost” much time and work (at stake the preference—strength question). Hence, where task-importance and eference—strength do pg: coincide: High task—importance + low preference strength = group goal achieve- ment is at stake Low task—importance + high preference strength = individual goal achievement is at stake )othesis D—4, if it holds anywhere, should hold under the latter condi- in. Non-support under this condition would constitute rather strong mtive evidence against power-dependence hypotheses about motivational estment. This approach, however, has the disadvantage of eliminating h-importance tasks from the analysis. 119 impetence and Gratitude arsus Power-Dependence An employed wife should have greater marital power than a n-employed wife, according to either the competence or gratitude rmulations. The same would seem to be the case under power—dependence sumptions; after all, she has greater alternative sources of re— urces outside of her marriage. However, the second aspect of )endence--motivationa1 investment——may actually be increased for ne employed wives, cancelling out the marital power which otherwise )Uld be gained from working outside the home. Hoffman's (1960) study of marital power of working wives Hcated that working was related to wives' marital power (a) if es had traditional marital ideologies or (b) if wives strongly re- ted traditional norms. However, if wives were more ambivalent in ir rejection of traditional norms, they garnered l§§§_power from ir working than in the other two cases. Hoffman suggests that / may try to reduce the “threat” of working to their husbands by 19 extra—compliant. Another possibility is that, for some women, 19 able to work is a valued goal in itself. They have high moti- onal investment in being able to work. With the possible ption of couples having extremely equalitarian husbands, the wife's on of working is still a goal ”mediated by the husband”. That is, an effectively prevent his wife from working, either by forbidding r—-in more “liberal” situations--by more subtle means such as 1g her little assistance with home responsibilities, complaining 120 she has less free time to spend with him, or not being willing to wer housecleaning standards now that she has less time for house— :k. In short, he can making working such a “costly" option for a wife that it is no longer worth the investment. He need not ;ually do any of these things. Just the fact that he gpglg prevent ‘from working makes her dependent upon him, if she has high moti— ional investment in being able to work. Her high motivational estment in working increases her dependence, and may override any er she would otherwise obtain from the other ”resources” she gains n this alternative. A working wife who does not value being able to work (or who laps would rather not work) can always implicitly threaten to quit , thereby reducing the family's income, unless her husband is onably cooperative at home.75 Test D-CG 2: (Power—dependence assumption) Among couples. where the wife is employed, the relationship between the wife's relative income and her marital power will be stronger when her moti- vational investment in working is high and weaker when her motivational investment is low. A competence or gratitude formulation, on the other hand, I predict that such motivational investment would not affect the ion of income to power (since motivational investment does not the resource-contributions or her work—related competence nor esources or competence of her husband). 75This idea was suggested to me by Mel Katz, former statistical ltant for the M.S.U. Department of Sociology. 120 he has less free time to spend with him, or not being willing to r housecleaning standards now that she has less time for house- In short, he can making working such a “costly” option for wife that it is no longer worth the investment. He need not rally do any of these things. Just the fact that he could prevent from working makes her dependent upon him, if she has high moti— onal investment in being able to work. Her high motivational estment in working increases her dependence, and may override any er she would otherwise obtain from the other ”resources” she gains 1 this alternative. A working wife who does not value being able to work (or who raps would rather not work) can always implicitly threaten to quit , thereby reducing the family's income, unless her husband is onably cooperative at home.75 Test D-CG 2: (Power—dependence assumption) Among couples. where the wife is employed, the relationship between the wife‘s relative income and her marital power will be stronger when her moti— vational investment in working is high and weaker when her motivational investment is low. A competence or gratitude formulation, on the other hand, 1 predict that such motivational investment would not affect the .ion of income to power (since motivational investment does not the resource—contributions or her work-related competence nor esources or competence of her husband). 75This idea was suggested to me by Mel Katz, former statistical ltant for the M.S.U. Department of Sociology. in COL—"whi; «5.1 {LE ft m .n( .w (. 121 The foregoing is a specific case of a more general type of situation, one in which motivational investment is not directly ”attached“ to resources.76 Where motivation is invested in specific resources (e.g., of the other spouse), the several theories would pre- dict the same results. In the gratitude or competence theories, the value of particular resources (e.g., income or some specific skill) is not taken as a given; although common cultural standards are often assumed, ideosyncratic differences in evaluation are not excluded in principle. “Motivational investment” in resources is simply another term for describing the fact that resources differ in value, and hence in their ability to affect marital power Competence Versus Power—Dependence In the competence explanation, high resources are associated with high competence because resource-procurement activities provide learning experiences. There are several particular resources which, it would seem, ought to be associated with the learning of knowledge 76Another such case is motivational investment in power-as- an—end-in—itself (see Hypothesis 0—6). Hoffman (1960) essentially :ested a hypothesis similar to Test D—CG 2, substituting ”traditionalism" hr motivational investment. I suggest that her traditionalism ideology masure may actually tap intrinsic valuation of power. To make a strong ese for this, of course, I would need a more direct measure of intrin- ic valuation which is strongly associated with her traditionalism scale. would speculate that wives' motivational investment in having a career see Test D—CG 2 above) and both spouses' motivational investments in eing a powerful person, in combination, could account for many of offman's cases of high—resource, low-power wives. -....qmmev- w . :LMM l22 and skills relevant to marriage. Persons possessing the ”resources”77 of greater age, previous marital experience, and having a child from a previous marriage should have greater average competence relevant to many marital situations (although having had children might only increase competence in dealing with child-specific problems). There— fore, such persons should have high average marital power. These same “resources“, from a power—dependence perspective, would be expected to decrease marital power by restricting "alterna— :ives". Particularly for women, I would expect greater age to decrease heir “market value“ on the marriage market, reducing their potential lternatives to their present husbands. The label of “divorced" can lso have the same effect, as can children requiring care and financial upport. Such market-value considerations should be especially salient, would think, for newly married couples such as those in my sample. I The contrasting predictions of competence theory and power- apendence theory for these specific resources are as follows: Test C-D 4: (Competence assumption) Possession of resources of greater age, previous marital experience, and previous experience with children is directly re— lated to marital power, such that the greater the spouses' inequality in experience, the greater their inequality of marital power (to the advantage of the more experienced spouse). W 77The use of the term “resources” seems awkward here. Berger's ncept of "status characteristic“ is less awkward, but is not entirely itable in this particular case. Previous marital experience, for ample, is a resource in that it supplies learning opportunities but often negatively evaluated, i.e., it is a negative rather than a 5itive status characteristic. l23 Test D—C 4: (Power—dependence assumption) Age, having been divorced, and having children from a previous marriage are inversely related to marital power, such that the greater the spouses' inequality, the greater their inequality of marital power (to the advantage of the spouse who is lower on these characteristics). The power—outcome predictions of the above hypotheses may be ummarized as follows: ELATIVE RESOURCE Resource assumed to Resource assumed to )SSESSION OF SPOUSES: increase competence increase dependence l Equal power I Equal power Both high (equal) . A _____ - _ - _ _ _ _ T _ _ _ ~._ _ _ _ _ _ _ One high, one low Unequal power, to the . Unequal power, to the (unequal) advantage of the highjfiTadvantage of the low; resource spouse . resource spouse Both low Equal power I Equal power (equal) ' I .l ”e—9"= where the theories differ in their predictions gure 9. Predicted marital power, according to competence and to power—dependence theory, for the resources of previous marital experience, age, and having children from a pre- vious marriage. this comparison, the crucial difference in predicted power occurs in eunequal condition (one high, one low resource condition). Such a parison would provide support for one or the other theory, although dings could not be generalized beyond the particular resources ted. This concludes my discussion of the hypotheses to be tested in present study. The purpose of this chapter was not merely to l24 ievelop a perhaps arbitrary set of hypotheses. It was a theoretical effort which, I hope, may contribute to the development of marital Iower theory. The present impassé in one dominant tradition, the resource heory program of research, may prove to be due more to unclear heoretical assumptions than to problems of research methodology. I ave focused on the cognitive processes which have been explicitly * implicitly assumed to be the link between resources of spouses Id their marital power. My detailed examination of diverse and, at mes, contrasting theoretical formulations may provide clues about uitful directions for the future study of marital power. Following a description of the research methods employed in study, I will present the results of my hypothesis tests. CHAPTER VIII METHODOLOGY Overview of the Study e Panel Design Data for testing the preceding hypotheses was obtained by terviewing a stratified sample of newlywed Michigan couples in eir first and second marriages. Central to the design employed are two elements: (a) the terviewing of both_spouses in each couple, (b) at two different ints in time during their early months of marriage. Couples were terviewed by telephone shortly after their weddings and, again, ur or five months later in their homes. Figure l0 summarizes the ning of the longitudinal design and the numbers of couples at each age for which complete data were obtained. ELE.‘ 1’? W°§-_ -.5 TIME 1 _ §-§ W°§'_.9 TIME 2 Ilication l/2 hour l-l/2 hour 'marriage TELEPHONE INTERVIEWS PERSONAL INTERVIEWS ense with each spouse with both spouses (separately and together) 584 couples N = l56 couples N = 90 couples Ire lO. Overview of the Panel Design. l25 , - “3.1LW l26 Findings reported are based upon data for the 90 couples (l80 individuals) who participated in all phases of the research.78 Justification of the Design The choice of this design was not made lightly, since it implied high research costs. That is, four individual-interviews were required per couple instead of the one needed for the more common procedure (e.g., that used by Blood and Wolfe, l960) of inter— viewing one spouse at one point in time. Inclusion of both spouses: Bernard (l972) and Safilios— Rothschild (l969), among others, have summarized the evidence that marriage may constitute a different reality for husbands and wives. For example, spouses often give different answers to the same questions. These differences are not random but patterned, e.g., in my own study, husbands reported doing more household tasks and Wives reported making more decisions than their spouses gave them :redit for. Unlike Heer (l962) or Wilkening and Morrison (l963), I am IOt willing to assume that the reports of one gender are more reliable ban the other. If individual's perceptions are assumed to affect ehavior, then one should consider the perceptions of both spouses in rying to explain marital behavior. 78Actually, 94 couples were involved in the entire study. )ur were excluded from the analysis because the wives' personal iterview questionnaires were lost in the mail. 127 The interviewing of both spouses complicates the analysis of data but offers possibilities otherwise not available. In the present study, for example, this made it possible for me to identify areas on which couples' role preferences were in agreement at the beginning of marriage, eliminating these areas from power—scores (since a role- division which is based upon consensus cannot be said to reflect power). In addition, having data from both spouses allowed me to sub— mit theories of marital power to a much more rigorous test than would othenwise have been possible. Take the hypothesis from power—dependence theory that the spouse with the higher “market value” on the marriage market will have the higher marital power. Individuals who perceive themselves as having a low “market value”, by this hypothesis, would not be predicted to have low power unless the other spouse had a high “market value.“ 'Use of both spouses‘ answers allows the separation of theoretically different situations, e.g., where both have a low market value (equal-power prediction). The longitudinal dimension: The study of married couples at nore than one point in time offers the possibility of observing couples )ver time. Even for a study in which the time—span covered is one of mnths rather than years, there are several advantages to such a ongitudinal approach. Perhaps the primary advantages for the present eport are methodological ones. A marital-power measure of the degree 0 which an individual's initial preferences are achieved some months ater must include data about early preferences. Later recall of these vie are sel am am he Kei 9X1 in! 128 ferences are likely to be unreliable, both because of the difficulty remembering and because of the wish to convince oneself or the inter- wer that the marriage is satisfactory, e.g., that responsibilities divided according to one's preferences. Questions about why one ected a particular spouse or about one‘s ”market value” are likely be more reliably answered shortly after marriage, while the memory fresh. Sampling teria for the pling Pool The sampling pool consisted of all couples meeting four pling criteria who applied for marriage licenses between February l May 15, 1974, in seven urban counties in southern Michigan. These nties were Wayne, Oakland, and Macomb (the greater Detroit area), 1am and Eaton (Lansing—East Lansing), Washtenaw (Ann Arbor) and : (Grand Rapids). All marriage license applications during this period were lined and 584 couples identified as apparently meeting the follow- eligibility criteria: a) Both spouses between the ages of 22 and 40; b) Husbands employed in an upper—middle or middle-middle status occupation; 791t did not prove feasible to simultaneously consider wives‘ )ations as well. The requirement that wives also have high-status >ations would have reduced the sample size drastically. For purposes of sample selection, a simple technique of “mining occupational level was needed, one which allowed quick sort— »f the thousands of license applications while sitting in the various The CON 129 c) Both spouses either: 1) entering a first marriage or, 2) entering a second marriage preceded by divorce (rather than by widowhood); d) Both spouses born in the United States or Canada. These eligibility criteria were chosen with two aims in mind. first was to increase the homogeneity of the sample, that is, to trol for variables extraneous to the theory which are known to ect the dependent variables. For example, marital power and ision—task arrangements vary among social classes (e.g., see Blood Wolfe, 1960) and among cultures (Michel, 1967; Safilios—Rothschild, 7; Buric and Zegevié, 1967; Rodman, 1967). If one is to have con— ence that observed variations in marital power and decision—task nty Clerks' offices. The Revised Scale for Rating Occupation rner et al., 1949) was chosen because “ . . . its seven-point occu- ion scale [is] probably the most sophisticated short classification ilable“ (Robinson et al., 1969: 338). Couples were included in ; study only if the husband's occupation fell in the top three of seven categories. However, occupation descriptions on marriage enses are sometimes vague, and a few persons with lower-middle level ipations were included in the sample as a result. Husbands describing themselves as students or military per- el were also excluded, the former because a student situation was idered transitory and atypical, and the latter because no military allations existed within the region where it was financially feasible onduct interviews° 80A few of the 584 couples initially selected, it turned out, widowed and had to be disqualified. Michigan marriage license forms inguish between divorced and widowed persons only for women, by ask— for ”Maiden name (if a widow)“. One county refused to ask women if had been divorced, listing the maiden name of all previously married 1, and several others had grossly inaccurate records on this item. ie other hand, two counties on their own initiative collected divorce ‘mation from both men and women. In all counties it was necessary to upon answers given by the couple on the return-postcard (Appendix A). 130 structures are due to differences in the hypothesized independent variables, rather than in social class or cultural background, such controls are necessary. Therefore, persons born in other countries or working in lower—status occupations (the measure used for social class position) were not included in the study. It would have been desirable to control for several additional extraneous variables via sampling criteria, but this would have reduced the final sample size too much. The selection of predominantly urban counties also served to increase sample homogeneity by eliminating most rural couples. The sample proved to be predominantly white, although race was not a sampling criterion because few non-white husbands met the occupational criterion and because refusal rates were higher than for white couples.8] One of the aims of this research was the study of marital :ommunication (although the present report does not deal with this iata). The second purpose of the sampling criteria was to increase :he probability of finding explicit marital negotiation behavior, some— hing thought to be quite infrequent in the general population. Test— ng of hypotheses is not possible without adequate variation in the 81Refusal rates were particularly high for racially mixed arriages. Black couples were often willing to be interviewed by elephone but declined the offer for a personal interview, which may all have been due to the use of white telephone interviewers. Plans 3 employ a black interviewing team during the personal interviews are dropped because of the low number of black respondents. lhi var in( IN eng 131 independent variable, i.e., one cannot compare the effects of high versus low communication with a sample of low-communication couples. While the first purpose of the sampling criteria was to reduce variance with respect to extraneous variables, the second was to increase variance of the independent variables. The very young and the less educated,82 it was reasoned, are less likely to question traditional marital roles and hence to engage in the extensive communication often required to work out Iew roles. They were excluded from the sampling frame for this eason. Special effort was made to include many previously divorced ersons in the sample. The experience of one unsatisfactory former arriage83 was believed likely to increase marital discussion about ales, in an effort to avoid past problems and mistakes. Also, "evious marital experience may affect both “competence" and lependence“, two central variables to be tested with reference to 82It did not prove feasible to use education pgr g; as a mpling criterion. Education is known to be highly intercorrelated th occupation status (Blau and Duncan, 1967), an indicator which 5 employed as a selection rule. 83The previously widowed were excluded because they presumably :ked this experience. Within this age group, there proved to be very lwidows and widowers anyway. Persons in their third marriage or more were excluded for sons of sample homogeneity. 132 the marital power theories. Since such second marriages were a fairly small proportion84 of the sampling pool, simple random sampling was 10t appropriate. Sample Stratification The sampling pool was stratified into four marriage “types“ [5 illustrated in Figure 11, with approximately equal numbers of :ouples selected from each category for interviewing. 84Statistics for the state of Michigan (Michigan Department f Public Health, 1971: 324) indicate relatively high percentages of econd marriages for persons between 22 and 40. Second marriages as a per cent of all 1969 marriages in Michigan Ages Brides Grooms 20—24 10.8% 5.1% 25-29 41.7 26.8 30-34 58.3 50.0 35-39 62.9 57.4 25—39 49.9 36.9 All ages (including the above) 18.6 19.0 ese figures, however, include many people not eligible for the present udy (e.g., married to someone under 22 or over 40 or to someone having re than one previous marriage). Most important, they include working ass as well as middle class persons. In sorting through thousands of rriage licenses, many more second marriages were encountered among )se with lower-status occupations for every one found among higher- itus levels. In fact, the infrequency of eligible second marriages :essitated the extension of the study to many more counties than 11d otherwise have been included. 133 WIFE lst marriage 2nd marriage lst marriage a b HUSBAND 2nd marriage c d igure 11. Sampling design. The reason for including a high proportion of second marriages as already been discussed. The choice of a four-category sampling esign rather than a two—category one (first vs. second marriages) was ade on both methodological and theoretical grounds. This design insures, first, equal numbers of individuals of xch gender who have been and have not been previously married. Iother technique of achieving this goal of equal numbers, the in— usion of only both—first and both—second marriages85 (cells a and d Figure 11) was rejected because the “mixed” cases (cells b and c) ovide the best tests of certain hypotheses and hence are the most eoretically interesting. For example, it is hypothesized that be- ; divorced is a power—resource in marital exchange which tends to fect one's bargaining power. But it is relative resources which are Jothesized to affect relative power of spouses; no difference in ative power is predicted between both—first and both-second couples 85This design would not have been possible on practical grounds way, as both-second marriages are the least frequent and hardest to ain of all the four marriage types. 134 since resources are equal in both cases. Only where one spouse has a relative advantage is a different outcome expected. A clear rank—ordering of husband—dominance (or wife dominance) among the four cells can be predicted (e.g., in power-dependence theory, husband dominance should be the highest in cell b, medium in cells a and d, and lowest in cell c). This rank—order provides a more stringent test of the hypothesis (in the sense of being more clearly falsifiable) than does the prediction that both-first and both-second couples should not differ in relative dominance patterns (where the problem enters in of judging how much similarity it takes to conclude that there is no difference between groups). Characteristics and Representativeness of the Sample Although theory development rather than generalization of findings to a larger population was the primary objective of the study, a few comments should be made about the representativeness of the sample finally selected. Effect of sgmplinggproceduresiupon representativeness: The :ctual process of obtaining a sample of respondents involved four tages: the choice of counties, within those counties the collection f names of all couples meeting the sampling criteria, selection of ouples for the first interviews and-—from those interviewed—-the ioice of couples to be interviewed a second time. The seven counties were chosen because they offered a large )pulation base, encompassing four of Michigan‘s five most populous 135 Standard Metropolitan Statistical Areas (Detroit, Grand Rapids, Lansing, and Ann Arbor)?6 Flint and the smaller SMSA‘s of Saginaw, (alamazoo, Muskegon-Muskegon Heights, Jackson, and Bay City were not included. No claim is made that study findings can be generalized :0 Michigan as a whole or to urban situations in other states. The second stage of compiling the sampling frame required everal months before sufficient names of eligible newlyweds could e accumulated. Given the requirement that couples be interviewed s soon as possible after the wedding, interviewing necessarily over- apped with the gathering of additional names. This had two conse— Jences for sample representativeness. First, it was impossible to "aw a random sample within each of the four marriage types because re total list of names was not available before interviewing began. 1d second, since the couples married first were interviewed first, 86The combined population of these four SMSA's was 5,351,682 rsons in 1970 (U.S. Bureau of the Census, 1971; 24-3), with Detroit ing by far the largest with 4,199,931 persons in its three—county ea. Lansing and Ann Arbor were chosen in preference to the more Julous Flint area for practical reasons, Lansing because the study ; headquartered there, and Ann Arbor because work could be coordinated :h the bordering Detroit area. The rural fringes contained in the SMSA's of Detroit and 1 Arbor were not included, partly because they fell outside of the l-free Michigan State University telephone trunkline. The pre— finantly rural counties of Clinton (in the Lansing area) and Ottawa | the Grand Rapids SMSA) were excluded entirely. . 87In order to have a complete sampling frame prior to inter- w1ng——when it is necessary to interview people soon after marriage-- ggh larger population base would be needed. 136 interviewing quotas for the most frequently occurring marriage types were filled first. The result was that both—first marriage couples chosen had weddings in February and March, while second marriage 88 There is weddings included April and early May dates in addition. no reason to assume that couples marrying at different times of the year will vary in systematic, theoretically relevant ways,89 however, so this should not be an important source of bias. Couples chosen during the third and fourth stages of sampling (for the first and second interviews respectively) may be more repre- sentative of second marriages (than of both—first marriages). The lack of random sampling presents less of a problem, since nearly all possible second marriage couples were interviewed even when this re— quired extensive efforts to locate and persuade them to participate. This was necessary in order to obtain sufficient cases, particularly of the rarer both-second and wife-second marriages. Both—first couples, on the other hand, were numerous, the sample interviewed contained a high proportion of self-selected couples (i.e., those that were most Milling to participate).90 88During the last visits to County Clerks, only the names of second-marriage couples were gathered. 89There are differences in number of couples who marry dur- ng different seasons, varying from a low point in January to a high n June. The “June bride” phenomenon is probably age—related, including .higher percentage of young couples; these couples most likely are not ithin the age group of this study-~both spouses >22 and <40 years. 90This is more true of some counties than of others, however, epending upon the conscientiousness and persistence of the interviewers. he Kent County sample of first marriages is undoubtedly the best in his respect. In retrospect, it would have been advisable and possible 0 use random sampling of both-first couples for the second interviews. 137 Response rates: Of the initial list of 584 couples, 14 were liscovered not to meet the eligibility criteria for the study (three id not go through with their marriage plans, six had been widowed ather than divorced, and five were disqualified for other reasons). f the remaining couples, 47 were known to have moved out of the cea and 20 had no forwarding address or telephone, leaving a total f 503 couples as potential respondents. Response patterns of these >uples are summarized in Table 2. The 156 couples where both spouses had been interviewed nstituted the sampling pool for the second interviews. Perhaps cause the experience of the first interview had been non-threaten- 9 and/or interesting, over 70 per cent accepted the invitation for other interview before any follow—up calls were made. Many of ese were both-first couples, all of whom could not be interviewed ‘ financial reasons. Of the pool of 156 couples, four had moved out of the a and one was already divorced. Response patterns of the remain— 151 couples are shown in Table 3. Of these 151 couples, 85 per cent agreed to be interviewed 60 per cent actually were interviewed. Per cent of couples arviewed varied considerably by marriage type: both first marriage, )er cent; husband second marriage, 61 per cent; wife second marriage, )er cent; and both second marriage, 76 per cent. The higher rates the several second—marriage types were due to special follow—up phone calls, necessary for obtaining sufficient respondents. m— mm~ mFN mwm ow w¢_ amp NNF Amm_a:oo . mom " zv suwch mwucwpmmuu< nwpumpcou wucmpmwoo< umcosmwaH wucmwmmou< NFwmec:_o> 020 m mJMQMMm Fmpoh prOH meccaw_mh vsmopmom ngmopmom mwcxzwmm 139 la 3. Response patterns by marriage type: Phase II personal interviews. Total Total Total Total Contacted Acceptances Refusals Interviewed h lst 55 ples) 48 46 2 26 band 2nd 36 ples) 35 33 2 22 e 2nd 35 pies) 35 28 7 23 h 2nd 25 ales) 25 22 _;§ 19 \L 151 )1es) 143 129 14 9O acteristics of Respondents Characteristics of the 90 couples who participated in all es of the study are presented in Table 4. These are the couples provided the data used in the present report. 1 The majority of couples interviewed, especially in the d—marriage groups, were from the populous Detroit area. The marriage types--not surprisingly-~differed with respect to al characteristics. On the average, wives were youngest in first marriages, followed by husband-second and wife-second ages, with wives in both-second marriages being the oldest. The trend was true for age of husbands, except that husbands in .ou Zacmpzmwz 11111111111 Aw—uzv amen Lons< cc< 140 v .00 mcxmz mm we Fm mm H v mmcm pvoepmo Aouzv .ou cooam Awnzv .ou EmsmcH x FF & a & m & m_ wnz wmcm mcemcme omHmmstoo Homuzm Amruzw Ammuzm Ammuzm Ammuzm ililllliliilllix Jcmch FGCOWLma HH omega ”mgzp mmmwscms >n mpcwucoqmmm mo mowpmvcopowgmgo .q wreak 141 Tammsox a oo_ . m s mm a oo_ a _o_ llllllll AV *1. II AmUZv mu... [1 o -- ----------------------- s m ---------mfl--------- msz-e30exe3 e_sew-flmmwmwwm--- -------------------------Nmmzs-mmflmmm:aao Lasso - m a mm- mm m N— a? mm ----s em & mm III.I|lln-lll I'll. iiiw-mm-------- N me x mm mm N mv m e . m _N --------m-------------m- mm mm -mm------------mm-- mp Am ---------------------------Aomuzv ma_mm Ammquflammwimqmoxm --iiiiiiiiimmauVLm_u II-Il“|ll -----lemmmmwcaocq .mcmmmcmz mmmszlflmmflmmmWMI.chowmwmwmmm uiiiilllillllililtiiiii1111 m2Hosazm maze: ”oz0422m maze: "masz Nomuzw ~a_uzm Amwuzv NNNHZM mmuz .....iiiiiiiuii. _ sacoc aeN spam sm_ eessmzz om_ ate: .pm_ Loom mQHHmHmmcuFpm>mecoo wmzoqm mco .chmnFF mmzo m mzo "cmez ----------------m NF mm ------x--------i--------------AcoFmFFm; Lasso .UFFoepeu =chowp -eemes. .LWWZaa xoeoepco .pcmpmmpogm Fmpcmenczd om Fa mm Ammuzv =m>Fpm>mezoo= mmmzo m :Fom AcoFmFFwL o: .uFFoggwo :ccmuoez .sszmq xouosceoxco: .pcmpmwpocm FmpszmucsFico: R m? N FM AQMHZV:_.MLwn:._.: wwwzo m Eflom 3:33:51). .1 IIIIll-IIIIIIu-IIIIII-llllllllll & ma 144 . ------------------- meuzv oz s mm & Fm & m N .. -----1--------1--NwmmZM-wmw-ii . mmwmma < zoma QsHIu < m>Hs 2mmostu >z< on Nomnzu NmFuzw Nmmuzv “Nmuzv Nomnzm 4r the clearing up of questions and problems. Out-of—town interviewers were supervised by telephone and I personal visits every couple of weeks, when completed questionnaires are checked and picked up. The major task of supervision proved to : keeping the time schedule implied in the research design (so that e time span between the first and second interviews was similar for 1 couples).96 The Data—Gathering Process taining Access to rriage Records The use of marriage license records as a source of names was sirable, because such records allow immediate identification of JPIES meeting the sampling criteria. This saves the expense of send— 1 thousands of letters to persons who may not be eligible and also 96Because of this design requirement, any problems with apendable interviewers were serious problems. The most difficult as involved people who ceased doing interviewing but would not admit : when telephoned, thus delaying the assignment of new interviewers Lhe area. In retrospect, this should have been handled by having lleted questionnaires mailed in immediately. At the time, however, added cost of several dollars per couple for first—class postage judged to be prohibitive. Third-class mail was tried but proved e much too slow, often taking several weeks. Also, several tionnaires mailed third-class never arrived. 152 *ovides some data about all couples, allowing comparison of :spondents and non-respondents to identify sampling biases.97 It proved necessary for practical (as well as theoretical) asons to interview very recently married persons, since the dresses on marriage licenses are soon out of date. It also proved :essary to obtain recent records from the individual county clerks :her than from the Vital Statistics Department of the Michigan aartment of Public Health, because filing deadlines are seldom Imptly met by the counties. Although, by law, marriage licenses are public records in higan, access, in practice, requires the permission of the County rk. A telephone call explaining the purpose of the research ficed in most cases. In one county, cooperation was attained only "equesting the intervention of Mr. van Hamburg of the Michigan ic Health Service Vital Statistics Office. In general, the ler and busier offices were initially less enthusiastic about the usion of researchers into their work routine. Each of the seven offices was visited several times, a total 9 trips in all, for the purpose of examining the most recent 97The possibility of obtaining names from ministers of “liberal” “-middle class protestant churches was considered, with the hope that would provide a sample homogeneous with respect to religion as well cial class. Unlike more traditional churches, however, such churches do not keep records of marriages performed. 153 98 After sorting out the eligible couples, the rriage licenses. ctinent information about each was copied (names, addresses, ages, :upations, country of birth, license file number, county and date license application and/or marriage). Letters were then addressed . sent to each couple as soon as possible, often before returning Lansing in order to circumvent the two—day postal delay otherwise ountered. tial Contacts with Couples The letter sent to all 584 couples explained the purpose of study and asked their cooperation in completing ten questions on Istage—paid return card (see Appendix A), including their telephone er if they were willing to be interviewed. Other questions in- ed religion, education and immediate plans for having children ended as a possible additional sampling criterion although not in and used as such), whether each spouse had been married before—- if so—-whether divorced or not (to supplement incomplete marriage se data for sample stratification use), items for use in com- 9 the characteristics of respondents and non—respondents (race, t of time each spouse is employed, presence of children living 98Some offices only allowed us to use license applications, others preferred us to use the licenses. The former is pre- e because occupation descriptions are sometimes changed during ping of final licenses (e.g., “executive secretary to president st National Bank” may be changed to “secretary“). All informa— ad to be copied by hand onto our own forms. 154 the couple, nature of the major job activity,99 and whether the fit occupation was a temporary job or the beginning of a long—term 100 If the return-card was not received within two or three weeks, 101 .nd copy of the letter was sent. Cards were finally received 78 couples in this manner. In addition, many couples were con— by telephone when it was possible to obtain their telephone 3. This proved an effective technique of increasing the response 32 Answers to the “postcard questions“ were obtained during the Jart of the telephone calls, a technique which seemed to convince of the legitimacy of the study and reassure them that we did not to ask threatening questions. In all, 148 couples were tele— in this manner. 99”In your present occupation, which ppg_activity takes the ne?“ Respondents chose from categories adapted from Kohn (1969)—— and writing, working with your hands, or dealing with people—— i to provide additional information about the status—level of 'ons. This question proved inappropriate for self—administration lny respondents insisted upon checking more than one of the three es. 100This question, as well as the question about major occupa— ctivity, appeared to have some reliability problems. In several couple was asked the questions by phone before their postcard ived. The two sets of answers often differed, probably because t spouses responded in each case. IOIThe first letters were addressed to “Mr. and Mrs...” and sent lsband's address. The second letters specified the maiden name fe as well and were sent to the wife's address on the marriage Wives appear to be more conscientious about providing the post th forwarding addresses. 02A good deal of the success must be credited to Maureen Aquilina, many evenings on the MSU telephone trunk lines telephoning 155 a I Telephone Interviews Telephone interviewing began as soon as the first acceptances obtained from couples, in early March, and continued until mid- 1974. Each couple was interviewed as soon as possible after wedding, which usually meant one to two months thereafter.103 Each spouse was interviewed separately for half-an—hour by 104 :erviewer of the same gender. Couples were asked not to dis— :he interview questions until both had been interviewed and, possible, both spouses were interviewed within the same week in to minimize inter—spouse discussion. In all, 328 individuals were interviewed including 156 “com— 105 couples. Only these couples were eligible for the second )f interviews. 103By Michigan law a couple may apply for a marriage license 0 days prior to the wedding. In practice, the average is only six days beforehand. Given delays due to forwarding mail to resses and to honeymoon trips, etc., few couples received the until a week or two after their wedding. Nor did they (quite andably) answer it immediately but waited a week or two before ig the card. 104During pretesting, it became apparent that people tell It things to male and female interviewers. Husbands, in parti— eemed less than frank about how they would prefer to divide d tasks when talking with a female interviewer. Hence it was to handle the differential effects of interviewer gender by cross-gender interviewing situations. For telephone solicitation of interview participation, female callers seemed more successful with both husbands and 1051f one spouse declined to participate, the other spouse was ~viewed. Men were much more likely to refuse than their wives, :tages of the interviewing. 156 ase II Personal Interviews At the end of each telephone interview, the possibility of :eCOnd interview was mentioned and a tentative statement of interest :ained. A letter verifying the plans for personal interviews was led to couples in late June, 1974, together with an answer form pendix A) and a stamped return—envelope. The response rate to 5 letter was sufficiently high (close to 70 per cent ppig: to any low-up efforts) that a second letter was not sent this time.106 a telephone calls were made to second-marriage couples, since a Ier percentage was needed in order to meet the requirements of the ling design without too large a sacrifice in sample size. Among more numerous first marriages, more couples volunteered than could nterviewed with the funds available. The objective was to obtain nple of 25 couples in each of the four marriage types or a total '0 couples (See Figure 11). Each couple was assigned to an interviewing team of one man we woman. Interviewers contacted the couple by telephone some veforehand to make a definite interview appointment. Interviews m6This remarkably high response rate was apparently due to st interviews being non-threatening and of interest to respond— The success was such that it suggests the deliberate use of telephone interviews as a recruitment technique for future l—interview projects. 157 k place in the couple's home, usually on evenings and weekends, lasted for approximately 1-1/2 hours.]07 ‘ The bulk of the personal interviews were done between July and September 31, 1974. In order to maintain equal time intervals Ieen the Phase I and Phase II interviews, an attempt was made to srview first those couples that had been interviewed first during telephone interviews. This meant that first marriages tended to ontacted before second marriages. However, summer vacations of les made this scheduling ideal more difficult to attain. The interview format was as follows: the interviewing team led at the couple's home and, after some initial small talk, soon into separate rooms (the wife and female interviewer in one room he husband and male interviewer in another). Separate interviews d an hour and a quarter, on the average. This was followed by ninute joint interview of the couple by both interviewers, for .es of obtaining behavioral measures of couple interaction (not ed in the analysis in the present report). The questionnaires :1uded in Appendices C through E. Ninety of the hoped-for 100 couples were actually interviewed. 5 contained in this report are based upon interviews with these 107Interviewers living in the interviewing area were usually minimize driving time. Even so, since couples were quite cattered within a given area, many hours were spent in driving. thanks are due to Mark and Claire Sandler and to Werner von der drove from Lansing to the Detroit area for the final ”clean-up” IS. 158 Preparation and Analysis of Data Data were first coded onto data coding forms, carefully cked, and then sent to the Michigan State University card punch- and verifying service. In several cases, variables were constructed by combining )rmation from both spouses. Discrepancy scores (e.g., between ises' preferences) were calculated by simple subtraction. For a iled discussion of the calculation of item—specific power scores, Chapter III. Where several questions were grouped together to ide overall measures of marital power, simple averages were calcu- d (the sum of scores of component measures divided by the number tems). Averages rather than sums were used, since the number of )nent items varied from couple to couple, unlike most previous ' indices, only items on which spouses had initially disagreed included in overall power scores. Data analysis was performed on the Telefunken computer at :ibniz—Rechenzentrum of the Bavarian Academy of Science, in , West Germany. The SPSS (Statistical Package for the Social es) program was employed.- Use of this American program package Telefunken computer required some modifications, which proved . source of some difficulty. The program was only recently by Leibniz—Rechenzentrum and the manual describing Telefunken- : control cards contained several errors. In the following chapter, the results of this analysis and interpretation of the results will be presented. It goes with- saying that the small sample size, the theoretically informed non-random sampling design, and the fact that the theories tested not highly refined, precluded the use of multivariate statistical sis procedures. Therefore, fairly simple two-and three—variable tabulations will be presented. CHAPTER IX TESTS OF THE ”EQUAL-INPUT“ THEORIES In the following chapters, hypotheses are discussed in the that they were first presented in the preceding chapters. Indi- construction is dealt with in conjunction with the hypothesis an indicator first appears. Five of the eight ”equal-input“ hypotheses presented in Chap- vere tested for the present report. Four deal with the relation— f perceived competence and marital power and one with gratitude "ital power. Omitted were Hypothesis G, because no direct measure :itude was available, and Hypotheses C—4 and C-5, dealing with nces between spouses' judgments of task importance. Hypothesis C-4, in which task importance is treated as a scope In affecting the relation of competence to power, was not treated arate hypothesis for purposes of this report. Instead, the f task importance was included in the discussion of Hypotheses l. The strategy for dealing with task importance differed some— 1 that implied by Hypothesis C-4; instead of using importance rol variable, the impact of importance was assessed by compar— asults of hypothesis tests using the unweighted marital power :h results when power scores were weighted by perceived task (see Chapter III). Because task importance was thought to ble scope condition for power-dependence hypotheses as well 160 161 3., Test C-D 3), the use of importance-weighted power scores was limited to the competence tests. The majority of all hypothesis :5 in the present report include weighted as well as unweighted er measures. Construction of Marital Power Indices Since the two versions of my marital power index are employed 1e majority of all tables in this report, it is appropriate to 18$ their construction at the outset. lnweighted Marital Power Index _____.________________________ One measure of overall relative marital power is an index , based on eight specific task areas, calculated by the following dure: (a) for each spouse, role—discrepancy scores (the absolute "ence between Time 2 actual task division and Time 1 preferred iivision) were calculated for only those task items on which 35' preferences differed; (b) for each spouse, these discrepancy were averaged; (c) the husband's and wife's average discrepancy were then compared; (d) if they were equal, both spouses received tive power score of 3. If they were unequal, the difference n their average discrepancy scores was figured.. This difference ied to 3 for the more—discrepant spouse and subtracted from 3 for ;s-discrepant spouse. Hence, lower numbers indicate pjgp§£_rela— )WEY‘. Eual - nth ‘3.a index till/e I zero, ' zero) 162 ’ower Index Weighted 'tance second power index, weighted by perceived task importance, .tructed. This index had its base in the relative power :h contested-task (see Figure 5 in Chapter III) which were 1 by the individual's importance-rating of the task in : Figure 6 in Chapter III for weighting coefficients). The ghted relative power scores for the several specific tasks raged for each respondent. As with the unweighted index, which spouses' preferences had differed were included in order that the weighted index be consistent with the un- weighted-index scores were rescored so that lower (i.e., aers indicate higher marital power. Interpretation of Tables the tables which follow, results using weighted and un- indices are placed side-by—side, wherever possible. alues of these two indices are not comparable, e.g., 108 3.00 for the unweighted index and 0.00 for the The indices differ in another important way. For the 'or purposes of index construction, respondents having power on a specific task were given a score of 0.1 O). (Other possible scores were 4, 3, 2, l, -l, -2, his slight change did not have much effect on the final allowed equal—power scores to vary according to per- ce. Otherwise, all equal-power scores would have been 5 of importance (since 0.0 times any weight still equals UHWE a SC a SC one VEIS isp hypo sex whet the . supp. .2]: << / DErC‘ 163 ndex, marital power is ”zero sum” in nature; if a wife has 1.10 (10 points from equality). then her husband must have 3.90. For the importance—weighted index, this is not true; lredict the husband's exact score from the wife's or vice- 1 tables in Chapters IX and X are set up such that power to be the lowest on the left—hand side and the highest on nd side. ’To put it another way, a table which supports the ill have the highest means on the left and the lowest mean e right. the right-hand side of each table are comments about ot the data suppOrt the hypothesis. These comments describe 1 0f the trend but not the magnitude of the support or nOn- a notation ”curvilinear” is used for cases where the trend )ected direction between two adjacent ”conditions” but in direction for the third “condition.” Some relationships “curvilinear” may still provide some support for the f the deviation-from—linearity is very slight. h this introduction, let us proceed to the first set of sts--those dealing with perceived competence and marital Tests of Perceived Competence Hypotheses General Competence + Overall Marital Power rding to the competence formulation, the greater the etence of a spouse, the greater that spouse's marital powe Hypo task stre clos sion 1101i seem 164 be (and the lower the power of the other spouse). deals with general or overall competence, rather than with c skills and abilities. Since the competence theory role of competence perceptions in the yielding of power competent spouse, it seemed appropriate to use the ptpg: luation as an indicator. That is, one should be able to power from A's perception of B's competence: the higher f B's competence, the more often A should yield power to fore-—the lower A's power should be relative to that of B. 1r indicators of perceived overall competence dealt with ance and dependability of the spouse. Intelligence comes 10d and Wolfe's notion of overall ability to handle deci- lblems while dependability—-although mentioned by Blood and the explanation of the low power of black husbands-—would 1y relevant to my task—based power measures. question used to measure intelligence and dependability as administered during the Time 1 interview. A number of iing dependability and intelligence) were mentioned in the be-selection criteria and the respondent was asked about vw important was this to you, in choosing him/her rather lse?” Response categories varied from extremely important important. Pretest trials had shown that respondents were directly asked to ”rate” their mate but would willingly stion when it was posed in terms of mate-selection criteria. atest respondents were asked both versions of the question; 'ere very similar in all cases with respect to intelligence and answ term tent same simu atio expe DONE in l. 165 idability. Hence there is some basis for assuming that people this question in terms of evaluating the mate, rather than in their own mate-selection criteria. This test of hypothesis C is a P—centric test, i.e., compe- icators for only one spouse are used. A dyadic test of the thesis, employing competence perceptions of both spouses ausly, will follow (see Hypothesis C—l). Table 5 deals with the relation of P—centric competence evalu— overall marital power. According to Hypothesis C, one would pouse's overall relative marital power to be inversely related her evaluation of the other spouse's competence. That is, ld be lowest in condition (1) and highest in condition (3), Jsing the unweighted power index (see left side of the table), '.wixgs are consistent with the hypothesis for both types of Support is stronger for wives' intelligence ratings than ependability ratings of the husband. When task importance to account (see right side of table), the hypothesis is ily for intelligence but not for dependability. Wives who lusbands as relatively undependable (condition 3) apparently 'n way only in areas they consider to be unimportant. sults for husbands show strong support for Hypothesis C to dependability, when the unweighted power index is used, itionship does not hold up when task importance is taken Husbands who View their wives as dependable apparently s' preferences, but only in areas that the husbands 1.656 LwnEzc LGZqu .00. u Lmzon Fmavm .szoa szmF; muooFucF m z zthzz Ammnzv “mohzv Ame . zwa>,3 ma> M No.- No. FF. . zopuzv Ameuzv szuzz zoo . zvamsx . .1 moan AeomcFFF>Lzov oz mo.- Fo - mF wcoFuuocFo “cmugooEF unopLouEF “:ougooEF ”we *szoa quFLoE com: eaouFoaLa.aeu FF“ on ooz zFoEoLFXm :F vcwcp mF Lo umzszom Ame ANV z_v AgmmcFFF>L=uv oz AcmmcFFF>L=uV oz «:oFuquFo ooooFuata ago cm UCmLH mm pcmucomEF ucoyxomEF ucmugomeF ”Fo acwzo :oFLouFLo :anomme-oume m we mucoucanF oucmmFFqucF .m mucouoosoo zouzv Aownzv zemuzv zao No. NF. mo. szuzv Amenzv szuzv Ago oo.- ao.- m_.- FFm pm poz szEmLuxm Amv ANV AFV :oFmeFLo :oFuowFomuwuoe m we mucougo EF NpF_ano=amoo Lo gossanm H< mocaoaasoo wzFFImez wuzqhmosz :FHz u zz mm>F3 u zv mocoamoz mmzm ,AcoweFFF>coov oz FconomcFo , omquvaa mcu :F ucng mF mmF mm» mcoFuowLFo . cF uchF mF oaooFooLa we“ __m on ooz an: Lozoza zoo Fozom .Lozoo Lozan mumoFucF moon .oo.m u to m u zv m AeFuzv Ammnzz Ammnzv AN ma>e3 . «m.m mo.m om.m u zv zoFuzv Ameuzv z_mnzv zoo moeanmsx u so so as . “concomEF peopLomEF peopeomEF ”Co acozo FFo go “oz zFoEoLuXm .co pozzweom . sz ANV AFV :oFLouFLo :onomFomiouoe o mo oocmucooEF wucmmFFqucH “m wocmpoosoo 1111 zanzv Aswuzz Asmnzz znm . zv oF.m FF.m mF.m mw>Fz Am_uzz Amsuzv sznzv Aka . 2V Fm.N FF.N Fo.m muconmsz “:oucoqu ucowgoaam. “copgozaw ”to acozo FouFLoE coo: szEoLFxm Lo umzszom Ame ANV sz :oFLochu :oFuooFom-ouoE m we mucouco EF quFFnaocwmma ”z wocwuonou mz3mersmm sss ”s n...:.._.: 167 r to be unimportant. The hypothesis was not supported for gence ratings, with either the weighted or unweighted power In summary, Hypothesis C receives stronger support for wives ' husbands. Both spouses yield power to a "dependable" spouse, in relatively unimportant issues. Wives, but not husbands, war to an ”intelligent” spouse. This may indicate that wives the wishes of "competent” husbands but that husbands—-when ld—-do so for reasons other than competence. Or it might indi— t wives view "intelligence" as a competence that is relevant irticular task areas in question, while husbands do not view :ks as requiring any particular intelligence. s C-1: Dyadic General—Competence + Overall Marital Power Hypothesis C-l specifies that the relation between perceived e and marital power should be stronger when spouses ggpgg in :eptions of relative competence. I cannot provide a direct 11$ hypothesis because each respondent rated only the other : not his or her own competence. A related test, combining data from both spouses, however, e with the data available. Essentially it is a dyadic version sis C, with the independent variable consisting of combina— usband's and wife's evaluations of each other's competence. the specific dimensions of competence measured are dependa- intelligence. 168 Empirical Hypothesis C-l: Overall relative marital power will be related to perceived competence, such that the following rank order will obtain (from highest to lowest power of spouse A): l) A's rating of B's competence < B's rating of A 2) A's rating of B's competence = B's rating of A 3) A's rating of B's competence > B's rating of A contains the results of this hypothesis test. Note that the 1 group (1) are married to the husbands in group (3) and vice- Because of the dyadic nature of the indicators involved, pp; I power scores (see left side of table) of husbands and wives independent of each other. If the wives in group (1) have a ove 3.00, then their husbands (in group 3) must have a score 00. In short, the trend for husbands and for wives must be Hypothesis C—l receives pp support for the competence of pugs, in contrast to Hypothesis C for which wives—-considered are likely to yield to husbands they considered as very intelli— . appears that such yielding wives fall disproportionately into in Table 6; their husbands also rate them as very intelligent, usbands apparently do not consider their wives' intelligence Jnds for accepting their ideas about task division. Results for dependability ratings are more in line with the ;. Spouses with a dependability—advantage have higher power zen the unweighted index is employed. When task importance nto account (see right side of table), wives but not husbands 1.659 .00. n Lozoo Fozom .Lozon Loszz wumoFucF mswnszc Lmzoza .oo.m u szoo Foaom .Lozoa Lozmwz wumoFucF mem552: cwxoza zomuzz zomuzv Ammuzz zoo u zv Azwuzv zomuzv Ammuzz AzomcFFF>Lzuv oz no.1 oF.- co. moconmzz “LowcFFF>L=oV oz mm.~ mw.N mo.m Amwuzz zomnzv zoanv zmo u zv Ammhzv zomuzv zzmnzv Aeooc—__>ooov oz Fo.- Nz. oo. mo>Fz zeoozzzz>ozov oz zo.N zz.m zo.m .coFuuoLFo nilmmmmlel. mmflmmm Illwmmmmwoll "mo «zozo FouFLoE cow: mcoFuoozFo omzomm mQFwom illwmmmmmll "mo acmzo ooooopoozz :ozF ocFooz osom :ozF mcFooz ooooFooco :ozF oszoz oeom :ozF ozFuoz z, ocwzF mo Lozon cFom zoom Loews cFom oF ozoeF oz Lozon eFom zoom L2.3 tFom . zmv zmv Fv _ “my Amy AFV monomFFqucF ”m wocoumasou mocmeFFoucH um mucouooEou Ammuzv zomuzv Amzhzz zzo . zv Amonzv zomnzo zNFqu zoo u zv zeooo_zz>coov oz mo.- NF.- wo.- oooooooz mo» oo.~ mo.N mo.~ mocmomoz . zzznzz zzmnzv Ammnzv zoo u zz zozuzv zzmuzv Ammuzv zeooo__e>czoz oz W zo. no. mo. wo>Fz moz oo.N eo.m om.m FconomLFo omzomm mcFuoz FoqumE coo: MmmwwmmmMMml lilmmmmmmWIi wmflmmm ulimwmmmmnlt woquuozm cozF mcFuom wEmm :ozF mcFuom H owquowLa cozF mcFuoz mEom :ozF mcFuom :F newcF mH Losz: mem zoom zozoz wFum cF oeoeF «F cozoF: FFom zoom zozoz *Fom Amv ANV A: , Amv ANV A: zoFFFoooeomoa u< oucouoosou NpFFFgmocwmwo n< mucoumoeou muzzeooZH zm oszmez mz2mHmummm >m mwzoo 4FFo HFlu mmeIFoa>I 170 fit with the hypothesis. As the wives' dependability-advantage es, their power in important areas also increases (although the is not large in magnitude). To summarize, Hypothesis C-l receives support only for the nce of dependence but not for intelligence. This support is t stronger for wives than for husbands. ;is C—2: Task—specific Power + Use of "Competence" as a Rationale Perhaps a more appropriate way to measure competence is to ask Iirectly about task—relevant abilities. After reporting how bilities for eight specific tasks were allocated, respondents ed why they did it that way. Two of the thirteen possible hoices involved considerations of competence: “The person who's t this should do it” and “It's faster or more efficient this aspondents who gave at least one of these responses, among their ‘ee answers, were separated from the respondents who gave neither e response. Task-specific relative marital pOwer was calculated by: (a) ng each individual's role—discrepancy score on each “contested“ (the absolute difference between the preferred and actual :ion scores); (b) comparing the wife's and husband's role dis- for individual tasks. Equally discrepant spouses both task-power score of 4.00. Where spouses were unequal, the ifference between spouses' role—discrepancy scores was calcu- a more—discrepant spouse received a score of 4.00 plus the score, while the less-discrepant spouse was given a score 171 minus the difference score. This procedure is illustrated in 5 in Chapter III. The spouse who came the closer to achieving her preferences was considered to be the more powerful of the Hypothesis C—2 specifies that lower-power spouses should men- npetence considerations with equal or greater frequency than having greater power in that specific area. That is, according :ompetence formulation, the less-competent spouse yields to the )f the other because of recognition of the other's greater abil- Results for Hypothesis C—2 can be found in Table 7. Two :as are eliminated from this analysis, ”deciding what people to :o the house” because few couples disagreed about allocation of em and “doing the supper dishes,“ because few respondents con— competence as relevant to this task. Hypothesis C—2 is pp: supported in any of the twelve sub—tests 7. In slightly more than half of these, results are in the direction from that predicted, i.e., the mgpe powerful a spouse task area, the more likely that he or she would give ”compe— : a rationale. This pattern occurs for both husbands and wives .raditionally “masculine” and ”feminine” task areas.109 109Although not related to the hypothesis, the distribution by task area exhibits an interesting pattern. Examination of [s of high- and low-power husbands and wives within each task Table 7) reveals that husbands predominate in all of the ally “masculine” tasks (tasks 2, 3, and 6), as well as in one ally ”feminine" task (task 4). For the remaining ”feminine” and 7), husbands and wives are equally represented in the high- lp. In no task area do wives predominate more often than husbands. 171 ) minus the difference score. This procedure is illustrated in 5 in Chapter III. The spouse who came the closer to achieving her preferences was considered to be the more powerful of the Hypothesis C—2 specifies that lower—power spouses should men- npetence considerations with equal or greater frequency than having greater power in that specific area. That is, according :ompetence formulation, the less—competent spouse yields to the )f the other because of recognition of the other's greater abil— Results for Hypothesis C—2 can be found in Table 7. Two eas are eliminated from this analysis, "deciding what people to To the house” because few couples disagreed about allocation of m and lldoing the supper dishes,“ because few respondents con— competence as relevant to this task. Hypothesis C-2 is pp: supported in any of the twelve sub—tests 7. In slightly more than half of these, results are in the direction from that predicted, i.e., the mgpg powerful a spouse task area, the more likely that he or she would give ”compe- a rationale. This pattern occurs for both husbands and wives raditionally ”masculine“ and ”feminine“ task areas.109 109Although not related to the hypothesis, the distribution 3y task area exhibits an interesting pattern. Examination of :s of high- and low-power husbands and wives within each task Table 7) reveals that husbands predominate in all of the tlly ”masculine” tasks (tasks 2, 3, and 6), as well as in one llly ”feminine" task (task 4). For the remaining “feminine” and 7), husbands and wives are equally represented in the high- Ip. In no task area do wives predominate more often than husbands. HYPOTHESIS C-2: 172 Table 7 PER CENT 0F HUSBANDS AND WIVES WITHIN EACH LEVEL OF TASK- SPECIFIC RELATIVE MARITAL POWER, WHO CITED HE WAY COMPETENCE AS A RATIONALE FOR CONTESTED— TASKS WERE DIVIDED 7 * ' Task-Specific Marital Power* 11) 12) 13) Higher Lower 15 Trend in once given as a rationale Than Than Predicted a division of: Spouse Egual Spouse Direction? 1 icery shopping ( ) ( ) ( ) : ( = N=15 N=13 N=15 No opposite Husbands (N 43) 53% 54% 47% . direction) . _ (N=15) (N=13) (N=14) 'No (opposite , ””95 ("'42) 67% age; 437. ' direction) )airing things around the house ( ) ( ) : _ N=16) (N=21 N=B .No (opposite ”“Sba'lds (”‘45) 75% 67% 63% . direction) . _ (N=8) (N=20) (N=17) ' No ””95 ("'45) 75% 65% agar. ‘ (curvilinear) ing complaints to salesmen or 1 d or . _ (N=25) (N=10) (N=l7) . No (oppos1te ““Sbands (”‘52) 64% 50% 41% 1 direction) . _ (N=17) (N=10) (N=24) ' No (opposite Hives (N—51) 77% 59% 5131: ' direction) Ding in touch with relatives (N=23) (N=14) (N=16) : o ””Sba"d5 (”=531 30% 43% 251. ; (curvilinear) . _ (N=16) (N=14) (N=23) 1 No (opposite “"85 ("'53) 31% 21% 2g% 1 direction) ding who you will invite to house (omitted --- --- --- ng care of the car (N=19) (N=16) ”1:10) 1 No Husbands (N=45) 63% 56% 60% I (curvilinear) . _ (N=9) (N=15) (N=19) ' No 1ves (N743) 67% 47% 58% l (curvilinear) "ing supper (N 14) (N 13) (N 13) ‘ N _ = = = o '“Sbands (“'40) 64% 69% 54% f (curvilinear) . _ (N=13) (N=13) (N=l2) " No (opposite "V95 (“‘38) 46% 31% 33% 1 direction) the supper dishes (omitted) nts within each power level. ae) er measure is not weighted by importance. (The total number is indicated in parentheses above each per- If N=15 and the percentage is 53, then 53 per cent of the respondents cited competence as 1 . Percentage figures are based upon the number of who did not differ in preferences about who should do a given task are excluded. since no are could be calculated for such persons. 173 It appears that Blood and Wolfe were wrong in their assumption lower—power spouses yield power because they recognize the greater tence of the other spouse, at least with respect to contested deci- about who should do what tasks. If the less-powerful spouses are the less-competent ones, they certainly do not appear to recognize The greater use of the ”competence“ rationale by more-powerful as could well reflect a rationalization for their own power, power apparently has some base other than competence.HO The position that competence rationales are used as justifi— 15 is bolstered when one examines the percentages of husbands and who give this rationale (regardless of their relative power).“1 ' ""“‘““““" HOI also studied the association of the same two variables, ime using power as the dependent variable (not shown). Spouses ve a competence-rationale for the division of contested tasks gher mean power than spouses who did not give such a rationale, cases except for two: wives for the task of repairing things the house, and husbands for the task of car care. Interpreta- f such data is problematic, however, since the competence- 1les given did not specify which spouse was thought to be the impetent. ; l l H10f the couples who had disagreed about who they would to do particular tasks, the following percentages of husbands es cited competence as a rationale for the actual task-division at. Husbands Wives Which Per cent Per cent gender citing citing , mentions competence Rank competence Rank. competence EA: rationale order rationale order more often? itionally 11e tasks” epairing (N=45 couples) 69 % l 73 % l Wives .king complaints (N=52 couples) 54 % 4 65 % 2 Wives r care (N=45 couples) 6O % 3 56 % 4 Husbands 174 ves cite competence reasons more frequently than do husbands for two the three "masculine” tasks, while husbands cite competence more en than the wives for two of the three “feminine“ tasks. Such fig— 5 suggest that husbands are resisting participation in ”feminine” ks and wives are resisting participation in “masculine" tasks, and t both are justifying this by claiming that the other is better lified to do these jobs. | In summary, Hypothesis C-2 is definitely not supported. Since Ls of the two previous competence hypotheses provided rather mixed ults, it appears that perceived competence has some effect upon mari- power but--at least as measured in the present study--such compe- :e considerations provide only a fairly small part of the explanation the distribution of marital power. Other theories, such as the itude formulation which will be discussed next, may prove to have explanatory power. Husbands Wives . -—-—--—-—-—-—-— --—-*--—---—— Wh1ch Per cent Per cent gender citing citing mentions competence Rank competence Rank competence rationale order rationale order more often? _.—.—.._______—__—__._..__ 'female tasks" Grocery shopping (N= 43 couples) 51 % 5 57 % 3 Wives Keeping in touch with relatives (N= 53 couples) 32 % 6 25 % 6 Husbands Preparing supper (N= 40 couples) 63 % 2 ‘37 % 5 Husbands 175 Test of a Gratitude Hypothesis 'pothesis G—l: Status Mobility + Overall Marital Power The ”gratitude" formulation cannot be directly tested because direct measures of gratitude or respect were available. One indirect st can be made, however. It seems reasonable to posit that gratitude one's spouse should be greater if one came from a lower-status family an the spouse and less if one's family of orientation was of higher ltUS than the spouse. This is the “Cinderella syndrome“; Cinderella luld be more grateful to the Prince than would be a princess married a prince. And, because of this greater gratitude, Cinderella should—— ording to the gratitude hypothesis--have lower relative marital power n the princess. While none of my respondents were princes or prin— ses, it seems that the upwardly mobile—-those who ”married up” rela- a to their parents' status--should have lower marital power than :e who were downwardly mobile. In the present study, I divided respondents into three groups rding to status “resources“:' (a) those whose spouses had higher— 112 us occupations than their own fathers; (b) those whose spouses 112Occupations were assigned to one of the following cate— as, listed from highest to lowest status: (a) professional, techni- and kindred; (b) managers, officials and proprietors, except farm; :lerical, sales, and kindred; (d) craftsmen, foremen, and kindred; 1peratives and kindred; (f) service workers, including private ehold; (g) laborers, except farm and mine; (h) housewife, etc. 1rs were excluded from the analysis. Those few respondents not ently employed were given an occupational rating based on marriage- se information about their usual occupation or their university (if a student) wherever possible. 176 d fathers had equal-status occupations, and (c) those whose own fathers d higher-status occupations than their spouses. I expected marital er to be the highest for group (c)—-who would have the least reason be grateful to the spouse--and the lowest for group (a). Results of '5 test are presented in Table 8. This hypothesis is supported for husbands. Although the rela— nships for wives are not linear, the hypothesis receives some support that the downwardly mobile wives (group 3) have higher mean power n the upwardly mobile wives (group 1). This is true for both types power indices. This support of Hypothesis G—l must, of course, be interpreted 1 caution. Since the presumed intervening variable of gratitude is measured, the findings are open to other interpretations. I will return to the competence and gratitude formulations ihapter XI, dealing with tests between different theories. First, ver, I will present the findings for my tests of power—dependence ry. 177 Table 8 HYPOTHESIS G-l: OVERALL RELATIVE MARITAL POWER, BY OCCUPATIONAL STATUS OF SPOUSE RELATIVE TO THAT OF FATHER. NWEIGHTED POWER MEASURE zwer numbers indicate higher power. Equal power = 3.00. Relative Status of " i *‘s Spouse and “ ' "s Father (1) (2) (3) Respondent's Respondent's Spouse And 15 Trend in Spouse Respondent's Respondent's Predicted aan marital power* of: Higher Father Egual Father Higher Direction? 1 Wives 3.31 2.89 3.03 1 No . . (N = 83) (N=41) (N=27) (N=15) . (“WWW—‘1'”) I l Husbands 3.00 2.76 2.62 1 Yes (N = 86) (N=40) (N=13) (N=33) ; l 1 ER MEASURE WEIGHTED BY IMPORTANCE marital ower* of: Wives .04 .08 —.Ol er numbers indicate higher power. Equal power = .00. l 1 N° . . (N = 83) (n=41) (N=27) (N=15) 1 (curv111near) 1 Husbands -.02 -.06 -.16 1 Yes (N = 86) (N=40) (N=13) (N=33) 1 1 l l CHAPTER X TESTS OF POWER-DEPENDENCE THEORY Hypothesis tests relating to power—dependence theory will be esented in the order that hypotheses were first introduced in Chapter The first “dimension” of dependence to be considered is that of tential alternatives—to-the—marriage, followed by the second aspect dependence, that of motivational investment in goals mediated by the ruse. Both of these are dealt with in Part I, below. Part I. Alternatives and Motivational Investment othesis D-l: Alternatives + Overall Marital Power According to power-dependence theory, a spouse's marital power ld be directly related to his or her alternative sources of valued -objects outside of the marriage. Such "alternatives” may be either itial or actually utilized. Currently utilized alternatives (e.g., life having a job which provides income and prestige independent e husband) are subject to multiple interpretations. For example, ployed wives were found to have higher mean power than non-employed . one still would not know why this occurs. A gratitude or com— :e explanation would be just as plausible as a power-dependence Potential alternatives, however, are predicted to affect ‘ power in power-dependence theory but not in gratitude or T78 I79 competence formulations (since a spouse cannot be grateful for resources not actually received and since one cannot gain competence from fore- going experiences of resource-attainment). Hence potential—but-unused 'alternatives“ are theoretically the more interesting for the present :tudy. Two indicators of potential alternatives were selected. Both elate to the explicit power-dependence assumption (see Heer, l963: l38) hat the possibility of divorce enters into people's ”calculations.” Jouses who believe themselves to have a good chance to find a suitable ate, should the present marriage dissolve, are less dependent upon the ‘esent spouse. Since dependence and marital power are hypothesized to einversely related, such spouses should have higher power than those ose potential alternatives to the present marriage are less promising. This assumption will be tested from both a P-centric and adic approach (the latter, which includes a broader range of indica- ‘s, is discussed under Hypothesis D—2). Indicators for the P-centric t were the following questions: ”Some people say that women have an easier time finding a good spouse and others say that men have it easier. In your own experience, does it seem that men or women have the easier time (finding a good spouse)?” Following a question about number of people ever dated more than just casually: ”Of these, about how many would you say were very close or serious relationships (including your present spouse)?” questions were judged to tap prospects of a person on the “marriage 1t.” The first--although phrased in terms of men and women in al—-was assumed to tap the respondent's perceptions of his or her ial situation. Since these couples were newlyweds, this memory 180 lOUId have been reasonably fresh. The second question about the Fstory of serious dating relationships can be assumed to indicate 1arket value” as well, i.e., a person with much successful past expe— ence with the opposite sex probably would succeed in future encounters d would find it easy to remarry, if so desired. Hypothesis D—l predicts that a spouse's marital power should higher under high-alternative conditions (condition 3 in Table 9) an when potential alternatives—to-the—spouse are low (condition l). Support for Hypothesis D—l is rather mixed. For Question A, 'es fit the predictions. However, the magnitude of the relation :omes extremely small when task importance is taken into consideration e the lower portion of Table 8). Husbands' data do not_support othesis D-l. With respect to Question B, it is husbands rather than wives support the hypothesis. This support disappears, though, when task— Irtance is used to weight the power scores. Apparently, husbands many serious dating experiences prevail only in task areas which consider to be relatively unimportant. In summary, wives appear to gain some power when they perceive women have an easier time finding a good spouse than men and hus- appear to gain some power when they have extensive dating expe— a.T]3 Wives with extensive dating experience, on the other hand, power--a reminder of the persistence of the double standard. H3Question B, on dating experience, could also be interpreted experience which develops competence. The predicted trends for would be the same if a competence assumption were substituted. .oo. u Luzon szcm _ .Lmzca memw: wpmuwuzm mgwnssc Lb304¥ .oo.m u Lmzon pmzcm wzo .1 .szoa Lwcmwc mums-PUP: menEJE L. 1.— _ n V Akw u 2v 1 uz Amw z . Amwuzv Smnzv GE: E u E H “MM“: AMMNV mos €5.53: 2.22.533 oz _ 8.- 2.- 8.- eggs: I a; _ V cm 5 :3: E - 5 _ 1 u Am_uz n _ :75 $3: 3?: :w 5 $232.; ” . a: OWN 3:3 Agmwcw—?>L=uv oz _ mo. FF. P0,- wm>p3 wuwmoaaov oz _ mv m 23325 0.35 .6 m v .8 m N .8 F W. 1mg 39: .8 m empowuaaa Amv “NV AFV uwuu_uwca Amv cw szcp mH . :? ucwxe mH u m m LgmcoLpMFme «mo—o Lw> +0 s mass: .m.:osumm=o a to ome—o Lw> to Langsz "m :owpmwaa ‘ 1; _ W _ 1 ”2 ANN u zv mm . 3N“: 3.3: :75 E u 5 m _ am”: :3; 2:. v sags: remarriage oz. 3.- 2.- 8.- 2522: . 285:238 oz H R.~ 3 N 8 N _ 1 u . 5",; :5qu 32: :m u E ,_ __ :75 Gmuzw a3: 5 zwai we, . 8.- No. 8. 0:3 , mm; . 8a m; MN m .2 LI. II- ! 1! ~ 1/ I) [I I! .630 SEE :3: «cowpuwgwo wammm wEmw wck memmm "do aszo _mqumE :mwz . m:o_uomcwa memmm wsmm och swwmmm to a F . umuuwnwga xwm :30 mwxww :uom xwm _ kuuwvwca xwm :zo wwxwm cuom xmm cw vcwch mH cw newsk mH muzm omkzuwmz mx3m~pm «M30; 4o mz2mr m wrnmh T82 efore examining the implications of these findings further, let us roceed to Hypothesis D—2. Here the same two indicators of “alterna- ives“ (among other indicators) are considered, this time for the couple ; a unit. Perhaps the pattern for relative alternatives of spouses ll provide a clearer picture. pothesis D-2: Relative Alternatives + Overall Marital Power Since ”power“ in power-dependence theory is an explicitly lational concept, it is relative dependence which ”determines” rela- ve marital power. Alternatives, as one dimension of dependence, )uld exhibit the clearest relationship to marital power when measured 1 relationships). In addition to Questions A and B (see the previous hypothesis), other questions will be included in testing. The wording of these 1dically (e.g., differences between spouses in number of serious dat- ( ;tions and the manner of scoring are as follows: ”People decide to get married in different ways. For yourself, when did it occur to you that you might like to (re)marry-- before you met the ”right” man/woman, or didn't it even occur to you until after you met him/her? Scoring: (l) Self before, spouse after (2 Both before, or both after (3) Self after, spouse before During the last few years, did you ever seriously consider not arrying (again)—-remaining single by choice? Scoring: (l; Self no, spouse yes 2 Both no, or both yes (3) Self yes, spouse no two questions relate to the alternative of not marrying. If being is an acceptable or perhaps even positively evaluated state, t provides a potential alternative to the present marriage. The 183 s attractive is the single state, the more dependent one is upon 1 spouse for maintaining the marriage status. However, if bgth uses value the status of being married highly (i.e., strongly wanted get married, as evidenced by deciding to marry even before a spouse found and never seriously considering remaining single by choice), n both are highly dependent. Dependence, when equal, should not d to inequality of marital power. Only when dependence is unequal uld a corresponding inequality of power be found. Questions A and B, about potential remarriage chances, were 0 rescored dyadically, as indicated below: Do men or women have the easier time in finding a good spouse? Scoring: (l) Both spouses agree that respondent's gender has a more difficult time (2) Mixed answers, or both spouses agree that there is no difference between the genders (3) Both spouses agree that respondent's gender has the easier time Number of very serious dating relationships. Scoring: (l) Self lesser number than spouse (2) Self and spouse equal (3) Self greater number than spouse In Table 10 below, results for these four questions are pre- 1ed. Predictions are for power to be lowest in condition (l) and est in condition (3), i.e., for mean scores to decrease as one 5 toward the right side of the table. ‘ In discussing the test results, I will begin with Questions 1 B (see the previous discussion of these indicators under Hypoth— D-l). Both deal with relative potential alternatives for spouses nd a suitable partner, in the event of divorce. Question A, when d dyadically, receives moderately strong support with the unweighted ~ . Asmmcwrw>gouv_ oz. . Aemmcw~T>L=ov_ oz. . Ncovpomcvo. owuozooeo cw ocmcp mH _ _ hoowuoocvo _ opwmoooov oz. _ . mw> _ 15341 Nvmuowowgm :ozuomcwo cw oomok mH AFm-zv ANmnzv flew-2o flaw - 2 so.- ~_.- so.- mucmamsz AeN-zv z_muzv ANm-zv how u z o_. _o.- mo. mm>e3 om20mm _o:mm mmoomm “do asozo Pouwcme :moz w>oz :oEoz Lo cos oo .mocweco xm :30 :H n< onmezo ”m onemmso zo_nzv “No-2v Aruzo AME u 2v 50.. mo.1 .11 moomomox AF-zv zwo-zv zo_-zv Am” u zv 1-- No. co. mm>wz w Loocow mLoch< Looomw ”we *Lw2o Fouwcme Ewoz,W m.ucoo:oommm Loguo m_u:mocoomwm so; Lormmu Loo Loose: ”mocmo mwmooom Homomm momooom U Amv ”NV A_V W momoo m ooo o Cruz?» we?» Lowmmo ozp w moco>vo wo ucm>w on“ or .Locueoo oroouwom o meets?» Low meucouoo .H poms. muzm omkszmz mx3msouv oz Acmw:w_w>couv oz ‘1‘ meowooaavo oaoowooca :? vowsk mH mw> ~:o_oowewo omoo_uaea cw vouch mz _ A_muzv ANmuzv Aemuzo ham 1 zv . oe.~ No.m mm.~ moeaomzx . . flew-2V A_muzv Ammuzv flaw u zv ” om.m Nm.N NN.m ma>w3 wmoomm Fmomm Ems-r smumocw e_om Ame ANV mop or m Tgmoowpo—oc ovumo ome—o co moowcom Lw> we smoEoz um onHmmzo _ _ Ao_nzv ANo-zv A_uzv ”ms - zv ” mm.~ mm.N --- moeanmsz U “Fuzz zwonzv Ao_uzv _ -11 ~—.m _w.m .1IIIIIIIIIIIII 1|IIII1IIIIII|1 Loooww moozmc< Lovemw m_u:mo:oommm Locus m_u:ov:oomwm Loo Lowmom cod Lovemx “mocmo mmmooow "wocmm mmmooom “My ANV A_V :wocmw wavu Lowmoo wou m>oz coEoz co :wE oo .wocmwco xo :zo cH H< zoCmmso ooLo>Lo mo poo>o on“ ow .LocuLmQ mFompwom m m:_o:ww Low Fmvucouom .o Seas mm3mHH~k<4mz >m mmzom 4Hho ”N19 mHmmIHom>z J.E35 .00. u Lozoo szam .Lozoo Lozmz; mumowocw memoEoc Lozoza .oo.m u Luzon _oocm .Lozoo Lozmzz muaowocw mcooaoo Lozoza Acmwcw—v>gouv oz mmz «:ozooaawo empowvwca cw vcwcp mm Aoowcz—T>L:uv oz Agmmchw>L=uV oz ~=owwuwgro owuowowxo cw noose mH . _ - - " zem u zv . ALF-zz zoeuzz zemuzv zso u zz zaaw=r~w>a=ov ” zNz-zv Ame-2V zom.zv mucanmsx ” o_.- _o.- w~.- mocaamsz oz _ me.N om.m mm N . _ - u so - 2v . sznzz zoo-2v ALF-2v zsm - zv Atma=z__>esuV . qu-zz zoo-2V za_.zz z mm>_3 . mo. mo. mo. mo>_3 oz _ mN.m om.~ _m m . _ _ o: mmoomm mEmm mm» wmoomm "do agozoo quzems cow: mcowuowgwo o: wmoomm wEom mm» omooam ”mo acmzoo _muwcoe coo: .mwz ozom one :pom .oc m_om oouowooco .mw» w_wm one zoom o: Lme Amy zwz z_v Cw ecaah mo Amy ANV z_v wmo_o:u o o— :_m no zoohmmoo ooz:_msoa amowmcoo z_m=oetmm so» u_o no zoopmmoo . . _ AMNHZV zoeuzz zoo-2o zoo - zv ” zMN-zv zoo-2o zo_.zz zmm - 2v H mo.- ~_.- mo. venous: wa> _ om.~ No.m sm.m moeaomsz _ . are some sN-zv so - e U 375 $75 eN-zo so - 5 _ . «0.. oo. oo. mwszz no» _ oh.~ N~.m oe.m _ mmzomm wEwh wEmm wcowuuogwo mmsomm meh wEmm wmoomw each as» zoom owoo_omea gaze use zoom seek Louoo :_ ococp ma Lopez Lo_rcom czam c_am ezwm zmv ANV a; vs ozu ome oo Loumm L LmE w; ca :0 on Lauuo uw Uwv :wzz "u zozhmmao umowcgms ucwmweo oz» on w>wumzowu~m meucwuoo a mo m_mcwm mowconom .HH you; zmv Amz zzv "u zozhmmzo ommzoeme “common wzu oo w>_oocawuzm -_o:ouoo m we w_m:wm mcwc_msoz .zz Hzr the hypothesis. The persistence of the “double standard“ is still 1mewhat in evidence, though; wives who exceed their husbands in dating ;perience have the lowest power of all wives. Another possible explana— on is that this group of wives contains disproportionate numbers of der women and women with children. Both age and having children from previous marriage may serve to reduce alternatives—to-the—present~ “riage, cancelling out whatever power advantage would otherwise have :rued from their dating experience. The effect of age and children 1 be examined in Test D-C 4, Chapter XI. Part II of Table l0 deals with remaining single as a poten- 1 alternative to the present marriage. Results for Question C ongly support Hypothesis D-2 when unweighted power measures are used. introduction of importance-weightings reduces the support, although ier—alternative spouses (condition 3) still have considerably more 2r than lower-alternative spouses (condition l). Question D, a more—direct question about whether singleness Ierceived as a viable potential alternative to marriage, provides ‘ limited support for the hypothesis. Although the equal-alternative ses (condition 2) do not fit the predicted trend, higher-alternative ses (condition 3) do have higher power than lower—alternative ses (condition l), as predicted. When the weighted power index is :ituted for the unweighted one, husbands provide no support at all 187 for the hypothesis. For wives, the trend is now totally consistent with the hypothesis. In summary, the indicators used provide mixed support for Hypothesis D—2. The hypothesis is supported for one of the two ques— tions about potential for finding a suitable alternative partner, in the event of divorce, that of relative ease of finding a good spouse (Question A). The support is stronger for wives than husbands, however, in that the relationship for husbands is reversed when the importance— veighted power index is used. Of the two indicators dealing with ‘emaining single as a potential alternative to marriage, the hypothesis eceives more support for Question C, about relative timing of the deci- ion to marry. The data provide some support for the hypothesis even hen the weighted power measures are used, although the trend is no onger linear; wives ”gain“ in power only when they have a relative ' dvantage (condition 3), while husbands “gain“ when they have either 1 equal or a relative advantage (conditions 2 or 3). Ipothesis D—3: Motivational Investment + Overall and Task—specific Marital Power Hypotheses D—3 and D-4 deal with that aspect of dependence ich is termed motivational investment. The higher one's motivational vestment in goal—objects mediated by the other spouse, the more aendent one is--according to power-dependence theory—~the less power a will have, relative to the spouse. Hypothesis D-3 considers moti- :ional investment levels of individual spouses, with relative moti- ional investment of couples to be treated in Hypothesis D-4 which lows. 188 Two types of indicators were used for motivational investment. 'he first are aimed at tapping overall motivational investment in the iarriage, while the second are specific to the eight task areas. VI first asked respondents: A) “Some people think that married life is easier on women, while others say it is easier on men. In your 9wn_experience, does it seem that married life is easier on men or women?“ his question was asked during the second interview, after the couple as been married for some months. That question was followed by an- ther one: B) ”What about your gwn marriage? (Would you say the same thing? Right now, who probably has it a little easier, you or your spouse?) lswer categories were the same for both questions: ‘ (l) Easier on own gender (2) Both the same (3) Easier on other gender Question A is assumed to measure the individual's CL level hibaut and Kelley, l960), the comparison point against which they aluated their own marriage. A wife who thinks that women in general Tally have it easier in marriage, and who also thinks that she has easier in her own marriage, presumably finds that her marriage is tut the same as others in this respect. If, however, she thinks that in general usually have it easier but—-in her own marriage--she has easier, then presumably her marriage is more beneficial than most. latter wife could be said to have a higher motivational investment ier marriage than the former wife. l89 This point may be clarified by presenting the scoring rocedure which I used. I considered a respondent's answers to bgth uestions A and B, as illustrated below for male respondents. Scoring of Questions A and B in combination: (l) Husband has it easier than do "most husbands,” e.g.: own marriage——husband easier most marriages--husband harder gr_both spouses the same or own marriage--both spouses the same most marr1ages—-husband easier (2) Husband has it e uall eas (or hard as do ”most husbands,” e.g.: own marriage--husband has it easier {most marriages—~husband has it easier or own marriage——both spouses the same other marriages——both spouses the same (3) Husband has it harder than do "most husbands,” e.g.: own marriage—-husband harder most marriages-—husband easier gr both spouses the same OY‘ own marriage—-both spouses the same most marriages—-husband has it harder hort, a spouse (in this example, a husband) is assigned to category whenever his own marriage compares favorably to marriages in gene- as he sees them. Such hUsbands are assumed to value their marriages ly, i.e., to have a high motivational investment in preserving such lvantageous relationship. They therefore should be more dependent eir wives and lower in marital power, as compared to husbands in cry (3) who have less to lose, should the wife leave. IIIIIIIIIIIIIIIIIlIIlIIIIIII-----‘--------------“-‘*--‘—--=I=s”' l9O In addition to the indicator just described, Table ll also ludes responses to Question B alone. As can be seen in Table ll, this hypothesis is not supported either husbands' answers or for wives' answers. Without exception, )ondents who said that they had it easier than their spouses (group lave higher mean power scores than those who say they had it harder. could be interpreted in two ways. First, it may be that the hypoth— simply does not apply to marriage situations. Or it may be that indicators used are not actually tapping motivational investment; 3 these questions were asked after several months of marriage, I i suggest that “having the easier time" in marriage may be a ggnse; a; of having higher relative power. Because I believe that these :ators did not succeed in tapping motivational investment, I have ded them from the following hypothesis about relative motivational‘ tment. The findings in Table ll may also be contrary to Blood and 's assumption that more—powerful spouses ”earn“ their power by ‘buting to satisfactions of the spouse. 'At least, the more- 'ied spouses--e.g., those in group (1) who perceive that they are more advantageous position—-do not show any signs of yielding The second type of motivational investment indicators were c to the eight task-areas. A particular division of labor, I ‘ Id, could be a goal—object in itself. If a wife cares very much ho does the dishes--e.g., has very strong preferences that this shared equally—-she has a strong motivational investment in ].9JI .00. n Lmzoa Fanam .szoo smcmwz oumorocz mLoazzo Lmzoza .oo.m u Lozoo szom .Lozoo szmzz ouoowo:_ mLooaoc Lozoza _ _ zm-zv zomuzz AMN-zv sz " zv _ 1 u Amw u zv “:ozooozwo . . m N . zonzz zoo-zv ANN.zv mossz oo_oooooz oz _ oN.m o_.m zz m o >.z oz . mo. No. No . _ zmz-zv zoo 1 zz W 3-5 of: E-s so - oz EeEzoe ” as: so»: w some; zooooooooo . No- oo. o_.- mooooooz oz " MN.N NN N mo N muzmoooov oz _ .1 1 . o szoo . . . Houzeoa cmoz ” on o zone :moz .cozuumeo Looomu “mo: m< concow % a . Ncozuumcwo Lmocmw wwwmzwmwm hhmwmw mo ago: z u. omuu_owco oEom we oEmm wam Msomcwo oopowoozz mamm we pmoz ooze oz vouch mH “mo: cosh u oz so or vows» mH pmoz : cw Lmzmmu zzwm swoon: w—wm Lmzmom wam same: o E . E E E E zooooom oz zozocmm co mo ozzsoe :pwz omen Eoo mm mauvxsme czo "m + . d monHmmso . zzmuzz zomnzv zNzuzz zoo u zz fizooom__>coov oz H No. op. zo. mo>w3 zooooooooo . ANN-2V zomnzz zomuzv zzo u zz mooooooov oz. No.- oo.- oz.- mooooooz Noozoooeoo ozom zo osom azz N_om co "co .Lozo vaumomgz Lwogm: wmsoom oo< wammm oz oooez mo tzom zoo zNV zzz mwmao m L30 Lo :0 :o cooooo oezz oozoooe mo no zoozmmzo muzm amzzwmmz mm3mz~ zHhoz mo 4m>mz >m mwzom zHHo ”m1o mHmmzzoazz FF wzomz l92 iat goal. Help with the dishes is of great value to her, and she is werefore quite dependent upon her husband; he determines whether or at she can achieve this valued objective. This interpretation leads 3 a prediction which differs from common-sense reasoning, namely that 1e greater one's motivational investment in a particular division of abor, the lower one's power is likely to be. The measure of motivational investment for each task-area 1s a Phase I question about preference strength. After respondents are asked how they would prefer to divide the eight tasks in their own 1rriage, they were then told: ”People often feel more strongly about some jobs than about others. For example, a person might have very strggg pref- erences or opinions about whg_should do the dishes but have no opinion at all about whg should keep in touch with rela- tives.” a eight tasks were then repeated, and for each task the question was ;ed: ”How strong are your preferences (opinions) about wh9_should task name) ? Would you say Very Strong, Strong, Somewhat Strong, or Not at all Strong?” For this hypothesis, task—specific measures of relative mari- power were used (rather than an overall power score). The unweighted ‘cator was constructed by: (a) calculating for each spouse a dis- ancy score (the absolute difference between Time l preferred sion—of—labor and Time 2 report of actual division) and; (b) com- Ig discrepancy scores of husband and wife for a given task. If es had equal discrepancies, each received a score of 4. If dis- ncies were unequal, the difference between discrepancies of spouses Ilculated; the more discrepant spouse received a score of [4 + the |IIIIIIIIIIIIIIII33T___________________________________—__"77 77 """ I93 ce—score], and the less discrepant spouse was given a score of difference—score]. This scoring procedure is illustrated in in Chapter III.“4 Unweighted task—specific marital power, like unweighted over- tal power, has two characteristics. If one spouse is high, r spouse must be low. And the higher the score value, the e marital power. In addition, no power score was calculated les who agreed in their preferences. Hence numbers of respon- Table 12 are lower than 90. The weighted power measure was constructed as illustrated in , Chapter III.”4 Respondents were first assigned a relative )re for each task on which they had disagreed. Equal power ue of 0.l. The highest possible relative power was —4 and - t possible power was +4. Power scores were then multiplied hting factor based on the respondent's answer to the following irdless of who does it, do you think this is an important >r decision?” egories and the corresponding weighting factors were: Very Important .20 Fairly Important .l7 Slightly Important .13 Not at all Important .lO | ‘4However, the direction of scoring used is opposite that pter III. I assigned lower scores to higher—power spouses, make this indicator consistent with other power measures. T94 'ect of this weighting was to increase a respondent's power ge or disadvantage for ”important” tasks and to decrease it for hich were perceived as less important. The assumption behind ocedure is that prevailing in important areas represents more Jower than prevailing in unimportant areas. Weighted task-specific power scores, like the weighted overall 1dex which is based upon them, do n9: assume that one spouse se” power in direct proportion to the other's power “gains.” d who prevails in an area which he considers to be important credited with high power. His wife, if she considers the task be unimportant, will “lose” much less power than her husband Findings for both weighted and unweighted power are presented l2. Results (see Table l2) are very mixed. Using the unweighted sure, husbands' data support the hypothesis for tasks l, 6, l of which are traditionally ”female” tasks. Wives correspond ictions for tasks 2 and 8; here, however, the only large power 3 occurs between ”somewhat strong” and ”not at all strong“ 15. Wives with very weak preference strength about who should per dishes and repair things around the house are the most achieve the kind of division of labor they prefer. dith both unweighted and weighted power measures, an inter- 115 tern emerges for several tasks. Power is low when 15For unweighted power measures, the tasks which follow this -, for husbands, tasks 2, 3, and 4 and, for wives, tasks 3 lZ. HYPOTHESIS D-3: TASK-SPECIFIC RELATIVE MARITAL POWER BY LEVEL OF MOTIVATIONAL INVESTMENT IN TASK-DIVISION OUTCOME. TED POWER MEASURE Strength of Preferences About Who Does This Task (I) (2) (3) Is Trend in rital task—specific Strong or Somewhat Not At All Predicted accor in to task: Very Strong Strong Strong Direction? cery Shopping 5 usbands (N=43) 4.25 3.8l 3.79 : Yes (N=8) (N=l6) (N=l9) . ives (N=42) 4.00 4.06 4.l2 : No (opposite (111:3) (N=l7) (N=l7) - direction) airing things around the house 1 lusbands (N=44) 3.83 3.55 4.00 1 No (curvilinear) (N=23) (11:11) (N=lO) ' l lives (N=46) 4.40 4.40 3.94 1 Yes (N=l5) (N=l5) (N=16) 1 ng complaints, if necessary, to smen or landlord 1 I Jsbands (N=Sl) 3.6 3.04 4.27 1 No (curvilinear) (N=13) (N=23) (N=15) ; ives (N=52) 4.72 (4.23) (4.33) ' No (curvilinear) (N=lB) N=22 11:12 .' ng in touch with relatives : sbands (11:53) 4.2; 3.60) 3.79 , No (curvilinear) (11:9 (N=25 (11:19) - ves (N=53) 3.95 4.30 4.43 1 No (opposite (N=l9) (N=20) (N=l4) 1 direction) ing who you will invite to 1use (omitted) --— --— ——- | care of the car - bands (N=45) 3.8] 3.47 3.43 : Yes (N=2l) (N=l7) (N=7) _ es (N=44) 4.29 4.43 4.33 :No (curvilinear) (N=l4) (N=2l) (N=9) . ng supper 1 >ands (N=39) 3.86 4.l8 3.57 1' No (curvilinear) (N=l4) (N=ll) (N=l4) s (N=4D) 4.33 3.77 4.27 ' No (curvilinear) (N=12) (N=13) (N=l5) 1 1e supper dishes 1 inds (N=48) 3.78 3.69 3.57 I Yes (N=9) (N=l6) (N=23) 1 - (N=48) 4.59 4.50 4.10 ; Yes (N=l7) (N=lO) (N=21) . s indicate higher power. Equal power = 4.00. Table l2. Continued. 1.9 6 OWER MEASURE WEIGHTED BY IMPORTANCE Strength of Preferences About Who Does This Task . . . (l) (2) (3) Is Trend in can mar1tal task-spec1f1c Strong or Somewhat Hot At All Predicted pwer* according to task: Very Strong Strong Strong Direction? Grocery Shopping l I Husbands (N=43) .02 -.O3 -.08 1 Yes (N=8) (N=16) (N=l9) Wives (N—42) -.Ol .00 .Ol ‘ N0 (opposite (N=8) (N=l7) (N=l7) j direction) Repairing things around the house : Husbands (N=44) {-.03) (-.07 -.01 : No (curvilinear) N=23 N=ll) (N=l0) , Wives (N=46) .05 .07 -.02 ' No (curvilinear) (N=l5) (N=15) (N=l6) ' 1 Making complaints, if necessary, to 1 salesmen or land lo rd Husbands (N=5l) -.05 —.lS -l.l7 ' Yes (N=l3) (N=23) (N=15) 1 I Wives (N=52) .lO .03 .05 , No (curvilinear) (N=l8) (N=22) (N=l2) . Keeping in touch with relatives ‘ Husbands (N=53) .03 —.06 -.04 : No (curvilinear) (N=9) (N=25) (N=l9) , Wives (N=53) -.02 .05 .07 1 No (curvilinear) (N=l9) (N—zo) (N—14) )ecidi1ng who you will invite to the house (omitted) --— --- --— l 'aking care of the car Husbands (N=45) -.o4> -1.13) (-.00 : No (curvilinear) (N=2l (N=l7 N=7 Wives (N=44) .04) ( .07) ( .0? 1 No (curvilinear) (N=l4 N=Zl N=9 , ' | reparing supper 1 Husbands (N=39) -.04 .83 —.07 ' No (curvilinear) (N=l4) (N=ll) (N=l4) I Wives (N=40) .05 -.O4 .03 : No (curvilinear) (N-12) (N~l3) (N»l5) ‘ I ing the supper dishes , Husbands (N=48) —.02 —.05 -.06 1 Yes (N=9) (N=l6) (N=23) ) Wives (N=48) .10 .05 .24 I No (curvilinear) (N=l7) (N=lO) (N=21) 1 I scores indicate higher power. Equal power = .00. 197 rences are strong and higher when preferences are somewhat strong, edicted. However, power decreases again in the weak-preference Thus it may be that very strong preferences do indicate a moti- nal investment level which is sufficiently high to increase dence (lower power). But the effect of very weak preferences otivational investment) upon power is apparently offset by other 5; perhaps people with very weak preferences make no effort to nce their spouses or even to mention what their task—division ences are. At any rate, it is difficult to identify any clear overall ns in the data on Table l2. Perhaps some of the confusion can rified when relative preference strengths of spouses are studied. ill be attempted under Hypothesis D-4 below. asis D—4: Relative Motivational Investment + Overall and Task-specific Marital Power Relative motivational investment should be inversely related tive marital power, according to Hypothesis D-4. That is, the having the higher motivational investment should have the lower power within the couple. As for Hypothesis D-3, in which When weighted power scores are employed, the tasks are tasks d 6 for husbands and tasks 3, 7, and 8 for wives. I can find rn as to which tasks exhibit this trend and which do not. Such e about equally divided between traditionally “male" and jobs and between tasks about which more men than women have references and vice-versa. I98 es were considered separately, the present hypothesis was tested both general and task—specific indicators of motivational invest— Two general indicators were used. The first dealt with a ion about motivational investment in marriage itself (a goal which viously mediated by the other spouse). The question was: ‘In general, it is safe to say that married people get more )ut of life than people who never get married." dents were asked to indicate whether they strongly agreed, agreed, eed, or strongly disagreed with this statement. In order to a measure of relative motivational investment, responses of hus- d wife were combined as indicated below. Scoring: (l) Respondent agreed more strongly than spouse Respondent and spouse agreed (or disagreed) equally (3) Respondent agreed less strongly than spouse relative marital power is predicted to be the highest for group the lowest for group (l). Results of this test are in Table l3. e rates for this question are quite low. However, the available early do not support the hypothesis. Motivational investment is I, rather than inversely, related to marital power. A second general indicator of relative motivational invest— : a summary measure of discrepancy between spouses' task-specific 1ce strengths. Discussion of this will be postponed until after tion of preference-strength discrepancies for separate task i In the previous hypothesis, task-specific preference strengths 'duals was discussed. The rather mixed results suggested the 1599 Table I3 HYPOTHESIS D-4: OVERALL RELATIVE MARITAL POWER BY RELATIVE MOTIVATIONAL INVESTMENT IN BEING MARRIED. TED POWER MEASURE Responses of both spouses to the statement: Married people get more out of lif 1 (2) (3) Self Agrees Spouses Self Agrees Is Trend in More Strongly Agree Less Strongly Predicted .rital power* of: Than Spouse Egually Than Spouse Direction? Wives 3.07 3.14 3.24 : No (opposite (N=67) (N=15) (N=29) (N=23) ‘ d1rect10n) I iusbands 2.67 2.86 2.93 I No (opposite (N=67) (N=23) (N=2w (N=l5) : d'VECt‘0") lumbers indicate higher power. Equal power = 3.00. ASURE WEIGHTED BY IMPORTANCE ital power* of: ‘ves .Ol .lO .04 1 No (curvilinear) (N=68) (N=l5) (N=29) (N=23) 1 sbands —.06 —.03 —.04 | No (curvilinear) (N=67) (N=23) (N=29) (N=l5) : I nbers indicate higher power. Equal power = .00. 200 of dyadic measures. Relative task-specific preference strength es were calculated by comparing husband with wife. Scoring was as ows: (l) Respondent stronger preference than spouse (2) Respondent and spouse equal in preference strength (3) Respondent weaker preference than spouse 1 all tables, the group predicted to have the highest power is the with the highest number (group 3 in this case). See Table l4 for indings about relative preference strength and task-specific power. This table reveals no clear pattern for the unweighted power res. Except for tasks 6 and 7, the data are not consistent with Ipothesis. In general, it appears that motivational investment :k—division is not a power-disadvantage (except for car care and 'ing supper, where the spouse who felt more strongly is lg§§_likely ieve his or her preferred task-division). 0n the other hand, the —sense alternative prediction that the spouse with stronger pref- ; will get his or her way is not supported either (except for 1 . With the exception of wives on task 7, the importance—weighted leasures are not related to relative motivational investment in dicted direction for any task. Nor is there any consistent among the so-called ”curvilinear” relationships, except that s attain their highest power when preferences of spouses are in tasks 3, 4, and 6). In general, I can make no generalizations about Table l4 :0 say that the data do not fit the hypothesis. Several explana- ght be offered for this. With respect to importance—weighted 201 18 14. HYPOTHESIS D-4: TASK- SPECIFIC RELATIVE MARITAL POWER BY RELATIVE MOTIVATIONAL INVESTMENT 0F SPOUSES IN TASK- DIVISI ON EIGHTED POWER MEASURE Relative Strength of Preferences About Who Does Task (1) (2) (3 Self Stronger Self Weaker Is Trend in 1 task-specific Preferences Preferences Predicted pr* by task area: Than Spouse Both Egual Than Spouse Direction? I Grocery Shopping : Wives (N=42) 4.21 3.93 4.08 'No (curvilinear) (N=l4) (N=15) (N=13) .' Husbands (N=43) ($.08) (4; O7) ( . 'No (curvilinear) = 3 11- 15 N=15) 1 Zepairing things around the house . Wives (N=45) 4.08 4.20 4.26 :No (opposite (N=12) (N=10) (N=23) 1 direction) Husbands (N=45) 3.74 3.80 3.92 :No (opposite (N=23) (N=10) (N=12) 1 direction) aking complaints, if necessary. 0 salesmen or landlord 1 Wives (N=52) ( . ) (4.78) (4.15) ' No (curvilinear) N=21 N=18 N=13 Husbands (N=52) 3.85 3.22 3.71 1 No (curvilinear) (N=13) (N=18) (N=21) ' eping in touch with relatives : Wives (N=53) 4.0 4.53 4.13 1 No (curvilinear) (N=Z3) (N=15) (N=15) 1 Husbands (N=53) 3.87 3.47 3.91 1 No (curvilinear) (N=15) (N=15) (N=23) , riding who you will invite to 1 house (omitted) --- --- --- ing care of the car Wives (N=43) 4.55 .55 4.1 1 Yes (N=ll) (N=l6) (N=16) ' iusbands (N=43) 3.88 .44 3.45 1 Yes (N=16) (N=16) (N=ll) 1 varing supper : lives (N=39) 4.46 4.25 3.89 . Yes (N=13) (N=8) (N=18) 1 I usbands (N=39) 4.11 3.75 3.54 1 Yes (N=18) (N=8) (N=13) g the supper dishes 1 ives (N=48) 4.79 3.72 4.64 ‘ No (curvilinear) (11:19) (N=18) (N=ll) " ‘sbands (N=48) 3.36 .28 3.21 - No (curvilinear) (N=ll) (N=18) (N=l9) ' ores indicate higher power. Equal power = 4.00. 202 Table 14. Continued. POWER MEASURE WEIGHTED BY IMPORTANCE Relative Strength of Preferences About Who Does Task (l) (2) (3) Self Stronger Self Weaker Is Trend in Mean task-specific Preferences Preferences Predicted power* by task area: Than Spouse Both Equal Than Spouse Direction? 1 1) Grocery Shopping 1 l Wives (N=42) ( .02) (—.02) .01) 1 No (curvilinear) N=l4 N=15 (N=13 Husbands (N=43) (-.02) ( .01) —.ll : No (curvilinear) N=13 N=15 (N=15) 1 2) Repairing things around the house 1 Wives (N=45 .00 .Ol .04 : No (opposite (N=12) (N=10) (N=23) 1 direction) Husbands (N=45) -.05 -.03 -.02 : No (opposite (N=23) (N=10) (N=12) 1 direction) 3) Making complaints, if necessary 1 to salesmen or landlor ' Wives (N=52) .03 .10 .03 : No (curvilinear) (N=21) (N=18) (N=13) . Husbands (N=52) .02 -l.10) (~.07) : No (curvilinear) (N=13) (N=lB N=21 , 4) Keeping in touch with relatives 1 Wives (N=53) -.00 .10 .02 I No (curvilinear) (N=23) (N=15) (N=15) Husbands (N=53) -.02 -.08 -.02 : No (curvilinear) (N=15) (N=15) (N=23) , 5) Deciding who you will invite to the house (omitted) -—— --- --— 6) Taking care of the car Wives (N=43) .07 .08 .04 1 No (curvilinear) (N=ll) (N=16) (N=16) 1 Husbands (N=43) .03 ' —1.20 —.08 I No (curvilinear) (N=16) (N=16) (N=ll) 1 7) Preparing supper Wives (N=39) .07 .02 -.02 1 Yes (N=13) (N=8) (N=18) 1 Husbands (N=39) .00) (1.031 (-.07) i No (curvilinear) (N=18 N=8 N=l3 , 8) Doing the supper dishes 1 Wives (N=48) .12 .24 .07 I No (curvilinear) (11:19) (N=18) (11:11) 1 Husbands (N=48) -.09 .03 -.10 1 No (curvilinear) (N=ll) (N=18) (N=l9) ‘ *Lower scores indicate higher power, Equal power = .00. 203 measures (in both Table 14 and in Table 11), it may be that rence-strength and importance ratings are not independent of each Even though these two ratings were obtained during different liews, reducing the chance that respondents would strive to give :tent answers to the two questions, respondents may have inter— ! these questions as having similar content. Before further ch is done, the relationship between preference-strength and ance should be checked. This explanation does not account for the inconsistent find— ith respect to the unweighted power measures, however. Nor do :a for weighted measures indicate that spouses with strong pref- : consistently give high importance ratings; if that were the 1ower scores should have been highest in condition (1). Another possibility is that respondents are most likely to high preference—strength for tasks about which they are the most ent, i.e., for tasks where the spouse-—even a month after the -—has exhibited strong determination not to give in or compromise. this were the case for a large number of respondents, power :hould have been the lowest for condition (1). No such clear is found. In conclusion, Table 14 does not support the hypothesis. have been the result of a combination of the above-mentioned ions. Or it may be that preference—strength about task divi— iot a sufficiently ”powerful" kind of motivational investment, 1t most spouses do not place high value upon attaining their 204 116 Finally, the explanation may simply be sired division of labor. it the hypothesis that motivational investment is inversely related relative power does not hold for married couples. The final table for Hypothesis D-4 utilizes average preference- ength scores as the indicator for motivational investment. This sure was constructed by comparing the two spouses' preference-strength res on each task item. If identical, each spouse received a score 3 for the task. If preference-strength differed, the absolute dif- ence between the wife and husband was computed. This difference was ad to 3 for the stronger—preference spouse and subtracted from 3 for weaker—preference spouse. These relative task-specific preference- ngth scores were then averaged for each respondent across all tasks hich spouses had different role preferences. The resulting averages then dichotomized for purpOses of easier display in Table 15. Although Hypothesis D-4 was not supported for separate task , it does receive support when all task areas are considered her. This support remains strong only for wives when the weighted measure is used. When it comes to ”important” task areas, hus- apparently get their own way if they feel strongly enough about ile wives get their own way only when they do pp§_have strongly 1references. H6Previously married women have higher average preference- th scores than other women, while previously married men do not t this pattern. It seems reasonable that these divorced wives, v have had an unsatisfactory division of labor in the previous ge, would begin a new marriage with clear and strongly held ideas 10w tasks should be divided. However, the mean marital power of isly married wives did not differ from that of other wives (see 27 in the following chapter). 205 Table 15 HYPOTHESIS D—4: OVERALL RELATIVE MARITAL POWER BY AVERAGE RELATIVE MOTIVATIONAL INVESTMENT OF SPOUSES IN TASK-DIVISION NWEIGHTED POWER MEASURE 1 Relative Strength of Spouses' Preferences About Who Should Do Contested Tasks ‘ (1) Self Equal 2) or Stronger Self Weaker Preferences Preferences aan marital power* of: Than Spouse Than Spouse I Husbands 2.90 2.70 1 (N=86) (N=51) (N=35) ' l Wives 3.23 2.99 I (N=86) (N=58) (N=28) ‘ )wer numbers indicate higher power. Equal power = 3.00 ER MEASURE WEIGHTED BY IMPORTANCE n marital power* of: Husbands -.10 -.05 1 (N=86) (N=51) (N=35) 1 Wives .07 .01 1 (N=86) (N=58) (N=28) 1 er numbers indicate higher power. Equal power = .00 Is Trend in Predicted Yes Yes No (opposite direction) Yes 206 Considering the inconsistent findings with respect to specific sk areas (Tables 12 and 14), any interpretation of such averages must made with caution. Speculatively, I would suggest that the same eference-strength answer may have a different meaning for wives and sbands. That is, a high preference-strength may represent higher ivational investment for wives than it does for husbands. Although ety per cent of these wives were employed, it is likely that satis- tory domestic arrangements were still more consequential to them n to their husbands. Having more at stake, they are more dependent n their husbands for achieving a satisfactory division of labor than e-versa. If this is true, then the hypothesis is basically correct (there is a gender-related response bias in answers to the preference- ength question. This bias could probably be corrected (if it does exist) by :ing the task items within a larger group of items. This group 11d include issues of greater probable long—range consequence than of the task items, e.g., such things as large financial and career sions. These additional items would give a comparison point, ist which a respondent could assess the task items. A larger num- Jf answer categories should be provided to allow respondents to distinctions gmppg task items. If women really do have stronger 1rences than do men about task division, then this should show up the question format is changed in the way suggested. To summarize the tests dealing with motivational investment, ieses D—3 and D-4 are generally not supported. The one supportive 19 just discussed must be interpreted with caution. Whether the 207 lack of support for these hypotheses is due to measurement problems is a question which requires further research. Hypothesis D-5: Relative Alternatives + Relative Motivational Investment Alternatives and motivational investment are not entirely independent concepts, as was discussed in Chapter VI. Present-time levels of motivational investment——how strongly one values something-- are obviously influenced by one's past experiences about how scarce that "something“ is. :able" For example, a husband who has had many ”accep- women from whom to choose a wife--who had many alternatives in :he past--may value the rewards of marriage with a particular wife less han he would if she were the only close relationship in his experience. 0 the extent that the past exchange relations which provide the basis f motivational investment level are similar to the present alternatives arceived to exist, present alternatives and present motivational invest— ant should be inversely related. Whether particular types of alternatives are related to moti— tional investment in particular goal—objects in this manner, is an pirical question. Without knowing the historical development of a rson's value system, one cannot predict which alternatives should be lated to motivational investment in particular goal—objects. Such Dirical tests are limited in scope and, of course, cannot provide ‘ong support or non-support for the overall theoretical process by ch motivational investment levels are postulated to arise. Still, is of some interest to discover whether the measures of alternatives 208 and motivational investment level used in the present study are interrelated. Part I of Table 16 displays the relationship of my measures 3f alternatives to motivational investment in the status of being narried (a goal which is obviously mediated by the spouse). To make :hese cross-tabulations more readily understandable, the cell predicted .0 have the highest per cent in each row is marked with a ”#“ symbol. For section (A) in Part I of Table 16, what support does xist for the hypothesis is not strong. The small N's in groups (a) 1d (c) prevent a stringent test of the hypothesis. In sections (B) and (C), where the N's are more satisfactory, 1e hypothesis is not supported. Neither is it supported for Question although there is some tendency for wives with lower relative alter- tives (group c) to have higher motivational investment as predicted ee the bottom row). In summary, Hypothesis D-5 is not supported when my measures alternatives are related to motivational investment in the status being married. Part II of Table 16 deals with motivational investment in lieving a particular division of labor. The indicator used is that average relative preference-strength. Higher numbers indicate that respondent has stronger preferences than does the spouse (i.e., higher relative motivational investment). As mentioned under )thesis D-4, results for tests employing this indicator must be arpreted with caution. 2(39 Table 16 HYPOTHESIS D—S: RELATIVE MOTIVATIONAL INVESTMENT BY RELATIVE ALTERNATIVES PART I. Motivational investment measured by responses of both spouses to the statement: Married people get more out of life than people who never get married. ALTERNATIVES QUESTION A: In your own experience, do men or women have the easier time findin a cod 5 ouse? __________________JI__£__._B____ Motivational investment (1) (2) (3) question I: Married Spouses agree: Is Trend in people get more out Easier For Other Answer Spouses agree: Predicted of life Women Combinations Easier For Men Direction? I Husband agrees 1 I I more strongly # 37% 50% No (opposite direction) Both agree equally 43%# 25% : Yes Wife agrees more : strongly --- 20% 25%# - Yes 100% 100% 100% (N=1 couple) (N=51 couples) (N=8 couples) ALTERNATIVES QUESTION 8: Number of very serious or close dating relationships in past 1tivational investment lestion I: Married (1) (2) (3) Is Trend 1n ople get more out Pred1cted life Wife More Equal Husband More D1rect10n? __________________________.__________________ l Husband agrees more strongly 36%# 31% 43% I No ______________________________________________ 1 Both agree equally 36% 50%# 38% __ ; Yes Wife agrees more 197 197# i No (opposite strongly 27% . ' direction) 1 1 0/ 99% 0 S 100% (N=22 couples) (N=26 couples (N=21 couples) cell which is predicted to have the highest percentage, within the row. )1e 16. Continued. ALTERNATIVES QUESTION C: When did it occur to you to lre)marry--before 7 or after you met the 'right person. tivational investment (1) (2) (3) astion I: Married )ple get more out Both The life Husband Earlier Same Time Wife Earlier Husband agrees more strongly 8%# 46% 39% Both agree equally 39% 43%# 44% Wife agrees more strongly 54% 11% 17%# 101% 100% 100% (N=13 couples) (N=37 couples) (N=18 couples) ALTERNATIVES QUESTION D: Did you seriously consider remaining single by choice? ivational investment (1) (2) (3) stion I: Married ple get more out Wife, Yes Both Yes or Husband Yes, lifg Husband No Both No Wife No Husband agrees more strongly 33%# 41% 25% Both agree equally 44% 41%# 42% Wife agrees more strongly 22% 18% 33% 99% 100% 100% (N=18 couples) (N=39 couples) (N=12 couples) 2811 which is predicted to have the highest percentage, within the row. Is Trend in Predicted Direction? No (no difference) No Is Trend in Predicted No No (no difference) No (curvilinear) 2lLJ. Table 16. Continued. PART II. Motivational investment measured by the . average discrepancy between spouses' preference strength (how strong the1r preferences were about specific task areas . role preferences differe ALTERNATIVES QUESTION A: Motivational investment (1) (2) measure II: Mean reference- Spouses agree: strength discrepancy between arder For spouses* for: Respondent's Other Answer ender Combinations Wives (N=74) 2.66 3.04 (N=ll) (N=62) Husbands (N=74) ——— 2.91 (N=1) (N=62) ALTERNATIVES QUESTION B: Motivational investment (1) (2) measure 11: Mean preference— strength discrepancy between Self Fewer spouses* for: Than Spouse Egual Wives (N=88) 3.18 2.93 (N=33) (N=31) Husbands (N=88) . 3.05 (N=24) (N=32) ALTERNATIVES QUESTION C: )tivational investment (1) (2) easure 11: Mean reference- :rength discrepancy between Self Earlier Both The louses* for: Than Spouse Same T1me Wives (N=86) 2.88 3.13 (N=23) (N=46) Husbands (N=86) 3.11 2.87 (N=l7) (N=46) ALTERNATIVES QUESTION D: singleducsisdsez :ivational investment (1) (2) sure II: Mean reference- ength discrepancy between Self No, Both Yes or uses* for: Spouse Yes Both No Wives (N=88) 3.42 2.96 (N=l7) (N=46) Husbands (N=88) 3.07 3.04 (N=25) (N=46) 00 = spouses have equally strong preferences. 'er 3.00 = respondent has stronger preferences than spouse (i.e. investment). ~ der 3.00 = respondent has weaker preferences than spouse (1.e., investment). The average scores included only tasks d the division of on which spouses In _our own ex erience, do men or women have the eaSTer time finding a good spouse? (3) Spouses agree: Easier For Is Trend in Respondent's Predicted Gender Direction? --- : No (opposite (N=1) direction) 3.33 ' No (opposite (N=ll) : direction) Number of very serious or close dating relationships in past (3) Self Greater Than Spouse Is Trend in Predicted Direction? 2.98 : No (curvilinear) (”:24) 1 2.83 ' No (curvilinear) (N=32) ‘ When did it occur to you to (re)marry--before or after you met the 'right” person? 13) _ Is Trend in Self Later Predicted Than Spouse Direction? 2.89 : No (curvilinear) (N=l7) 1 3.11 1 No (curvilinear) (N=23) - Did you seriously consider remaining 13) . Is Trend in Self Yes, Predicted Spouse No Direction? 2.93 : Yes (N=25) I 2.58 1 Yes (N=l7) 1 , higher relative motivational lower relative motivational 212 The hypothesis is not supported for alternatives Question A. Although the relationship for Question B is slightly curvilinear, it is )enerally consistent with predictions. The spouse with a relative 1dvantage in previous dating relationships (group 3) does have lower elative motivational investment than a spouse with a relative disad- antage in dating experience (group 1). Differences between husbands 0d wives are also evident here. Wives have higher relative motiva- ional investment only when they have less dating experience than their Isbands, while husbands have higher motivational investment when their .ting experience is either equal to or less than that of their wives. is may be due to disproportionate numbers of previously divorced men falling into the group (1). Because they have been married before, eir dating experiences are likely to be more limited than their hus- 1ds'--at least when their husbands have not also been previously ‘ried. Wives in this marriage type had stronger relative preference 'ength than wives in other marriage groups.H7 Hence the findings _____.__________ H7This can be seen in the figures below: Relative Marital Experience (1) (2) (3) More than spouse Less than spouse (self divorced, (spouse divorced, spouse not) Same as spopse sejf_not)_ _ I preference— ““““““ ‘ T ‘ ‘ ‘ “ en th discre anc 'een spouses for: Wives (N=88) 3.28 2.95 2.99 (N=22) (N=43) (N=23) iusbands (N=88) 3.01 2.72 3.05 (N=23) (N=43) (N=22) 213 Question 8 may be due more to previous marital experience than to ng experience; although these two types of experiences are similar ;ome respects, previous marital experience seems more likely to act preference-strength about the division of household tasks. Hypothesis D—5 is not supported for Question C but is supported Question D. As for Question B, the data for Question D reveal er differences. Wives have higher relative motivational investment when their alternatives (with respect to singleness as an option) lower than their spouses' (group 1). Husbands have higher relative vational investment when their alternatives are either equal to or than their wives (groups 1 and 2). Again, I suspect that pre— sly married women are disproportionately represented in group (1); 1 many of these women not only have children but also have custody ' 1eir children, the option of remaining single by choice is less ble for them than for other wonen. Since these women seem to have r relative preference-strength (see footnote 117), it is likely Jrevious marital experience is what raises the preference—strength a group (1)-women about task division. However, it is not implaus— ;hat women who never would remain single by choice—-who are highly ge—oriented--would have stronger preferences about task division. larger sample size, the relative effects upon preference-strength 'ious marital experience and perception of singleness as an alter- could be checked by introducing marital experience as a control In summary, Hypothesis D—5 received little support. Those ns in which it is supported, I suspect, can be best explained 214 terms of previous marital experience. Relative marital experience, itself, may well be interpreted in terms of alternatives. Divorced rsons may have fewer potential alternatives to the present spouse lOUId they divorce again), i.e., they may be in a poorer ”bargaining sition“ than non—divorced persons. This speculation is not supported, vever, when relative marital experience is run against relative mari- 1 power (see Table 27). To conclude this section of power—dependence hypotheses, it 1ears that the support for the power-dependence theory is mixed. 10theses dealing with alternatives (D—1 and D-2) receive more support n those dealing with motivational investment (D—3 and D—4). Part II. Power Use Across Multiple Issues The second set of power-dependence hypotheses is based upon work of Dahlstrdm (1966). He suggested that the relation between r and dependence is conditioned by factors which influence the er incentive” of the more-powerful partner and the “compliance 1tive" of the less—powerful one. For Dahlstrdm (unlike Emerson, , power and dependence are conceptually independent concepts rather being defined in terms of each other. Although they are postulated inversely related, this relation is subject to predictable varia- Of the five hypotheses in Part II of Chapter VI, three were in the present study. Hypothesis D—6, about the effect of intrin- luation of power for its own sake, and Hypothesis D—9, postulating 215 exponential function in the relation of dependence to power, were tested. The three hypotheses which were tested deal with the act of multiple issues upon the amount of power used. For purposes this exploratory research, I have focused upon two-variable tests. ther research should be done to check whether the relationships with "age power use which were found would hold up when dependence is ‘oduced as a third variable. Before presenting tests of the hypotheses, the dependent able of “amount of power exercised" will be briefly described. nt of power exercised or ”used” within a couple is different from all relative marital power. Here the concern is not with phigh se has more power but with how much power is exercised, regardless 10 has the ”advantage.“ Other things being equal, for exampie, I ider power use to be lower in couples who make half-way compromises in couples who do not. I The indicator for amount of power exertised was constructed $3 king the absolute difference between a respondent's overall rela- Jower score and 3.00. The resulting score is the same for husband ’fe and simply reflects how far they deviate from an equal—power ion.H8 H8In future research, I plan to use a somewhat different :or which may prove to be more sensitive. It involves figuring 'iation-from—equality for task—specific power scores first, and lculating an average from these scores. The indicator I have this report probably underestimates the amount of power exer- n couples with equal relative marital power. 216 10thesis D-7: Number of Areas of .agreement + Amount of Power ercised When more than one area of disagreement is present, trade—offs .ween spouses are now possible. A spouse who would otherwise not e in on one issue may be willing to do so in return for the other's pliance on another issue. Hence, the average amount of power“9 rcised is predicted to increase with the number of issues at stake. In the present study, spouses could disagree about who uld do all eight tasks. At the other extreme, they could agree It who should do all tasks (although only two couples didizo). majority of couples differed on anywhere from one to four task IS. For purposes of Table 17, these couples were placed in 1p (1) and those who differed on more than half of the task areas roup (2). Table 17 (p. 220) reveals support for Hypothesis D-7. The ter power use among couples who disagree on a large number of areas may, as Dahlstrom would suggest, point to the existence of —offs. This interpretation must remain tentative for the moment, er, as no direct measure of number of trade—offs is included in table. (I did ask couples ppy they arrived at their division of ligPreliminarily, “power“ in this context should be taken to Tat A will get his or her way in a specific, joint task—area e the fact that B disagrees with that particular way. 120No power scores could be computed for these couples or other couple where the data about task division was not e. 217 labor for each task, with one of the answer—options being ”We decided that one of us would do it and that the other would do some- thing else, in return”. This “trade-off“ option was chosen by ygyy few respondents, however). Another explanation for the findings in Table 17 is that :ouples who disagreed about many tasks were also the ones who, when ;hey disagreed, had widely discrepant preferences. Such preferences vould increase power—use scores, given the nature of the power index mployed. This possibility will be discussed later, in conjunction ith Hypothesis D—9. In summary, Hypothesis D—7 was supported. The greater the Tmber of areas of disagreement between wife and husband, the greater IS the average power use within the couple. Whether or not this 1n be attributed to the presence of trade—offs, as suggested by .hlstrbm, can not be determined without further research. pothesis D-8: ‘Average Discrepancy in tivational Investment -+ Amount of ver Exercised Trade—offs and the exchange of power resources, according to 1lstr6m, ought to be most evident when the interests of spouses vary. A and B feel strongly about the ppme disputed issue (have high ivational investment), a trade-off is unlikely. But if A feels e strongly about issue 1 and B feels more strongly about issue 2, 1 each has something to gain by a trade-off; the ”compliance in— .ive” of each spouse is increased. Power use should, therefore, 218 increased, as motivational investments of spouses become more crepant. . One could argue, of course, that such an exchange of power ources is somehow different from "power", as the term has been 1 up to now. If spouses agree to take this exchange willingly, 1 may actually change their initial preferences—-reducing the gap 'een what they want and what they get, which is the way I have ured power. "Amount of power exercised“, therefore, must be rpreted with this possibility in mind. For the test of Hypothesis D—8, the measure of motivational stment discrepancy used was somewhat similar to the average rela— preference strength indicator (see Hypothesis D—4, Table 15). 1 the present interest is only with the gegygg to which spouses' 'ests (strength of preferences) vary, and not with yhg has the ger preferences, a different calculation procedure was used. ach item on which spouses had different task—division preferences, )solute difference between their preference—strength scores was ed. These absolute difference scores were then averaged across sagreed—items. The resulting average was the same for husband fe and simply represented the degree to which the couple ed from a situation of equality. Results for Hypothesis D-8 are contained in Table 18 (p. 220), Jothesis receives some support, to the extent that couples with 'erage motivational-investment discrepancies do exhibit higher se than couples with low discrepancies. However, the highest 219 el of power use is under condition (2) rather than under condition , as predicted. This unexpected result may be due to measurement )lems or it may be due to disproportionate numbers of couples with 11y differing task—division preferences falling into group (2), a 1ation which would inflate the power—use scores. At this point, it IS appropriate to proceed directly to the consideration of such erence differences. thesis D—9: Average Discrepancy in erences + Amount of Power Exercised The final factor to be considered with respect to trade-offs 1 multiple issues is the variable of how far apart the preferences e two spouses are. The smaller the extent to which the interests erred outcomes) of spouses coincide, the greater should be the nge of power resources (e.g., trade-offs). This implies that pay compromises will occur more often when spouses have only :ly different preferences than when preferences are far apart. pancy between spouses in preferences (Hypothesis D—9) and in‘tWEY th with which these preferences are held (Hypothesis D-8) are nplementary aspects of the degree to which spouses' interests 1e. As with the previous hypotheses, the less the interests of coincide, the greater should be the chance of tradeéoffs, and her the total power use across multiple issues. Findings for Hypothesis D—9 are presented in Table 19 (p. 220). )le reveals support for the hypothesis. Note that power use ; increase much between the low- and medium-discrepant groups. 1y couples must be fairly discrepant in their task—division 220 Table 17 HYPOTHESIS D—7: OVERALL AMOUNT OF POWER EXERCISED, BY NUMBER OF AREAS ABOUT WHICH SPOUSES' PREFERENCES DIFFERED. Number of task-areas where s ouses references about task—flivision differed (out of eight areas) 15 Trend in (l) (2) Mean amount of power* Low High Predicted exercise (1-4) (5-8) Direction? Couples (N=87) .04 .36 1 Yes (N=56) (N=31) *Higher numbers indicate higher power use. Table 18 HYPOTHESIS D-8: OVERALL AMOUNT OF POWER EXERCISED, BY AVERAGE DIFFERENCE BETWEEN SPOUSES IN STRENGTH OF TASK-DIVISION PREFERENCES. Avera e discre anc in s ouses' reference stren th lacross task-areas where preferences differed) (1) (2) M(3)h Is Trend in ' Predicted dean amount of power* Low Medium (0.0—0.9) (1.0—1.9) (2,0:— ~2h .9) Direction? exercised Couples (N=87) .08 .35 . :No (curv11inear) (N=32) (N=46) (N=9) ' 'Higher numbers indicate higher power use. Table 19 HYPOTHESIS D—9: OVERALL AMOUNT OF POWER EXERCISED, BY AVERAGE SIZE OF DI SCREPANCY BETWEEN SPOUSES' TASK- DIVISION PREFERENCES. Average size of discre epancy between spouses' task- division references (average excludes task- ~areas where spouses' pre erences agreed) (1) (2) (3) Is Trend in n amount of power* Low Medium High Predicted rcised (1.0—1.4) (1.5—1.9) (2.0-4,0) Direction? Couples (N=87) .14 .15 .33 : Yes (N=55) (N=20) (N=12) - 'her numbers indicate higher power use. 22l ferences before the predicted higher power use occurs. The h-discrepant couples (group 3) are those who--on the average--are least two points apart on the five-point preference scale. This IS, for example, that one spouse prefers to divide a task equally e the other prefers to have one person do it all. These high- repant couples may, as Dahlstrom would suggest, be engaging in eéoffs where he “wins” sometimes and she ”wins“ sometimes. It is lly possible, however, that the more-powerful spouse is “winning“ 1f the time. Further work should be done to check out which is lore common situation. Hypotheses D—7, D-8, and D—9 deal with interrelated aspects wer use among married couples. A fruitful avenue for further‘ rch would be to perform an analysis of variance, identifying the ~ in which number of areas of disagreement, discrepancy in ‘ences, and discrepancy in the strength with which preferences ‘ld, operate together to affect the total amount of power exer- To summarize Part II of the power-dependence hypotheses, )unt of power exercised within a couple was found to increase Ie number of areas of disagreement and with the average discrep- ‘tween spouses' task-division preferences. Partial support was und for the hypothesis that power use increases as the motiva- investment levels of couples become more discrepant. In , Dahlstrom's notion that power use will be low when the :s of spouses coincide and high when their interests vary, l support. Whether or not this increase in power use can be 222 ributed to the existence of trade-offs in multiple-issue situations not be determined without further research. CHAPTER XI RESULTS OF CRITICAL TESTS BETWEEN THEORIES The preceding two chapters have revealed rather mixed patterns upport and non—support for the several theories of marital power. )urpose of this chapter is to attempt to determine whether one ‘y can be said to have greater explanatory power than another. A er of test situations were devised, in which contradictory results be predicted by different theories. The use of the term ical tests” for such test situations is something of an over— nent, given the exploratory nature of the study and the experi— l nature of the indicators. Still, it is a promising avenue in :uch test situations build in the possibility of finding negative' ce, i.e., of increasing the testability of theoretical formu- s. This chapter is organized in the same manner as Chapter VII. )f alternative theoretical processes will be presented first. :ond section deals with specific “resource” Variables, for which inks with theoretical processes can reasonably be inferred. 223 224 Alternative Processes etence versus Gratitude Theories ___.____________._.______.._____ G—Cl: Relative Competence and :ive Income + Overall Marital 0f the three hypotheses presented in Chapter VII, two were d in the present study. The sample size was deemed too small to y the two control variables required for a direct test of hypo- ; C-G 3. I Test C-G l predicted that the effect of perceived competence Iarital power would operate even when relative income is controlled. and Wolfe assumed that higher-resource (higher—income) spouses kely to be more competent. They also assumed that higher-income 5 would receive more gratitude and respect from the other spouse. both the gratitude and competence assumptions would predict power for a higher—income spouse. Obviously, there must be differences in perceived competence eople within the same relative income levels. If the widely ted relation between relative income and relative marital power ale 25 for yet another example), is due totally to gratitude of er-income spouse, then these within-group differences in per- :ompetence should have no effect upon marital power. The com— formulation, on the other hand, would predict that perceived ce would still be related to marital power when relative income olled. The gratitude formulation, then, predicts no difference among 2 conditions in Table 20, while the competence formulation 225 :ts that mean power will vary, being the lowest in condition (l) we highest in condition (3). This raises the problem of determin- yiJnuch difference in mean power is required to disprove the “no ~ence" gratitude hypothesis. For this purpose, a statistical )f significance level was required. A two-tailed t-test for [ting the difference between two means was employed (degrees of rm = N, + N2 — 2), with calculations performed on a small ti—Underwood Programma lOl computer. The sample undoubtedly t meet all the assumptions of this statistic. Given the present or a simple and pragmatic way to differentiate between theories loratory research, the t-test was chosen in preference over such ime—consuming techniques as Tchebychev's inequality. Differences I mean power scores which were significant at the .lO level or were taken as support for the competence formulation, and hat failed to reach thissignificancelevel were considered tov the gratitude hypothesis. However, any statements that the ie hypothesis is ”supported” must be interpreted conservatively, I as no direct measures of gratitude were available. Results for Test C—G l are contained in Table 20. For the ce of dependability, the trend is in the direction predicted )mpetence formulation. This is true for both levels of rela— lme, that in which the husband has a very large income—advantage wife (i.e., makes at least twice as much annual income as she does) TEST C-Gl: 212(5 TABLE 20 OVERALL RELATIVE MARITAL POWER, BY PERCEIVED COMPETENCE RATINGS OF BOTH SPOUSES, WHEN RELATIVE INCOME IS CONTROL EIGHTED POWER MEASURE I HUSBAND MAKES A 1T DEAL MORE INCOME l WIFE lean marital power* of: Wives (N=2l) Husbands (N=21) HUSBAND DOES pp: A GREAT DEAL MORE 1E THAN WIFE ean marital power* of: Wives (N=59) Husbands (N=59) lUSBAND MAKES A DEAL MORE INCOME IFE n marital power* of: dives (N=22) lusbands (N=22) SBAND DOES NOT BREAT DEAL MORE THAN WIFE marital power* of: ves (N=59) sbands (N=60) COMPETENCE A. (l) (2) Self Lower Both Rating Than Same S pouse Rating . 3.19 (N=ll) (N=7) —-- 2.8l (N=3) (N=7) 3.29 3.05 (N=19) (N=29) 3.08 2.92 (N=9) (N=3l) COMPETENCE B. 3.10 3.6l (N=3) (N=l5) 3.38 2.39 (N=4) (N=l5) . 3.3 (N=20) (N=22) 2.37 2.78 (N=l7) (N=23) Dependability (32 Self H1gher Rating Than Spouse Which Theory is Supported? (N=3) . 2_55 Grat1tude (N=ll) 2.85 (N=ll) ' 2'7] Gratitude (N=l9) Intelligence 3.l3 (N=l7) 2.95 (N=20) Neither one Gratitude mbers indicate higher power. Equal power = 3.00. 227 at in which the husband does not have such a large advantage.]2] r, this trend does not approach statistical significance. The ude hypothesis, therefore, cannot be rejected. For the competence of perceived intelligence, no support is for the competence hypothesis. This is not surprising, given ;ative results for Hypothesis C—l (see Table 6). As in Table 6, is higher in condition (3) than in condition (l) as predicted—- y when the husband has a large relative-income advantage. ;his relative income—condition, the trend comes closer to meet- competence prediction than in Table 6 where income was not led. However, the difference between the equal—competence and gher groups is statistically significant at the .05 level, not )redicted direction. Hence, neither the competence nor the ‘atitude) hypothesis can be said to be supported. Under the second ralative- income condition, differences be— ans are not statistically significant, and the gratitude hypo— annot be rejected. This income group contains a diverse set as, which may account for the inconsistent means. Subdividing 1p might reveal some trends which are obscured by combining re wives make more income with cases where incomes are equal ly to the husband's advantage. 2IThe latter group includes couples where the husband has a come-advantage, where spouses have roughly equal income, and les where the wife makes more money than the husband. 228 In summary, the gratitude hypothesis receives stronger support toes the competence formulation. Trends are consistent with the ,ence hypothesis for the competence of dependability, however. -G 2: ”Relative Effects“ petence and Income A somewhat different strategy for assessing relative effects independent variables (Rosenberg, l968) requires that the sistent” cells be compared. The competence assumption would : that a more—competent, lower—income spouse would have higher ;han a less—competent, higher—income spouse, while the gratitude .ion would make the opposite prediction. This exact comparison be made, inasmuch as there were too few higher—income ines to 122 separate group. The following values, taken from Table 20, low an approximate test of Hypothesis C—G l: (a) (b) ’ Hu$band > Wife Wife > Husband (or Husband somewhat > Wife) . petence advantage: ’ Wive < Husband Husband < Wife Income advantage: ean marital power of wives: 2.50 3.29 (N=3) (N=19) 1 1e value for condition (a) above had been omitted from Table 20, f the small number of respondents. Means were used for only one ecause the dyadic nature of all indicators in this particular not allow the trend for husbands to differ from the trend for th sets of means were reported only for ease of interpretation 0 and other tables where this occurred. 229 The very small number of couples falling into condition (a) as any assessment of relative effects impossible. The figures :h are available are in the direction predicted by the gratitude 1thesis, however. l-Input versus r—Dependence Theories The gratitude and competence formulation share, I submit, a )n implicit assumption that marital power is “given up" in ex— 18 for benefits received, and that this exchange operates accord— .o a norm that exchange—inputs should be equal. That is, a spouse ontributes fewer ”resources“——particularly fewer tangible urces“——to the marriage feels obliged to make up for this in— ity by giving greater deference and compliance. If this is true, l low-power spoUse should not perceive the exchange of benefits fair” (e.g., “it is only fair for John to make more of the ons, since he's making the money”). The gratitude and competence ations, then, would predict no differences in perceptions of 35 among different power levels (Test GC—D 2). Nor would they : differences in “fairness" among individuals who perceive the exchange as unequal with respect to some specific ”input” such rt (Test GC—D l) since such inequalities are “equalized out“ in rall-picture by operation of this norm. The spouse who ”has it a easier” in the marriage gives more in other ways. Power-dependence theory would make no such predictions. It asonable to predict (following Homans, 1961) that inequities 230 i power—outcomes or of effort-inputs, for example) are not necessarily ceived as ”fair”; specifically, I would predict that the more—advantaged tner will perceive the exchange as fairer than will a less—advantaged tner. In addition——since Homans suggests that inequity may “bother“ are—advantaged exchange partner, albeit less than it bothers a ;—advantaged one—-I also make the tentative prediction that the .l—exchange situation will be perceived as the fairest of all. GC—D l: Relative rt + Task—Specific sfaction Direct questions about “fairness” proved awkward during the est interviews, and respondents who did answer such questions were 'illing to rate the division of labor for any task as unfair. quently, questions about satisfaction were substituted. These ions were administered during the final in-home interview by of a paper-and—pencil check list which the respondent filled out vacy. Following a question about how the eight task areas were d between husband and wife, respondents indicated for each task satisfaction with that division of labor. The question used was atisfied are you with who does this?“ Four answer categories were ed, varying from extremely satisfied to not at all satisfied. satisfaction answers were assigned higher scores. The independent variable for Test GC—D l was a question which a used previously in this report as an indirect measure of moti— 1 investment (see Table ll). Following a question about whether 231 Irried life is generally easier on men or women, respondents were ked: ”What about your gyp marriage? Would you say the same ing? (Right now, who probably has it a little easier, you or your ouse?)“. As I have indicated earlier, the data in Table 11 ;gested that this question did not tap motivational investment. :her, it appears to tap perceptions of relative effort—inputs. lCh relative effort may be the outcome of the power balance; if so, results of Table 21 which follow should be similar to Table 22, which power is directly related to satisfaction.) Results for Test GC-D l are presented in Table 21. The null 1titude) hypothesis that satisfaction will not differ according to :eptions of relative effort should be considered to be rejected if st scores are significant at the .10 level. Such tests, of course, be viewed simply as pragmatic devices to assist in data inter- itTOh. I consider the power-dependence hypothesis to be supported aver satisfaction means are the lgpeg: in condition (1) of all the conditions and when the difference between condition (1) and [ condition (2) or (3) is statistically significant. That is, the st satisfaction could either be in the “both the same effort” :ion (as Homans would suggest) or in the “self easier“ situation. which correspond completely to Homans' ideas are indicated, how- by double asterisks in the right—hand column of Table 21. Table 21 reveals striking gender differences. Wives fit power— ance predictions for almost all tasks, including all of the lonally “female” tasks. In the remaining two task areas, the trend Table 21 TEST GC-D I: PERCEIVED FAIRNESS 0F TASK DIVISION (AS REFLECTED IN SATISFACTION), BY PERCEPTIONS 0F RELATIVE EFFORT INVESTED IN THE MARRIAGE. Relative effort, as perceived by individual respondent Who has it a little easier rioht now _ . Zou or iour Siouse? hich Theory t d? (1) (2) (3) W Sp use Easier Both the Same Self Easier Suppor e (N= 38 husbands, (N= 39 husbands FACTION WITH WHO (N= 12 husbands, TASKS: N= 31 wives) N= 40 wives) N= 17 wives rocery Shopping 1 Mean satisfaction* of: 1 Husbands (N=89) 3.58 3.45 3.41 ' Equal-input Wives (N=88) 3.07 3.33 3.59 f Power ‘ ‘ pairing things around the house : Mean satisfaction* f: l Husbands (N=89) 3.17 3.61 3.36 1 Power—dependence** Wives (N=88) 3.26 3.28 3.65 1 Power A ‘ l l 1 1 :king complaints to salesmen lan dlo Mean satisfaction* of: Husbands (N=89) 3.50 3.37 3.03 1 Neither one 3.10 3.23 3.59 Power-dependence Wives (N=88) zping in touch with relatives Mean satisfaction* 0 Husbands (N=89) Wives (N=88) iding who you will invite the hous lean satisfaction* of: Husbands (N=89) 3.25 3.45 3.24 ,' Neither one Wives (N=88) 3.37 3.38 3.47 1 Equal-input# l 09 care of the car aan satisfaction* of: Husbands (N=89) 3.42 3.42 Wives (N=88) 3.20 3.38 3.59 . Equal—input# 1 I 3.18 3.00 3.03 3.47 2.94 I l l I l 2.97 1 Neither one I Power-dependence l 1 l 1 l 3J0 lNfitMrom 1 rhgswmr an satisfaction* of: Husbands (N=89) 3.50 3.37 3.34 ,' Equal—input Wives (N=88) 3.00 3.37 3.34 ' Power I l l 1 Equal-input ; Power—dependence 4 J the supper dishes n satisfaction* of: 3.37 3.23 Husbands (N=89) 3.25 Wives (N=88) 2,65 3.20 3.53 nbers indicate higher satisfaction. Lowest possible satisfaction is 1.00 and highest possible lon is 00 d corresponds with Homans' predictions that equality (the middle column) should have the highest tion of all lot statistically significant the trend is in the direction predicted by power-dependence theory 233 responds with power-dependence theory although it is not statistically :ificant. Husbands, in contrast, fit the power-dependence predictions ' with respect to one traditionally “male” task, that of repairing gs around the house. Could it be that husbands believe in the of equal-inputs while their wives do not? These unexplained erences suggest the usefulness of in—depth case studies. If such a / were pursued, I would suggest the inclusion of a larger proportion )uples with non—employed wives than was the case in the present e. It may be, for instance, that only employed wives, but not wives, fit the power-dependence predictions; those employed wives lso do most of the household tasks may constitute the majority of atisfaction wives in group (1) in Table 21. In summary, the power-dependence formulation was consistently ‘ted for wives and the equal-input formulation was supported for ldS with respect to most traditionally "female" tasks; in other reas, however, neither theory was supported for husband. 3-0 2: Task—Specific Marital + Task—Specific Satisfaction The second test of equal-input versus power—dependence assumptions ith relative power and level of satisfaction in specific task areas. to the procedure followed for the previous hypothesis, the power— ice hypothesis was considered to be supported whenever mean satis- was the lgweg; in condition (1) of all the three conditions, 19 that the difference between condition (1) and either conditions 3) were statistically significant. 234 The equal-input hypothesis that the perceived fairness of iange (measured in terms of satisfaction) will not vary among equal- er and unequal-power situations is supported for husbands in three :he four traditionally ”female” task areas, as seen in Table 22. traditionally "male” tasks, husbands are generally consistent the power—dependence hypothesis (although the trend is statistic- significant in only one of the three tasks). As in the previous 3, gender differences are apparent. Wives are consistent with )ower—dependence hypothesis for all “female“ tasks (although two. 1e four are not statistically significant), and are consistent with -qual—input hypothesis for most traditionally “male” tasks. There are some similarities between the patterns in Tables 21 2. In both, there is some tendency for each spouse to fit the -dependence predictions with respect to tasks which were tradi— ‘ly assigned to that spouse, and to fit the equal-input predictions .her tasks. This trend is stronger in Table 22, however. —GC 3: Relative Potential Alternatives lative Current Alternatives + Overall 1 Power Spouses having sources of ”resources” (other than from the ‘) are said to have ”alternatives.“ For example, an employed 5 higher alternative sources of income (and other valued es) than does an unemployed wife. A direct relationship between e's income and her marital power, however, could be accounted >ne or all of the marital power theories which I have discussed. : experience could have increased her competence in the eyes 2 3 5 Table 22 TEST GC-D 2: PERCEIVED FAIRNESS 0F TASK DIVISION (AS REFLECTED IN SATISFACTION), BY LEVEL OF RELATIVE POWER IN SPECIFIC TASK ARM Relative task-specific power (l) (2) (3) Self Less Self More ISFACTION WITH Power Than Power Than Which Theory DOES TASKS: Spouse Spouse Equal Power Supported? Grocery Shopping Mean satisfaction* of: Wives IN=4ZI 2.86 3.53 331 Power-dependence (N=14) (N=15) (N=13) Husbands (N=43) Equal—input Repairing things around the house I 1 I I 3.33 3.27 3.54 ' (N=15) (N=15) (N=13) ' I I Mean satisfaction* of 1 I I I I Wives (N=46) ' 3.35 3.13 3.33 Equal—input (N=l7) (N=8) (N=21) Husbands (N=45) 3.13 3.50 3.48 Equal-input# (N=8) (N=16) (N=21) Iaking complaints to salesmen ' 1nd landlord , Mean satisfaction* of: , Wives (N=52) (2.80) 3.47) (3.60) 1 Power-dependence** N=25 (N=l7 N=10 . Husbands (N=52) 3.12 3.48 3.10 1 Neither one (N=17) (N=25) (N=10) 1 zeping in touch with relatives ' Mean satisfaction* of: ' Wives (N=53) 3.09 (3.88) 3.43 l Power-devendence (N=23) N=16 (N=l4) ' Husbands (N=53) 3. 25 3.22 3.14 ' Equal-input (N= 16) (N=23) (N=14) ' :iding who you will invite I the house (omitted) 1 ing care of thec : Mean satisfaction* rof: 1 . Wives (N= 44) 3.22 3.30 3.19 , Equal—1nput (N=18) (N=10) (N=16) 1 Husbands (N=45) 3.00 3.16 3.25 . Equal—input# (N=10 (N=19) (N=16) 4 raring supper . lean satisfaction* of: 1 Wives (N=39) 2.92 3.31 3.31 1 Equal—input# (N=13) (N=13) (N=13) . Husbands (N=40) 3.31 3.21 3.08 I Equal-input (N=13) (N=14) (N= 13) l 1 supper dishes Ian satisfaction* of: Wives Zn: 435 2.84 3.13 3.38 I Equal—Input# (N=25) (N=15) (N=8) I Husbands (N=48) 3.07 (3.36) (3.25 I Power-dependence (N=15) N=25 =8 . lmbers indicate higher satisfaction. Lowest possible satisfaction is 1.00 and the highest is 4.00. Satisfaction scores are reported only for those couples who differed in preferences k division. d corresponds with Homans‘ predictions that equality (the middle column) should have the highest tion of all not statistically significant, the trend is in the direction predicted by power—dependence theory. 236 of her husband, her income could have increased his gratitude and respect for her, and/or the fact that she has an alternative source of income other than from her husband could have decreased her dependence 1pon him. Potential (but currently not utilized) alternatives, on the ther hand, are predicted to be related to marital power only in power— ependence theory. Some support was shown in Table l0 for Hypothesis -2, that such alternatives should be directly related to marital power. 1 itself, Table 10 did not provide a rigorous test, however, because ’ the possibility that the people with high potential alternatives e also the ones who have high currently-used alternatives. If poten- al alternatives are indeed directly related to marital power, then is relationship should persist even when the effect of present-time ternatives is controlled. The indicator chosen for the ”control variable"-—present—time :ernatives—-was that of relative income. Within groups of couples ing similar relative incomes, the power-dependence theory predicts t marital power is directly related to relative potential alterna— es, and the competence and gratitude formulations predict no dif— ence in power. The null (competence/gratitude) hypothesis was considered to upported if no two means were significantly different (at the .10 l). The power-dependence hypothesis was considered to hold if r was the lowest in condition (l) and highest in condition (3) and 1e difference between one pair of adjacent conditions was statisti- I significant. TEST D-GC 3: Table 23 OVERALL RELATIVE MARITAL POWER, BY RELATIVE POTENTIAL ALTERNATIVES. WHEN RELATIVE INCOME 0F SPOUSES IS CONTROLLED UNWEIGHTED POWER MEASURE PART I. QUESTION A. Potential for finding a suitable alternative partner Do men or women have the easier time finding a good spouse? (l) (2) (3) Spouses agree: Spouses agree. Harder For Easier For Respondent's Other Answer Respondent's Which Theory Is Gender Combinations Gender Supported? WHEN HUSBAND MAKES A GREAT DEAL MORE INCOME THAN WIFE Mean marital ower* of: Husbands (N=19) -:- 2:77 1:55 , Power-dependence (N-O) (N-l7) (“*2) ‘ (but note the Wives (N=19) 4.35 3. --- , small number of (N= 2) (N=l7) (":o) . respondents) WHEN HUSBAND DOES NOT MAKE A GREAT DEAL MORE 1 INCOME THAN WIF 1 Mean marital ower* of: Husbands (N=51) --- 2.87 2.83 | (N=1) (N=42) (N=8) . Wives (N=5l) 3.18 3.08 --—. 1 Equal-input” (N=8) (N=42) (N=1) . QUESTION B. Nunber of very close or serious dating 1 ‘ L' in the past (I) (Z) (3) Which Theory Is Self Fewer Spouses Equal Self Greater Supported? WHEN HUSBAND MAKES A GREAT DEAL MORE INCOME HAN WIFE Mean marital ower* of: Husbands (N=22) 2.00 2.135 2.80 1 (N=5) (N=7) (N=10) , . Wives (N=22) 3. 20 3.14 4.00 . “Elmer °"e (N= 10) (N=7) (N=5) I WHEN HUSBAND DOES M 1 MAKE A GREAT DEAL MORE INCOME THAN WIFE ' Mean marital ower* of: ' Husbands (N=61) 2.71 3.07 2.79 ' (N=18) (N=24) (N=l9) Neither on e Wives (N=61) 3.19 2.85 . (N=20) (N=23) (N=18) 1 2138 Table 23. Continued. PART II. Remaining single as a potential alternative to the present marriage QUESTION C. When did it occur to you to (re)marry-—before after ou met the “ri ht' n? y 9 person (I) (2) (3) Self Earlier Both The Self Later Which Theory Is TMnSwme Sme TMnSmme wwmmw? WHEN HUSBAND MAKES A GREAT DEAL MORE INCOME THAN WIFE flew* 01’: , Husbands (N=22) 3.48 2.36 2.2 ; (N=6) (N=10) (N=6) 1 P d Wives (N=22) 3.75 3.64 2.52 1 ower- ependence (N=6) (N=10) (N=6) . I WHEN HUSBAND DOES N91 1 MAKE A GREAT DEAL MORE ' INCOME THAN WIFE ' Mean marital power* of: Husbands (N=59) 3.10 2.99 2.50 : (N=9) (N=35) (N=15) | P d d Wives (N=59) 3.50 2. 2.90 1 °wer‘ ape“ e“ce (N=15) (N=35) (N=9) 1 I QUESTION D. Did you seriously consider remaining single by choice? 01 (2) <3) , Self no, Both Yes or Self Yes, Wh1ch Theory Is Spouse Yes Both No Spouse No Supported? WHEN HUSBAND MAKES A GREAT DEAL MORE INCOME THAN WIFE Mean marital power* of: l Husbands (N=22) -—- 2.84 2.57 (N=3) (N= 12) (N=7) ' Wives (N=22) 3.43 3.15 -—- . P°Wer‘depe"dence (N=7) (N=12) (N=3) 1 WHEN HUSBAND DOES NOT l MKE A GREAT DEAL MORE INCOME THAN WI | I Mean marital ower* of: 1 Husbands (N=6l) 2.95 2.97 2.44 1 (N=l9) (N=32) (N=10) . P d Wives (N=6l) 3.56 2.97 3.05 1 ower- ependence (N=10) (N=32) (N=19) I *Lower numbers indicate higher power. Equal power = 3.00. #Trend is consistent with power-dependence predictions, but is not statistically significant. 239 As can be seen in Table 23, power—dependence theory receives consistent support for Questions C and D and some slight support for Question A. Question B provides support for neither theory, as the trend is statistically significant but not in the direction predicted by power-dependence theory. A comparison of Table 23 with Table 10 suggests that control- ling for the effects of relative income strengthens the relationship between potential alternatives and power for Questions A and D (the trend for the latter is not entirely consistent with predictions, which was not the case in Table 10). There is little change for Ques- tion B (which does not support the hypothesis on either table), and the relationships for Question C appear to be weakened. To summarize the findings for Test GC-D 3, power-dependence theory was found to be supported over the competence and gratitude formulations. Competence versus Power-Dependence Theories ______________E So far, the two theories given the most attention have been the perceived competence and power—dependence formulations. Each, when :onsidered separately, showed rather mixed results. The test situations hich follow attempt to assess whether one theory can be said to have reater explanatory power than the other. “— est D-C 1: Relative Competence and elative Dependence + Overall Mar1tal Jwer At this point, it is useful to ask whether relative dependence .e., motivational investment and alternatives) will each be related 240 to marital power, when the effects of relative perceived competence is controlled, and vice-versa. Table 24 contains data for all of the previously employed indicators of relative dependence, in combination with those for relative competence. The right-hand column of each sub- table summarizes whether or not the trends are consistent with predic- tions. No attempt was made to assess statistical significance of these trends. Since task—importance will be dealt with later, I will dis— cuss only the results in which the unweighted power measure was employed at this point. The competence of ”intelligence" had not fit predictions (see Table 6) and-—not surprisingly——does not have the predicted effect upon marital power when the relationship is controlled for dependence. Competence—formulation predictions for “dependability,” which previously had been supported in Table 6, also receive some support in Table 24. In no case is the predicted trend found under all conditions of relative dependence, however. The pattern of findings for relative dependence is somewhat :learer than for relative competence ratings. Alternatives indicators land C and motivational investment indicator B, all of which had reviously fit the power—dependence hypotheses (see Tables 10 and 14, espectively), continue to do so under most conditions of relative Dmpetence. In addition, alternatives indicator B--which previously id displayed a curvilinear relationship, somewhat in the direction )posite to the predicted (see Table 10)--now fits the power-dependence 'pothesis for some conditions of relative competence. Heuzv . Aauzv AmF1zv No_uzv _ Amuzv ANruzv Amnzv 11111111111111-1111111 N eoeeeeeod e? , mmnz .floeuzv Ae_1zv oz - meeeemsx E ereo 1 deememsx _ E sm.N mm.N eo.m E em.~ om.~ mm.m E mm.N ee.~ mm.~ .......... ascends: 1-1-1-111m-mmflmwmmmm1s1111-111-1111111111111mmmm11n- WMZM1NMHMZM-1NWMZM1- “mnzv Am_1zv Awnzv _ Amuzv Am_1zv-flflflmmm at seed 1 md>e3 _ oz - muse: E Ne.m em.m No.m ee.~ _N.m mo.m E qm.~ ma.m Ne.m ............. md>ez 1E1- 1-1111u1111111r 111111 1111111 111-11y- 11111 111111. 11- 11w 11111-- -11 -1 1111 dueemEEeoeeH Am --- n a a meoeueeeod _ _ :oeeeueod e? E em1z _~e_1zE Neequ Nmuzv Nanzv NEE1ZE Nm1zM Edoeuzv dwuzm Nenzv e? s_eo 1 meeeemex E »_=o - meeeemex E E oe.m ew.~ oa.N , E_.m em.N mm.~ _ MF.N eo.m mo.m ........... wdemgmsz l1111111111111111111IIIIIIIIIIIIIIIIIIIIIILI11111111 111 11 11111 1111 111 a eoeeeeeou E m :o_eee ewnz E 412% “mnzv ~o_an _ Aenzv Neean Nmnzv a Amuzv Am_1zv AanzM at s_eo - muse: E1eou e_ sees md>t3 E _ mm.~ ea.N Em.m Nw.N ma.~ mm.m E om.m m_.m om.m .............. mm>e3 _ mcowpmfimt mcwpmu _ _ H E xuwfiwnmucwawo A< mzowcwm mo .02 Am 1-111111-111-11-1-1-11111-1111111111111111c111111111~mmzq--Nmmz“11~mw2M11r Nflm2M1~wmsz1~m_nzq-$1-NonzM-1~flm2M11Nonzq1111 a a m meowpwu:ou _ . Nmnz a? >Fco 1 mvcmnwzz E oz . _ Rm.m m~.N NR.N E No.N mm.N mo.m E 1111 oo.m 1111 .......... mucmnmzz 1-----11111111111 _ . _-MNNZ n Aonzv N_1zE Aenzv11n~m_nzv Aewnzv Ne_nzv _ Amnzv11Nenzv-1~mmmw mo> 1 mm>w3 . oz E E 1111 oo.m 1111 E Fm.N Na.m _o.m E mm.m mm.m mo.m ............. mm>w3 11111111111111111111L11111111 _ E 11 _ 1111111 wocmm?__wch Am . A Nenz E Amnzv Awnzv N012“ EANNHZE NNNHZV noenzv . Aonzv A_nzv Aonzv mu mm> E _ , ox._ _m.N 1111 E me.m mm.~ No.m . 1111 oo.m 1111 .......... mvcmnmzz oz - J N cote _ Nanz , Acuzv--flenzv1-douzv EAN_1ZE nomuzv AeanE . Amuzv Amnzv Awnzv mm> E 1Tucoo cw »_co E 1111 oo.m 1111 om.~ mo.m RN.m _ om.¢ m_.m om.¢ ............. mm>e3 _ wmmzoam uoom me_eeee mete Auwpwnmucmawa A< memmm mm; 0:: A< "we ocammme ”m_ mcammwe mucmummEou mm>wemcstF< "can: «ease Hap wocwvcm mo1cwzoa wucwuw Eco 2 szcm Lose; cwcmwx Fmsvm Luzom sugar: szcm Luzop ” muzmpmazou o o Lowsk Luvs: FmpOH wmzo m such cm; w: cash cmzoJ ”mm>HHHHm oz< muzmozmamo m>HH<4mm >m mmzoa 4HH<4mm 44<¢m>o ”F U1o Hmmh «N mFQMP ‘ 2412 ozE _ oz 1 mucmamzzn “1-11-11--- E mwmz1111 oz _ oz 1 mw>w3 E n ~N.m mm.N E _ E 11111111111111111r11111 11 1111111111 11F 1111 1 1111 11 111 1 . _ zOEUEeeod er E ewuz _ Aenzz anzv Awuzm .NmznzM szuzz Nmnzz 3 0z E zzeo 1 meeeemzz . E mm.~ mm.N mo.m _ az.N Ez.m mm.N E 1--11111111111111-111411m-m1M1wmmflmMmmmm1r11mWMZ1-n1~mmzM1~mflsz1~mflmzz1d11~111 oz _ zz azeo 1 base: E E mm.m em.N zm.m E Kw.N om.N EN.m E _ _ n _ E _ _ _ _ mmz E 3qu 375 2.5 .875 :sz 5.5 E mm.N FN.N oo.v _ «o.m vm.m mo.m qm.m E IlllllIlllllllIlllllr m_.m “ Em.~ a_.m _ E E E _ . 1-111111111111111111111111111111111111111q11mmmm11W1 Wmaq11~mmmq11~mmaq1u1~1111 mw> E 02 1 mvcmnmzr E _ mm.N wo.m om.m E om.N 111.111I11111I111111r111|1111 E mmnz W1~mmzv11~msz1-Nwsz1J1~1111 mw> . oz 1 mm>z3 N E o~.m mm.~ 11111111111111111111r mv.m mm.m om. _ r1 11111111111111111 1L1 111111 1111 11111111H1 E m zoEEEeEOU =w1n11mmm211E zan Azzan zmvzz _ szqu NNNHzV zanzv E 30> _ zzzo - ascends: z _ om.m mm.N Nm.m E em.~ mo.m as.“ _ -1--.1--m-z-fl1wszMHmmmw1E--Mmmz11w1~mmzz11Nmmzz1~mflsz-4--Nmsz1wmmmzz-~mflmzz-d- was H zz zzeo - mdszz E E mm.N E~.N N mz.N mm.~ me.m E vo.m mm.m oz.m _ n1zm11~mmmz11~mmaw-11Nmflmzz-Nmflmzw1flflmmzm1u11~msz mnzw1~mmsz1~wmsz1A11~ mEuzq1~mmmzz1flmmmmM1J11~mmmw11~ .oo.m n _ _ Amuzv Em.m oo.m 1NMNmzm mz.N mm.m om.m EE12M-~Hmmzz11mflsz1w---1-- Awnzv me.~ Luzon zmacm Ezuzv mz.N -Nflflmzw11~wmzz mw.m zoznzv eo.m mnzv NNMZM11Nmsz oo.m no.m o~.m Amuzz mo.m m¢1m mm.m Amnzw-1flmsz11~wmmw ¢_.m oznzm11~wsz11 m~.m -Nwsz1 zw.m mm.m m~.m m~.m Nammw Nammaq11flmsz o¢.m om.m .Lmzon smzmw: wuaowccw mswnszc Lm304a mvcmnmzz mw>z3 documEEzuzez Am mncmnmsz mw>w3 wwuvozo An wzmcwm mcwxmum quzm awwzwnmvcmawo A< 1cou x—maowcom Ac zan Nmsz11111111111111-1111111111-111-111-111--- mucmnmzx ............. mu>Ez udebmzzzueez Em wucmamzz ww>w3 Nmmzoam qu swuem Lo wtowmn zuzzwnaccmawo A< xcEmE o» wu_uma Au E _ mmuz . zozan zezan zean Aman zmznzv zozqu _ Amqu zNEqu zoqu zo. mo.1 om.1 eo. Eo.1 o_.1 mo.- oz 1 meadows: _ oz 1 meadows: . _ No.1 00.1 E . .......... moeanmsx E _ mmnz _ Aoan sznzv Aman _ Amqu Amzan Amnzv H zoan Amzan zzzuzv oz 1 mo>z3 E oz 1 mw>_3 E E 00. cm. 00. E m0.1 N0. zo. . No.1 no. F0. ............. mo>z3 E E E E E wocwmzzzwucz nm E _ E _ m :owuwocoo E F :ozuzucoo :z E zwnz _ Azzn V Az_uzv Amuzv E AmuzM An_nzV Amuzv E Aozuzv Awnzv Azuzv 2% z—co 1 mocmnmax _ Apco 1 mvcmnmzz E _ 00.1 m0.1 00.1 E F0. mm.1 —0.1 E N_.1 No.1 00.1 ..... . mocmnmzx o zoEEEezou a _ zwnz _ Azuzv Amnzv zozan . Azuzv zzEnzv Amqu _ Amnzv Amznzv zzzsz :P zzco 1 mw>w3 E oz 1 mo>z3 E H No.1 0m. N_. E 00.1 F0.1 No.1 E m0. zo. zo. .............. mm>w3 E _ E _ H mcoquEwL mCTumo _ _ E E _ zeEEEEaeszuz zz mzoztdm co .oz Am _ E E_ a :oEoEUeOU E . Nzuz E Am" E zznzv Nman _ zzzqu zemqu Amzqu E zoqu “Euz zoan cz zzco 1 mnemomzz E oz 1 mocmnmzz ” _ zo.1 m_.1 m0.1 E mo.1 EN.1 _0. E 1111 No.1 1111 .... . . meadows: _ Nznz _ zoan zEqu mouzv _ zoznzv Eeman mozuzv E Aman Azuzv Amnzz mmz 1 mozzz ” oz 1 mo>_z _ _ -111 No.1 1111 _ No.1 00. _0.1 . No. m_. _o. ......... .. mo>_3 “w _ u _ H ” mucmmzzpouCH Am oz 0 eozeEe=0d E m eoEoEUCOU zz E Nznz . anzE Hoqu mouzv . Azmnzv szan zozan E zoan “Euzv zoan cw zzco 1 mocmnmzz E z_:o 1 muconmzz _ _ m~.1 mo.1 1111 E #0.- m_.1 #0.- H 1111 No.1 1111 .......... moconmzz _. “NTz E 8qu :z 8an _ ENTzE STE 5an _ 8qu $qu $an moz 1 mw>zz moz 1 mw>_3 E 11-- No.1 1-11 _ 00.1 00.1 «0. E 1111 No. mm. ............. mm>_3 E E _ womzoom zoom meEezEc use» zoEEEEeeedado zz Edemed me; 0;: z< ”we ozzmmoe Emz wczmeme mozooomEou mozwuazcmuz< oucoocwmw 1Lmzo mucouooEou z Locmz: Fozom Lozoz Lozmwx _m:om Lmzoz Lo:0_I zmzom Lozoz ” muzmzmozoo aodoto 3m WE Ledzz zdzzz Eeooz Adv zeE zmv AUV ADE AME A00 ADE Amy wmsoom cosh Eozme zmzvm mmaoaw cock Emzoz ”mm>HHHH0 muzzohmz mm0mN3 E oz E _ _ E _ 11111111--1-11-1111111111111111-111111111111111.. oszM11Mw1.zE NN1 .zM1J11N E zm1 z .1N MWHzM1N11H1 N1mez11n1MWHzM-NWNHzM11M E E mw>Nz 1 mocmomoz u :oNuNo28 E m :oNuNo: :N aNco 1 mo>N3 _ :N NNco 1 mm>N2 E 1-11N1NWHN11E1MmHzM- N E E _ E E _ E E E _ _ _ E 11-1-11-1-1-1-------1--1-11111-1-111-11-11mmHz11T 1N E oz mocmnmaz E lLlllIIlllIllIl E oz mocmnmsz E NN z o1.zM11N m1 .zM11Nw11zM111NW mw1z N1zM11NN1zM11Nm1zM 1 Nm E E _ .111H111E11N N N012 E OT zM11N 0N1.zEE N 24141 11111111111m1mNM1m1 mcoNNNucou :N ANzo E m :oNNNmmmm11E 1 E EN NNco 1 mo>N3 _ E11No IIIIIIIIIIIIIIIIIIDIIIL o coNuNocoo E m :oNuNocoo cN :N NNco 1 mocmnmzz E NNoo 1 mocanmaz E 1111w1m1m1wmmNNNmmmm111nM1m1N1wmmwmwmmmm :N cho 1 mo>N3 . cN NNzo 1 mo>N3 E N1 .zM NNNu zE WHzM1 NN m1.zM1_11 NW 11L1111111E1 Ean z N M1_111szv E 00 MHzM1 N 1Nm N4 MHz WHzM1N HNEM1NO £1zE N NN1.zM- N N1.zE EN w1.zM1NN EN1.zM1 N mE1.zM- No EN1 .zM11N1H1M1111N ow1.zE EN -11zM- N NMHzM- NN NN1.zE NN1.zM1NN mE1 .zM1111N EN1.zE No mN1.zM11m11No NNu .zE mN1O1zM1-m1No N1 .zE N NT zE mo. E _ E _ lllllllrl .00. n W1 .zM11N T.zM1NN ~1zM N-H-M11 IITIIIIIIII ENoT1zE NW1.zM1 N oN1zM11N m1 .zM11N we. MHzM11No Nozoo Nozom WHzM11N N1 NHzM11No NN1.zM11N1H1M11111111111-11111111111111-1111111111111 WHzM11NN N1H1M11N WNHzM11N NHzM11111111111-11111111111111-111111111111- NT .zM1111H1M onM m1.zM m1.zM aHzM N1 -zM .Eozoo Losz; aumuNocN memossc swzoM « wozoomzz ww>N3 ddemmNNNuozE No ... mocmowsz mo>N3 NooNozo an wchNw chNopm NouNm NuN 1cou aNmzoNEwm No NNEaeebEdo Nz mocmomzz ............ ma>Nz uezuoNNNdozE Nm mocmnmzz .............. mw>N3 Nomzoom Noe LouNm No oEonn NuNNNooocoooo N< NENaE op moNowa Nu ZILES .00-m H LmZOn rmscm .szou swim}; wumofiuc-r wLwDEJC LwZOJi 111m1m1m1mmmmmmmmmm11 11111111111111111111 w1NWHz1E1NmNHzM1NWNHzM11NWHz11N 11111111111 E 1 1 H1 1111 -EEN-NE.NEE-H-MEEEN1.EW1-NE-.E. ----- EEEE-H-wm-EHEEE _ E OZ 1: NEE-2 1-EEEEEEENEE ENEMNE NEE-ENE .......... 111-a1: o N a mooNNNoooo E E qm1z E Eo1zE EoN1zE Eo1zE 11 1 1 E EmN1zE NNNHzM1NWNHzM :N cho 1 mw>Nz E oz 1 mw>N3 E E Nz U H E u 1 ( 1 H wucmmNNNmucN Nm IIIIIIIII ...IIIIIIIIII I E IIIIIIIIIII II I o N o mcoNoNocoo N N :oNNNozoo E mwuz _ EEN1zE EmN1z NN1zE N \ 1 E ENN1zE NmNqu Em1zE :N NEoo 1 mozmoozz _:N ono 1 mocoomoz E E om.N mo.m No.N 1 \ 1 1 _ NN.N NN.N No.m .. mozoomo: 11111m1N1m1WNNNNNmmmmN1111111N1mmNNNmmmm1 E mm1z _ Em1zE1NNN1zE NN1zE E11qu 1111111111111 1N111w1NmHzM1NMMHzM1NWMHzM :N NEco 1 mw>NzE oN NNzo 1 mo>Nz E U NN.N NE.m wN.m E \ ,1 oN.N mm.N NN.N .............. mo>Nz _ E E E E Nmmmgw>mE E E E cumcmNum _ E E E E wocmeNwLa E E H w E NNNNNooozoooo EN ozNooEoz Em _ E E E 11 11 - - 1 1 N 1 1 1 E oo1z E Em1zE1 NN1zE 1NmHzE1N11NNHNM1NNNHzM--NNHzM-E-1NNHzM--NszM1NNNHzM 1111111111111111111111111111111111111111 E E E NN.N Eo.m ov.N E wo.m Nm.N NN.N E NN.N om.N oN.N .......... moooomoz 111-11 11111111111 1- 1N 11111 11 1- 11111 1- a oo1z E EoE1zE EN1zE NN1zE E NN1zE NmE1zE NN1zE1-n1NmHzM1-NNHzM-1NNHNM oz 1 mo>Nz E oz 1 mo>Nz E NN.N mw.m wN.m E oN.N Em.m NN.N E oo.m mo.N mN-N ............. mo>Nz E E H E _ E mocmmNNNmucN Nm 11111111 111111111111 E N E 11 111 E m :oNNNoooo zN E No1z H NN1zM NN1zE Eo1zE E NN1zE ENE1zM Em1zE E Em1zE EoN1zE NN1zM oz 1 oozoomzz E NNzo 1 mozoomoz E E mo.N Em.m 1111 _ ow.N NN.N om.o E NN.N No.m mm.N . ...... mocmoooz 1 1 -o1ooNNNoNoo N 1 1 _ zo1z E Em1zE Eo1zE Eo1zE E Eo1zE ENE1zE NN1zE . No1zE NNHzM11NNHzM :N NNco 1 mm>Nz E mwz 1 mw>Nz E om.N . . _ No m MN m E mm.N NN.m oN.m E 1111 wo.N Nm.m ............. mo>Nz mmNN me E use egos umm NNNNNnoucmamo N< manmn umNLLmz N< EmN mgammms EmN mLzmomE NcwEumm>zN mucmummsou NocoNNo>Nuoz ”awn: *Nmzo szmN: Nmucm Luzoz Lmsz: Nmzwm Nmzoz szmN: Nazcm Nwzoz EmUZmNmmzou u u m u m wmzo m cmzN Luzoz NN5 m wmzo m :mzN Nu: N: ”NzMZNmmzzN MNNoz ENE ENE NEE mozmozmmme mo mmm3mzN ENNoz .NN NzE3 E oz 1 mm>E3 E _ 1 1 \ N0. . mm>E3 E _ E E \x E wooomEEEoucE Em / E _ E E \ 111111 1111m1m1m1mmmmmmmmmW11u1111111111111111111n1mm1m1fl1NMM1mM1 NWM1aM11NM1aM1111111 N1 1 1 1 M; .zE Nmmu1z $12E ... :E AEco 1 moooowsz oz 1 moomomzz E E E N 1 E 1 ....... mocoomzz E m :oEuEmmmm11w1mm1m1111MW1zM1NMM1aM11NW1aM1111w11111111111111111111110W1ZM1~WW1zM1~mNu.zM ........: «oz 1 mo>Ez _ :E AEzo 1 mo>Ez _ E ,E .... mo>Ez Emmozw>oE E E E E ;pmcooum E E E E wozozmmmgo E _ E E E zoEEEoooooooo E< o>EooEoz Em 111-11-1111111111111111--- E oo1zE Mm1zM11~N1zM11~m1zM1111EW1zM EME1zM11EN1zM1111EW1zM11Em1zM E1111 oz 1 mucoAWJI _ oz 1 mooonmzz E E oE ZM1111111111111111111-111111111-111111-111 l111111111111111111111 E1111111111111 2416 IIIIIIIIIl1llIIIIIIIIIIEIllIllllllIllllIIlITIIIIlllIMO mocmamzr oo1z E EmE1zM11EN1zE11Em1zE111-1N1zM1EWE1zM11EN1zM1111Em1zM11EN1zM11Ew-zM oz 1 mo>Ez _ oz 1 mm>Ez E E E0. E E E _ n11M1mmEMMmmmm1mE E oo1zEE1MN12M11~N1zM11~m1zM1111EW1zM EmE1zE11EW1zM1111Em1zM1~oM1zM11«M1zM oz 1 mocunmoz E AEco 1 moooomzx _ E E ........... mocoomoz _ E ooEoEoooo 1n1mo-z1N11EW1zM11EW1zM11~m1zM11 EW1zE11mE1zM11EW1zE1111~111111EN1zE11EN1zM oz 1 mw>Ez E oE AEco 1 mw>Ez E . mw>E3 oooomEEEoooE Em IIIIIIIIIIIIIIIIIIIIIIIEIIIIIIIIIIIIIIIIII mm>Ez mEEE Eo poo oLOE pom AHEEEomocwomo E< wEoomo omELLoz E< muz husband ' Husband,z_Wife Relative competence: Husband > Wife Wife > Husband Mean marital power of: Wives 2.7T (N=10) 3.27 (N=6) l I Wives' mean power is higher on the left-hand column above,—— lower numbers should be taken to ”mean“ higher power--indicating art for power—dependence theory over the perceived competence 248 formulation. Of course, one could have much more confidence in such findings if the numbers of respondents were larger. Test D—C 3: The Role of Task Importance As discussed in Chapter VII, there are reasons for speculat- ing that both the competence and power—dependence theories ought to hold most strongly under the ”high importance" condition, i.e., when the issues at stake are perceived as important ones. Since I did not employ task—importance as a control variable, I will check out this hypothesis by comparing the results of the unweighted power measure with those obtained when the power measure was weighted according to perceived task importance. If a theory receives stronger support with the weighted than with the unweighted indicator, it is reasonable to :onclude that it can predict marital power better for important than For unimportant tasks—~since “important“ task areas are given higher Ieights. I will first review the effects which task importance have hen each theory is tested individually, and then assess the effects hen both competence and dependence are considered jointly in Table 24. When competence perceptions of each spouse were considered aparately, “dependability“ was related to marital power in the pre— cted direction (Table 4). This support disappeared, however, when e importance-weighted index was employed. Dependability ratings of th spouses, considered simultaneously, fit the competence formulation t, when importance was taken into account, the support was reduced. 3 only situation in which ratings of intelligence fit the hypothesis 249 was when wives' ratings of their husbands were considered separately (Table 4). This trend persisted when importance weightings were employed. In general, however, it appears that the competence hypoth- esis has greater explanatory power for low-importance task areas than for areas that are perceived as quite important by the spouses. Measures of dependence (alternatives and motivational invest— ment), when considered separately from competence, also show the same effect. When the perceptions of individual spouses about potential alternatives to the present marriage were related to marital power, two situations--one involving wives only and one involving husbands only-—supported the power dependence hypothesis (see Table 9). The support for wives persisted when importance was taken into account but that for husbands was reduced although not entirely eliminated. When relative alternatives were related to power, the support which was found =or the hypothesis disappeared entirely for one indicator (Question A 'n Table 10), and was greatly reduced for another (Question C). In a :hird (Question D), the hypothesis received stronger support under igh—importance conditions for wives but not for husbands. The only indicator of relative motivational investment which ad been consistent with predictions was average relative preference :rength of spouses about task-division. When importance was taken Ito consideration, the support for wives appeared to increase while [at for husbands disappeared entirely (see Table l3). In summary, it appears that power—dependence theory holds re strongly under high—importance conditions for wives and under v—importance conditions for husbands. This is consistent with 250 ower-dependence findings in which the dependent variable was satisfaction ather than power (see Tables 21 and 22), where support was somewhat tronger for wives, especially in Table 21. Two explanations for this eem to be reasonable. First, it may be that variations in dependence ong women are greater than among the men in my particular sample of uples. Or, perhaps——as suggested in Test D—C 4 in Chapter VII——women swered the task importance questions in terms of their own personal als, while men interpreted them in terms of goals for the couple as a it. When Table 24 is considered, the difference between unweighted er and power weighted by task importance can be compared by contrast- ; the support or non-support patterns summarized in the far right—hand lumn of the respective sub-tables. The pattern of support or non- )port for the competence hypothesis, with respect to the competence dependability, usually shows no change for husbands (in four of the tests) and decreases for wives in half of the tests. The shifts erved, in terms of gross changes from support to non-support, are narized below: Number of tests in which support for the competence hypothesis is: Wives Husbands Increased by importance weighting 1 _ Not changed 2 4 Decreased by importance— 2 weighting 3 251 r the competence of perceived intelligence, there is little change en importance weights are added; few test results are consistent th the hypothesis in either case. Similar comparisons for power-dependence theory, between eighted and importance-weighted measures in Table 24, can also be e. The patterns of gross shifts from support to non—support or e-versa are as follows: Number of tests in which support for the power—dependence hypothesis is: Wives Husbands Increased by importance- weighting 4 — Not changed 4 7 Decreased by importance— weighting 3 4 ll ll In general, when dependence tests are controlled for relative etence as in Table 24, the previously noted tendency for the predic— ; to receive increased support under high-importance conditions for is found again. When competence tests are controlled for relative dence, the tendency observed before, namely that support was ed under high-importance conditions, is present to some extent. In conclusion, the role of task importance as a scope condi— ior theories of marital power appears to be a significant one. er, the effect of importance was not always in accordance with *xpected. The competence formulation, as adapted from the work ger et al. on status characteristics and expectation states 252 966, l972), had been expected to hold most strongly under conditions high task importance. When such task importance was high, I soned, competence would become a factor because there would be more ious consequences if task performance were done poorly; the group ld be more "task oriented" and thus somewhat similar to the task :used laboratory groups used in previous research. There are at st three problems with this line of reasoning. First, it may well d if one were to use the type of power measures that Blood and fe (l960) did, in which the person who actually takes more responsi— ity is considered to have more power. Second, the kinds of task is used for the present study may all_be relatively unimportant to : couples, relative to other joint concerns which they have. lly, it appears that--at least for wives--the rating of a task as ly important did not necessarily mean that they considered it ly consequential for the couple as a group. The last reason may account for the finding that power-dependence theory was supported strongly for wives under high-importance conditions. Tests Involving Specific Resources My attempt to find a way to test the theoretical process of l power separately from particular culturally—specific situa- was not entirely successful. Finding support for some indi— but not for others is a reminder, not only that my measurements efining, but also that such theoretical process—concepts as tional investment and competence cannot be measured in the t. There apparently are different norms for husbands and 253 ves with respect to what kinds of "competence” can earn one influence .g., wives yield to "intelligent“ husbands but not vice-versa). To e another example, potential alternatives to the present marriage reduce dependence in societies where divorce is a possibility but ld not be expected to do so in other societies. I was, however, successful in my attempt to test theories marital power in a manner which allows for potential falsifiability. approach of previous researchers in the “Resource Theory“ tradi— 1 failed in this respect; the "resource theory“ could never be acted, because any negative finding with respect to one "resource" d always be explained away as being due to some other "resource.“ In spite of my desire to avoid the use of specific “resource“ ables, there were a few cases where alternative theoretical formu- Jns seemed to clearly predict different results from possession given resource. The presentation of these tests constitutes the ’nder of this chapter. etence and Gratitude versus eDependence Theories -CG 2: Relative Income and Wives' tional Investment in Having a + Wives' Overall Marital Power Income is a ”resource” which has been found to be consis— related to marital power in Western, industrialized societies. ment of wives has been found to increase their marital power in ndustrialized and less industrialized countries. (See Figure 3 ummary of some relevant findings with regard to income and work participation.) 254 It appears that a good deal of the marital power of employed is associated with their income. As was already mentioned under -GC 3 above, the relation between wives' income (relative to f their husbands) and wives' relative marital power, by itself, at support any one marital power theory. This income could earn fe gratitude and respect from the husband (the gratitude theory). >erience of working could increase the wife's competence in areas It to domestic matters and/or the husband's estimate of her com— (the competence theory). Finally, having an outside source of (”alternatives") could decrease her dependence upon the husband dependence theory). It may well be that all three processes simultaneously to raise the marital power of employed wives :hat usually held by non-employed women. Since my concern was to assess the relative explanatory power al power theories, findings about the relation between income r alone (see Table 25) are not helpful. I searched for a test n in which different predictions would result from the different aories. Although power—dependence theory would predict that ise in relative income should be associated with an increase in power (because of better alternative sources of resources he marriage), a contradictory prediction could result (if the such ”alternatives” were offset by high “motivational invest- ie situation in question is one where the wife values having ior its own sake. Assuming that most U. S. husbands still I implicit ”veto power” over the wife's decision to work, 255 wife’s career is a goal which is "mediated” by her husband. She ependent upon him for ”permission" to pursue this career. If that assumption still holds, power-dependence theory d predict that wives with high “career investment" will “lose” of the power which they would otherwise have gained from having ternative source of income. The competence and gratitude theories predict that the relation between income and power remain unaf- d by such attitudes, since variations in motivational investment .ving a career should not cause changes in either the wife's per- d competence or in her income contribution for which gratitude e earned. Test D—CG 2, in summary, hypothesizes that the relationship an wives' relative income and their overall relative marital will be reduced when controlled for wives' career investment. mpetence and gratitude theories, in contrast, would predict that investment have no effect upon the size of the relationship relative income and power. Results of this test are found in Table 26. By comparing ference between higher and lower relative income conditions in 6, with the difference between higher and lower relative income e 25, one can determine which hypothesis is supported. The nce between wives' mean power in higher and lower relative onditions is .27, without the use of the control variable (see ). If the corresponding mean differences are noticeably smaller 256 Table 25 OVERALL RELATIVE MARITAL POWER OF WIVES, BY RELATIVE INCOME OF SPOUSES. IGHTED POWER MEASURE Relative income of spouses (l) (2) Other: Husband makes much Husband makes somewhat more income than more income, equal in— does wife (at least comes, and wife makes twice as much) more income marital “* of: ‘ves (N=83) 3.36 3.09 ' (N=22) (N=6l) numbers indicate higher power. Equal power = 3.00. 2557 umuLoanzm agoocu Lozurwz . Amw wreak zuwz umgnnsou . AFFHZV Amuzv uuwwmw muvmonaov oz _ Fm.m vmugoaazm acowcu wucuuwa IEOU\wu:pwumgw Amw agaveoweev empowewg one :_ ucmspmw>zr quLMU do powyww mH _ j “ nguo wLoE gone ucmamsz vcmnmzz cozy oguxm ecu was» I! cwup>oc szumL vpaoz H "ou.mcmzm: w—nmh cum: uwcmaeouv uuwvmw oz AmFuzv Amuzv wzowuuwgwu kuuvaL mwgm A m wcu cw uzwsumw>cw A < _ :ogum Lo wwcmmc uoc wwwz E was» m .mwmzonm cwmzuwn mocwxmawwm Remnzv AN_nzv Am_nzv Amuzv mo.m Nm.m mo.m wm_3 :mzp weoucw wwwz cusp mEouc_ Lozuo wLoE nous ucmnmax cwguo n m wees cuss vamamzz nmzoqu m>Hhem mumsoam ucmnmzs cusp z_m:ogum “mumsomm zmm3Hmm muzmmmmmHo oboe wwwgmmmwu wwwz .m mm3mzH mmwmzu mmumaz no: Locum; er: .nmJJthzou wH mmmmHk<4m m cwcuo wLos cuss nemamzr szuo wees cone vcmnmsz “mzoqu m>~p<4mm n m Ammwcmmmwo >_mcocumv as mguxm ms» umcp canw>oc o: H ”op mw>w3 we wmcoammm .< mm=mzH mummzH qHkoz .mm>~3 zuxz >m mm>H2 no mmzom 4~k<4mm 44o "N cola wwmp mN mFQmF 2558 Lows» szpwwz sz m_nmp sow: vegansoo :owuowcwv wuvmoaaov oz acozboazwe umpozewe as» c? «cwsumo>:m gwmgmo do poweww mm .uwpgoaasm zcowcp szuwwz Hw>onm u :o_umw:o Low F azocm ;pwz umLMQEouv oz ~co_powc_e napozuag wzu :w “cmsuww>:w cwwcmu vo puwmmm mz .umugoaazm agowzu szumwz zmw a_ask cum; ewcaasoo .cowuomcmu wuwmonaov oz ~=owuumgwu vaUWqu wzu :w acmEumm>cw cmmcmu do uumwwm mm ze_uzv znmuzv wo.m ez.m mew: can“ wEoucv wwwz can» mEoucv wLoE zone vcmnmzz mLoE :uzs nzmnmzz m Acoryznsa 3o_ Lo Ezwumzv AcowuwnEm ;m_; zcw> Lo carry 304 saw: .==L :o~ wzu Lm>o corp? ~w>m_ Lao wNwquom new ;wwcmu we Lego wmmw_a ”op mechm .mw>wz .m mzzmzH mumm N p .Aw>onm u :owumwso op szmcm wwzu w>mm 0:3 mw>wz mo »_:o umxmmv :cos mm upczuco . o cwmcmu seam a>me a“ u; :0 cwzoz= co “Casamtoecw to somzwtpm .o mm3mzH mmmmzu z,_nzv Amuzv Ameuzv ze_uzv w,.m wo.m No.m mm.m may: :mcu weoucw wees cogs vcmnmzz m mmwz cusp waoucw wLoE zone neonwzr a A Mac: op pa: uzmzo :weoz um use a J203 on w—nm on ou uzmso :mEon 3o; Acwe mm aumczugoaao chm o>mz cu ucmzo :mEon :mwz .oo.m u cwzoa _c:cm .szon cmcmw; wumuwncw mcwn23: Lugoza ”HZmszm>zH mummzz “do *LmZOm quzcws :mwz .mZOUZH m>~hzH mmmzwz ”we «cwzom Fmpwcms cum: umzoqu m>~bzH zmmmw employed, please characterize your level of career ambition ong run.” Respondents could check one of five options: very , medium or average, pretty low——no particular long-run 264 ons, or I don't plan to work. Those answering “high“ or “very (group I in Table 26) were considered to have high career in- it, while the remaining women were considered to have low career nent (group 2). As can be seen below, results clearly do not : the power—dependence hypothesis. SURE E. Level of wife's career ambition over long-run. l) 2 High or Medium Theory Very High or Low Supported nce of power— etween lower ier relative- ; | I groups: .66 .06 Neither one In summary, there is a striking trend in favor of wives with .ivational investment in having a career; such women accrue e marital power from a favorable relative-income position than with little career investment. This is contrary to the from such earlier studies as that of Hoffman (l960). Such strong findings, I posit, would probably not disappear est were refined (for example, by eliminating the few unemployed m the sample, by making finer distinctions with respect to re— come, or even by controlling for whether the wife worked full- art—time). 4y assumption that wives feel that their husband's ”permission” in order for them to have a career, simply may not hold for er—middle class respondents. This possibility should be checked ‘ research. Another possibility, of course, is that high moti— nvestment in goals mediated by the other spouse does not 265 crease dependence in a marital situation——or that this kind of pendence does not lower marital power. Although power-dependence theory has received some support the present study for the hypothesis that ”alternatives” should be rectly related to marital power, it has received almost no support “ the hypothesis that ”motivational investment” should be inversely ated to power. Either the theory requires some change—~at least a cification of the conditions under which it does and does not hold—— I have failed to measure the kinds of ”motivational investment” h are the strongest sources of marital dependence. This promises e a fruitful avenue for further inquiry. i The final test~situation involving individual resource-variables )involves power-dependence theory, this time in specific contrast I the perceived competence formulation. etence versus Power—Dependence Theories C—D 4: Relative Age, Marital Experience, and Having Children + Overall Marital Power The final three ”critical" tests deal with three “resources" :iated with certain kinds of past experiences, the experience simply comes with having lived longer and the experiences associated having raised children and having been previously married. All of are potentially learning experiences, which should——more often iot—-increase the competence of a person in a variety of areas. )mpetence theory would predict, then, that a spouse with more of a TEST C-D 4: Table 27 OVERALL RELATIVE MARITAL POWER, BY RELATIVE AGE AND PREVIOUS MARITAL EXPERIENCE AND PREVIOUS EXPERIENCE WITH CHILDREN. 266 'HTED POWER MEASURE A. Relative Age (l) (2) (3) Younger Than Same Age Older Than Which Theory Is Spous As Spouse Spouse Supported? arital power* of: :bands (N=86) 2.67 2.62 2.93 (N=ll) (N=26) (N=49) Pow r d d e - e en en e es (N=86) 3.02 3.38 3.33 p C (N=49) (N=26) (N=ll) B. Relative marital experience (I) (2) (3) Spouse Married Both Spouses Self Married Be— Which Theory Is Before, Self Not The Same fore, Spouse Not Supported? ital ower* of: ands (N=87) 2.79 2.84 2.82 (N=23) (N=43) (N=21) . Neither one s (N=87) 3.l8 3.12 3.21 (N=21) (N=43) (N=23) C. Relative experience with children (I) (2) (3) Spouse Has Children, Neither Has Self Has Child- Which Theory Is Self Not Children ren, Spouse Not Supported? :al ower* of: nds (N=53) 2.73 ( . ) 2.83) (N=25) N=lZ (N=16 ' Power-de endence (N=53) 3.17 3.15 3.27 3 " (N=l6) (N=12) (N=25) nbers indicate lower power. Equal power = 3.00. 267 :ular experience should be perceived as the more-competent and, therefore, should have the greater relative power. These three ”resources“ were selected because they may also ative ”status characteristics,“ i.e., traits which reduce one's 3n the “marriage market” and hence reduce one's potential alterna— :o the present spouse. Power-dependence theory would predict eing older, being divorced, and having children will all be ,ted with lower marital power. The test which follows relates the relative position of spouses e three variables to their overall relative marital power. Com— theory would predict that spouses with greater relative "experience“ ion 3 in Table 27) will have the highest power; power—dependence reinterpreting the same variables, would predict that spouses Iher relative "alternatives” (condition I) will have the highest In short, the two theories predict opposite trends. Ages of spouses were obtained from the marriage license appli— and were coded as follows: 22—23 years, 24—25, 26-27, 28—29, 2-33, 34-35, and 36—40. Codes for wife and husband were then If they fell into the same age-category, they were considered equal age (condition 2 in Table 27). Other respondents were ther in the ”older than spouse“ group (condition 3) or the than spouse“ group (condition l). As Table 27 reveals, results are generally in accord with endence predictions (although younger husbands have slightly Ier than equal-age ones and older wives have slightly higher 268 r than equal-age wives). Husbands apparently "lose” power only hey are older, while wives ”lose" if they are either the same age lder. The assessment of such gender differences could be improved )ntrolling for age level; for example, the advantage of being [er than one's husband may decrease in older age brackets. Also, a larger sample it would be desirable to determine the effects all versus large age differences; the ”older" wives in Table 27 sually only slightly older than their husbands, while some of the r“ husbands are considerably older than their wives. The second part of Table 27 deals with relative marital ience. All respondents in the present study were either beginning :t marriage or else they had been married once before and then .ed (or, in two cases, had an annulled marriage). A divorced married to another divorced person was considered to have an amount of previous marital experience. In about one—fourth of nple, the husband but not the wife had been married before and ther one—fourth, only the wife had been previously married. :ouples were placed into groups (l) and (3) in Table 27, depend— »n whether the individual spouse had more or less marital nce. Table 27 indicates that there are almost no differences in power among couples who vary in relative marital experience, so theory can be said to receive support. When the wife but not >and has been married before, husbands have a ygpy slight power Ie relative to other husbands and wives have an equally small sadvantage relative to other wives. 269 The final test deals with whether or not spouses have ildren from a previous marriage. Couples in which neither had an married before are not included in the table for this parti- lar test. The couples in group (2), where neither has children, a similar to the couples in groups (I) and (3) in that at least : spouSe has been married before. Also omitted from the table are u couples in which both had children and one couple which has ldren born some time before the wedding. The results with respect to children in Table 27 generally Jort the power—dependence hypothesis. As with relative age, rela— 2 experience with children exhibits differences according to gender. ‘e the wife has children but the husband does not, husbands have er marital power than other husbands and wives have lower marital r than other wives. The reverse situation is a different matter, ver; if the husband but not the wife has children, power scores iot different from those of couples where neither spouse has iren. Part of this gender difference is undoubtedly due to the that only about one-third of the husbands but almost all of the '24 who had children, had their children living with them. Having 1y of one's children, rather than simply having children, may turn > be the real disadvantage with respect to one's alternatives on :marriage market or—-for that matter-—for one's alternatives of '24Only one of the 27 wives who had children did not have her an living with her. .270 able to support oneself outside of marriage. This finding also ponds with Blood and Wolfe's couples, where the average power of was the lowest of all stages of the family life cycle for the in which there were young children. To summarize the results of Test C—D4, the data for relative 1 relative experience with children generally correspond to lependence predictions. Neither theory receives support with : to relative marital experience of spouses. CHAPTER XII SUMMARY AND CONCLUSIONS This report analyzes one major research program within family iology, that focusing on the “Resource Theory“ of marital power. implicit cognitive processes by which resource possession is assumed ue “translated“ into marital power were identified, centering around concepts of perceived competence and gratitude—and-respect. These ‘ls were contrasted with a third theoretical process suggested by ics of the Resource Theory, that of power—dependence theory. A case was made that the Resource Theory research program has hed an impasse, where each additional study only increases the number antradictory and inconsistent findings. 0n the hunch that this ;sé is at least partially due to the underlying assumptions of the ‘etical reasoning, the job of explicating and testing the assumptions IIS line of reasoning was undertaken. These assumptions were then juxtaposed in an exploratory way st ideas from alternative theoretical traditions. I attempted to en the perspective beyond that of ”classical" family sociology by 19 upon seemingly unrelated strings of theory development (e.g., ‘ et al., 1966, and Dahlstrbm, 1966). By so doing, I was able to ‘p a set of hypotheses for each of the theoretical processes under eration. 27l 272 Hypotheses were tested on data from a sample of 90 upper— Iiddle class, urban, newlywed Michigan couples ranging in age from ‘2 to 40 and including couples in both first and second marriages. Toth wife and husband were interviewed, with gender of interviewers etched with that of respondents. Couples were interviewed at two oints in time, by telephone shortly after their wedding and in their omes several months thereafter. In addition to the telephone and ersonal interviews, data were also obtained by a return—postcard od during the many telephone calls made to solicit participation. artly because of such efforts, response rates for the second inter— iews were very high. One benefit of the longitudinal research design was that it lowed identification of the role preferences about task division th which respondents had begun the marriage. These preferences re compared with later reports about the actual division of tasks. e average degree to which a respondent's initial preferences were iieved was taken as the measure of marital power, using only those ems on which husband and wife had begun marriage with different eas (see Figure l2 below). This approach to the measurement of 'ital power eliminates many of the shortcomings of the decision- ing indices which have traditionally been used. The 23 tables containing the results of hypothesis tests for ; sample of respondents do not lend themselves easily to a "quick” iary. However, there are some central points which deserve mention. 273 .00. n szba _m:cm .szoa szmw: mumuwvcw mcmnszc emzoza zezuzv Ammnzv Ammuzv zzw u zv mm> N0.: me. ~_. mw>w3 zemw=__2>tsov zo_uzv Amsnzv z_muzv zmw " 2v oz 00.- _o.- w_.- meeaamzx mcowuumgwo “cmpgomem acmuLomEH uceugomEH ”mo ango vaowvwgm mg“ __m an “02 z—mecwxm cw ucwgp mH Lo umczweom D zmv AWE A22 :owcwpvcu :owuuw_wmnmums a mo wocm mucmm___mu:H “so EH um wocwumneou Acmwcz_r>gzuv Amuzv zewuzv Aemnzv Azmuzv 02 No. Np. mo. ww>w3 Axwwcrpp>ssov Am—uzv Ameuzv Apmuzv Anm n zv oz wo.u v0.1 mw.- mucmnmsz Ncowuuwcwo «cmueomsz ucmucomaz acmuLomEH ”do acwzo ~muwame :mwz embozewea mg“ __m be “oz sFaEmcpxm cw ucmsp mH Lo uwzzwsom A8 E A: :owLwHwLU cowuuwfiwmlmumE m mo wucmuLo EH Nuw_womucwmmo n< wocwuwanu wthszmz muzwum_mc wczmmus op umm: nozums an“ we :owpmcym:~_H .N_ teamed .oo.m u gmzoa pmzcm .Lmzoa smzmw; wuouzucw mcwnszc xmzoza ze_nzv Ammuzv Ammnzv zum u 22 was em.w mo.m om.m mw>w3 Atma=,_z>c=ov zozuzv zmeuzv z_m"zv oz mm.N mm.~ me.N mcowuowcwo pcmpcomez ucmygomem ucmucomez empowewea as» __m be poz azaEatbxu cw tawny mH Lo umszweom zmv zmv A—V :OwquwLo :owuuwawiwumE m we wocw wucwmw__mu:H “so EH ”m wucwumneou zmuzv zewnzv zemuzv zmm " zv as» o_.m __.m mp.m ma>23 zm_nzv Amsuzv z_mnzv zem u 22 mw>z em.~ _~.N no.m mecmnmsx Ncowuowcwo “:epcomEH “:oucomEH acmucomsz ewuozeaea web PPM be Caz s_wsaebxm cw ucwek mH Lo pmzzmeom Amy zmv z_v N mm2m ugmzczoo uvzwnoe mzwmumuzmzovum :uuo census 'Ad1 wmaonm case Lozmw: cm IIIIIIIIIIIIIIIIIIIIIII 4<3ku< mmaonm .H 020.— ”M JBMfld named mez~o222 zz_:o “mm mwmozponz; umm— :wv mw_a:ou m meLmE ww> unmzczoa wmzonw :mzu szmw: :mzu meod 1. wucmuw EOU w>~um_.wx mmzoam \ if I _o JBMOd omkufiomma umozu ou m>wum_wg prsme \ :30 mo moucaomwc-maumum mw>zz mucmnmzz cmzoa _muwcme m>_um~mz «macaw mo Luzon —mqumE m>wumfimm muzwuwasou w>wumzwz mmHzOmzh zuzom 4x H hm<¢ 22'759 -xmmu mo cu :mLum m>wumpum posau Lo ;wmcoxum um wmzoam m. M ifiTfi”7-T_' mom mama so wpamu wow .zLLuE ou mxT— u~zoz uacu :owmwuuu we mcwemu w>wuapwm Au .mmzomm voom mcwvcve mg: 5.53 mm; cmucwm 5.23 2 map omen co m_ouu wwm ww~ wmma co ange“ wwm u w_LLmE mvzu op mu>mu -ucxmufim pawucwuo w>rumfimx smgaw; szmw; m—wm wwsoam lillilillolllllllnilell m M01 W5IH wkzmzzou 4<=ku< JBMOd JBMOd 1H _~_ ”01 JaMOd z Azpco pom.mew:uoazz umm_ :_V mw~qaoo u/ vszumw>cp .iliulilsiiluuiiuuuinllll chowum>wuoe w>wuu~wm caem_e Laemwz w_mm omsoam IIIIIIIIIIIIIIIIIIIIIIIJ .I 0 M JBMOd UBIH mm>wumcgwu~m w>wumpw¢ Luzmw: gm:m_z mpwm umzonm II 3u_)<:J: . mm>wz u - - . mucmnmsn u xwzoa —ou+LmE m>wmewz,/, ucwEumwscw pmcowum>_uoe a>zba_mm Luzon Fouzgae aszaa_wm mm>wumcgwu_m w>zba_wm av / ¢w wa ,x/mmM/w "nzmumz 33N30N3d30'U3M0d 280 :oEEELfiS 3......»me cw aucmawgumwuv acoE uuwgmmmwv mo uNwm wmmcw>< AN .Ammocmcmmwga :owmw>*u uxmmu :mV «cmewmgmmmwu mo mmwgm mo Lwazzz AF mumwcwucw M01 oNN woma co wpnmu mum mhszZOU 4<3Fu< asn JBMOd mozzozzm zx_co uwm mwmmzuoaz: umm_.cwv mm_a=oo u mm>_z u - - - - ,. mucmnmaz u ”ozmwm4 mummgmucw wammmm .mwmzo m mo wucw gw>_o cwzumz ummwucwxw ;m_: 304 szoa mo uczoe< / may» lonllllllllilllillllllllb z .x m a a W Mm was Mw NIQ IA) /\ .W 9.07. mw AmmammW «w m mpnmu—as wmoeumv MW mumewacw m_wm:oam mo wocomgw>wo ompuzommm mzmm1F0m>z .uwzcvucou .H EDNEONHdSO'HBMOd Hmzz »_co mm>w= n I I I I I I I mucmnmzz u "mzmomz IIJ d 0 M IIIIIIIIIIIIIIIII a a .Laguo AN :owuwecou H .wmmz mm meoucw .flw guns mu wu_zu «umm— um waxes vegans: AF :owu_ucou zozhuzamma uo2h2h<¢c uz~.ame=a we u>.um_ae Laea_; eases; Lazmwz Lazar; upwm mmaomw m—wm wmaonm vaFoLucou IIIIIIIIIIIIIIIIIIIIIIIJ m. "wI mam mmucaommc M m a,n_m=~p a>_u~_am ”2mm: \ l I \ cwzoa _muTLmE iHI\I % M 3322/ an\\\\ J J oNN «man :0 wpnmu uwm R I\\\I|\\\~ ‘ Cad zuemu_cwcmwm . 38.52:: 8: r m .H. 25.5 353 335 .W .4 \ Ion»: ov:u_uugu saw: wocwumnEou \\\ ucwumwmcou macruzru onHoHomzm mozmpmzzoo w>wuu_wz III mhzmzon z<=hu< amHuHemzm mummzkom>z mozzozzm .mumwmooxm zHhquFom Au “mm mmmn :o w_nmu wow .wmaoam uoom m m:_ucwm meme wammw mm; qucwm :u_:3 A< wmmpLLME Illllllllllllllllll mvzu cu mw>_umcgmu_m _~_u=wuo w>_um_az Luzowg Lmzmw; m—wm wmzoum 2| \\‘I(_m ‘rlIPe {no \ xx 1530 x \ ‘QI \ h .c A:o_uuwvmca « wv3u_umgo\mo:wuwnsou {J a. I» o X a: = o .— u u .— u an L a USIH “:uumwmcou mucmch mhzmzZQo 4<3hu< zanod .muzmgu _mmemm mEmm wzu vuuwnwzxm mucmamzz .covpmquanucw we mmmw Low .muwIn umwu Low mocwccwm cw umuafiu:_ mew mw>w3 z_:o mm>w3 u I I I I I I mucmnmaz u ”QZmaMJ mm>wumchuFM III meucmuo w>wumzom Lmzm_z szmw; m_mm mmzonm IIIIIIIIIIIIIIIIIIIIII41 _m d m lIIIIIIIIIIIIIIII a I mm U- zozhuHamzm mozhzhwumcgmupm _m_ucmuo w>wum_wz Lmzmw; memw; . :1 U2. 3.5.50 m mam mmuczommc w_nwocmu w>mum_wm uzmzz d cwzon Puuwcme m m>_um_oz a J ,zmou-o mm :m U- mw>wumccwu~mx onHUHQWmm muzmozmmmoIzmzoa w>wum_mm IIL ompuHamma n~nut.DL_c BGOIIIVUO/EONEIEdNOO ’SA 33N30N3d30'H3M0d 283 emm woma co upnuu wow mu” mama an anau mow .cowuumvusa szurw su_: «:ounzmcou uoz 3sz8 I! ucueumw>cw Luwgmu m.ww_= saw: 304 A 1 0 o L.M T.» w B .4 m a” . 3 D. . "m. _ -..... NN.+ IIIIIIIIIIIIIII m u _ ma .=owuwnsm megmu e m w cssImco— fat»: 3 9 WM 9 H19. .1. I. .953 0.3 ..U M murmuzo mzwzgoz :oon _ uzonm comcwao n.0wwz «a + u ZOHPuNnumm moakupcm Lawtmu m.mw_z “cprmm>=z gmwcuo m.wewz 5:. :3 =3: :3 d s W m... ..m m x “a m.» mm~.+ :mzu mmu_H “ m.» \ a m I a m n x \ ww ww \ x. _ "w _ "M \\ x I. 1:8 I93 .I \ ~ my. _ mp. z x H mm . mm +9. \ 9 ...IP 9 1”! a \\ m. In. CN.+ :2» 295 m. In... N m u M . _ onHoHommm muzwozmamnImmzoa 4<3Hu< .mmzm muzaowmmI mucmamzz n :owuwmoa wEoochw>wum_wL mw>_: I - I I szoq m.wmw3 / acmsumw>cv qucmu m.mwwz uHmHUMAw quNNAHH: mmHmONz» m>Hhz .... unzmomz BOnlIlVHS/HDNHIHdNOO 'SA 33N30N3d30'd3M0d 0mm mama :o wfinmu mwm .comuuquLn szumw :u_z ucwumrmcou uoz 2E311 wow woma co wpnmu wow .=0wgupvwga aucmvcuawu Ieazon ou somopu mhzwv—zou ouzowewaxm -u_gme w>wum—wz Am wrlllilllln ”Ill/(l. flamercgme maow>ogn Eogm :wLu_rzu w>~z Lmzuuzzv :mLu—wzu saw: wucuwgwaxa a>_u~pwz Au mum w>wu~_wm z< wLOE ULOE C a wmzoam “WI .m < \I?: will lf|~ ‘IIIUII all! 1‘ 0 < . H JaMDd JBMOd mougzommx uwunpchwucmwco xu w>wparwa Lmzmwz Cesar: _mm wmzoam ._II o M Id 0 M 3 J mm "- zohpanmzm muzmhmmzou mw>wumcuwu~< a>wum_mx Luzmm: szmw: wmzoaw wrwm IIIIIIIIIIIIIIIIIIIIIIJ zl m d m a J H onhuHammq mquGmeunImmzom mmFuHouxm .mmzwucwm umuuwpmm $0 acmeeam mu>w3 u I I I muzmnmzs u 533 $3235 a>_e~zam///////// A vqu Acmguzwco .ommPLLme maow>oca .wmmv mwos30mwc w>wum_um I m~wmzh0m>z .m_ massed "ozmamq 33N313dN03 'SA 33N30N3d30‘83M0d 285 It may be, too, that one theoretical model describes ilies better during the earlier stages of the family life cycle ile another is more applicable at later stages. Since the exchange mitment between spouses is a long—term one, one may have to study ger time periods than the several months encompassed by my study, order to adequately evaluate the applicability of exchange— oretical process models for describing families. Equitable trade- 5 may occur, for example, in which a benefit received may be repaid hs or years later, a reminder to methodologists that "longitudinal” ot necessarily equal to just more than one point in time. Finally, one promising strategy for further research would :he identification of those couples which best ”fit“ the power— ndence and perceived competence models, respectively. If these out to be different (non—overlapping) groups, an examination of common characteristics within each group and of the characteristics 1 differentiate the two groups might prove very useful. Such an )ration of the conditions under which each theory "operates”——in of such things as stages of the family life cycle, normative fs about the roles of men and women, career orientation of hus— and wives, the larger context of social and economic circum- es, etc—~would constitute another step toward the goal of being to describe and explain marital power processes within families. I fairly clear idea of how to identify couples for whom power— lence, perceived competence, or some other models are applicable, 286 further step would be possible: research aimed at assessing the sitive and negative consequences of such processes, both in terms the life—quality of the individuals involved and in terms of the lications for the communities and societies in which these ilies live. E QUESTIONS OF FACT ..... Your confidential identification no. Please check (V) the o_ne best answer to each question. .“ During your _fi_rs_t_ three months of marriage: a. Where will you be living? C] Wayne County; DOakland County; I: Macomb Co.; I] Washtenaw Co.; E] Ingham Co.; [:1 Eaton Co.; E] Kent Co. B outside of southern Michigan D other county in southern Michigan D. How many hours will each of you be employed outside the home? SISE’SM’H DIIII'“21.:I.§,I:I'“"* CI} .ifihtm‘. ‘3... c. Will any children be living with you (includEi'ngp younger brothers & esisters?) [:I yes CI no Has either of you attended a college or university for two or more years? HUSBAND: (:1 yes C] no WIFE: C] ye s C] no 0 you definitely plan to have a baby (or adopt a child) within the next l-l/2 years? [3 yes [I no F EITHER OF YOU IS EMPLOYED OUTSIDE THE HOME: Is your present occupation (as listed on the marriage license form) a job or the beginning of a long-term career? HUSBANDl—l ':j'o bjob career career In your present occupation, which gig activity takes the most time? Reading & Writing (composing Working with your hands Dealing with people letters, programming, math (typing, playing piano, (selling, teaching, calculati 1110 k ypunching) tc.) ID (check one): D [I 'check one): :1 El I: D this a first marriage for both of you? 1:! yes (90 to #6 below); I: no NO-a Did the previous marriage end in divorce HUSBAND. De syes So no (rather than annulment or widowhood)? E: By [111 Are there any children? ........ ggggmo'a Dye e::yeSE]':n" no ional questions (if spouses differ, check 2 boxes): Religious backgroun nd: E] Jewish, Orthodox; (:l Jew is,h other [:1 Catholic traditional; D Catholic 'modern' C] Orthodox 1:] Protestant fundamental; Protestant other ttl Other religion [:1 0 religion Race/ethnicity: El Spanish- -Amer1‘can [jEn tother White (:1 Black (:I ON (:1 Native American (Indian) ou would like a report on the results of our study, where should we send it? Addre e can we call you for a simple telephone interview? Phone No. are good times to call? (example: Mon., Tues., KI Weds. from 7—l0 PM; Sunday afternoon) THANK YOU VERY MUCH I ! First Class Permit #359 E. Lansing, Michigan BUSINESS REPLY MAIL . No postage stamp necessary if mailed in the United States ; Postage will be paid by: E. BENSON DEPARTMENT OF SOCIOLOGY MICHIGAN STATE UNIVERSITY EAST LANSING, MICHIGAN 48824 MSU Acct. l l—3742 APPENDICES APPENDIX A LETTERS SENT TO COUPLES 288 IICHIGAN STATE UNIVERSITY East Lansing, Michigan 48823 EI‘PAR'I'MENT OF SOCIOLOGY . BERKEY HALL ear Ire at Michigan State University, we're beginning a study about the early ages of marriage. You are among several hundred couples chosen from the blic marriage license records, and we're asking for a few minutes of your me. Won't you plgpse take 5 minutes and fill out the enclosed postcard? Y THE POSTCARD QUESTIONS? Although many studies have been done about teen- 2 marriages, much less is known about people who marry in their 20's and 's. For example, we lack very basic facts about the situations in which Iples begin marriage. Yet these situations make a great deal of difference—— en the decisions which have to be made are different, depending upon whether 2 or both people are working, whether major career decisions have been ,tled yet, whether there will be children to plan for, and so forth. RE IS ALSO A SECOND PART OF THE STUDY, involving 30+minute telephone erviews with each of you at your convenience. Of course, this is en- ely voluntary——if you gpg interested in being interviewed, please fill your phone number and preferred times at the bottom of the postcard. interview is both brief and interesting, mostly about decisions during early stages of marriage-~how soon you expect certain types of decisions :ome up, which decisions you see as the really important ones in your own lation, and so forth. You would pp: be asked personal questions about * income, intimate matters, or the like. I do hope you'll take the time to return the postcard very soon-- by , if possible. Even if you are not interested in be- ing interviewed by phone, please fill out the rest of the card-- this information will help us find out which are the most and least common situations. In return for this help, we'd very much like to send you a report ebout what we learned as a token of appreciation. Included will 1e some interesting comparisons of first and second marriages. Iurse, gll_information you may give us is kept in the STRICTEST DENCE. Your answers will be placed together with those of many other es. Reports of the study will give information only about groups of es, in order that no individual or couples could ever be identified. you please let us hear from you soon? We look forward to your reply! Very truly yours, Elizabeth E. Benson Study Director (5l7) 332-0924 289 [GAN STATE UNIVERSITY East Lansing, Michigan 48823 I'MENT OF SOCIOLOGY . BERKEY HALL IOU were telephoned this spring for the Michigan State University marriage I, both of you indicated interest in a second interview——if a convenient :ould be arranged. That's why I'm writing—-to ask you when it would be Iient and also to tell you more about our study. It promises to be an sting study, involving couples from seven counties, and we hope you'll be 0 participate again. COND INTERVIEWS will be done this summer. Each couple who decides to parti— will be personally visited in their home by a man and woman from our staff, rearranged time. It will take no more than an hour. The questions deal )me of the same subjects as the telephone interview--in particular, with )n—making during the first few months of marriage. I think you'll find our iewers friendly and informal and the questions enjoyable, with more oppor— for expressing your ideas than is possible in a telephone interview. As I we will pp: be asking questions of a very personal nature. TRYING TO FIND ANSWERS TO QUESTIONS LIKE THESE: at do couples find to be the most effective ways of making their more difficult decisions? Which 'styles of decision-making' work best for which kinds of people? (We hope to publish an article on this in a widely read magazine, so that others may benefit from this knowledge.) people with greater experience—~either due to having lived more years or having been married before--use different approaches to handling iecisions? V do couples arrive at a way of dividing up everyday tg§k§_and responsi— Iilities? What strategies do very busy couples use to cope with these ;asks? 1 does the nature of one's employment, and the physical and mental fatigue t creates, affect the handling of home responsibilities? e, any information you give us will be kept in the strictest confidence. utside of our small staff will ever see your answers, and you will never ified or individually described in our written reports (of which you will a copy). We will describe only groups of couples or groups of individuals reports. ZTURN THE ENCLOSED SHEET NOW in the stamped envelope provided. This will 1 general idea of when an interview would be convenient for you, so we how many interviewers we need each week in various areas. About a week Id, when you have a better idea of your time plans, your interviewers will 2 and make a specific appointment. We thank you very much for your help Iring, and look forward to talking with you in the summer! Sincerely yours, Elizabeth E. Benson Study Director (5l7) 332—0924 290 HIGAN MARRIAGE STUDY - l974 artment of Sociology iigan State University t Lansing, Michigan 48823 |LY RETURN AT YOUR EARLIEST CONVENIENCE: l. We will be interviewing throughout July and August. 15' there any particular week or part of the summer which you would especially prefer? 2. Are there any weeks which definitely would pp: be convenient for an interview? Which times of the week are usually the most convenient for both of you? PLEASE CHECK SEVERAL CONVENIENT TIMES: Late Late Early Late Early Evening Morning Afternoon Afternoon Evening (around 9) I I L I I I ]> _< *1,‘ Please check your address & phone number below and correct any mistakes (to make sure our interviewers can locate you): we” r” / APPENDIX B PHASE I° TELEPHONE INTERVIEW 29l IDENTIAL INFORMATION 974 Newlywed Study Couple N°°-—--—————————— epartment of Sociology l___Husb. 2___Wife ichigan State University County Marriage type: l___wife lst, husb. lst wife 2nd, husb. 2nd 3 wife lst, husb. 2nd wife 2nd, husb. lst PHASE I - TELEPHONE INTERVIEW SHORTLY AFTER MARRIAGE of Interview , l974 AM tarting time : PM nterviewed by INFORMATION FOR INTERVIEWERS Any children from previous marriage? ..... Yes No Any children living with couple? ......... Yes No Respondent plans to work? ................ Yes No Occupation listed on marriage license: w 292 .INE 0F INTERVIEWER'S INTRODUCTORY SPEECH .__________________......__________________ a. This is b. from Michigan State University. We contacted you recently about a study of newlyweds. c. Did I pick a convenient time to call? (Could you spare half an hour to answer some questions about the decisions you two have made in the past and those you expect to be coming up in the future?) NING THE INTERVIEW __________________ Let's begin by talking about some decisions which will probably come up, everything from buying a car to deciding about having children. We're mainly interested in the timing of decisions--what things come up even before people get married, how soon different decisions will probably be coming up in the future, and so on . . . . (turn page .———__; ) INSTRUCTIONS FOR DECISION QUESTIONS (pp. 2-5) You will repeat the questions in the large box lO or ll times, once for each subject. lst subject - read eVery word of questions. later subjects - read only underlined words m m v N — A ... o .mL; o—Im .mLc mIv .mL; me .L: F .L: N\— xa .ucmgwwwzu umzzwsom .Lmzzewm umzszom .1 .I [III IIIIIII a on mmsou ow :wzz meme? Am . . memo» o— ..mzx m ..m a» m to N .mea m cw ..moE e ..moE m :_zowz wcoom 20; u:oa< ”mazes Lo mm» LHI/ e on em :wuuom Lo .Amz v xzm mflco_mzum m :3qu 63.83 umaw Icmucz Lo :o_mTowu m m ..mLc m Lo N .Laoz cm oma_ oz» =_ so ozone ”oz LH\\ W u go mczucmumcw ecu mo “can use um> czzumwc v . .mczoz o_ .Lzochm .szVpcmuz zpcmm: Lao mew ucwcwmm_u Lo cmzzswm :0: Au wucwcwwmzv wuwzc Lo ”zen :ox uzzoz . .wcx oz w . ..mcz m ..mcx m Lo N .mez m :_ ..moE m NaoOm :0: u:on< me2 0H ow .mzo ..woe m =_zovz Azosmm Homzzzm Hmzuuz oz do moves on FHLz m:o_mzowu oz“ so: nooumcoucz nowLFM m.uw mmzau Ion Lo mums on on meowmzuou 0: one wcwzu wmzmown was“ mH ucz cm ucmzouv m m :uuuom Lmzou cmuuo Lw>o ..mcz o_ op m Im_mz Amm no» u—zo uwz_mu so» o>mc maze: cue zo: uaon< w .zmz mzu mo w—uuw_ .30» XMw so» v—zoz mmcwvcmum :o w>mz cum :0: / Laos» wzo z? mzzo “scam :wz_mu on on gum xw so oo ”zozmzumo oz as an: ego do —Fc ..mzz m Lo 3 Nozzzos "waxes Lo mm» dz. \ uoz mH m m m Va v e F. “can—L095 Nope Jazz moon oz: .33; m5 250.; o: 3:95 :o» 35339.3 uoz mo comm 5:5 :ozIImoBFFEcho um: 223:0; 533: «:onm fizz 8 m m m “:33qu .m N N N “:33qu F F F acmuLanF .bxu 3:95;me .8 3:0; .5: a: 95:.» ”Song «23 AN 3: 2: 83 ”3323. 925348 “F: zoo $5885 2:: 4 G we: .3: $2.. 5: 9 2 E: 3.75 3 233328;: N m m anmtohEooc: Zn on uoz FF: u: no: PF PF .8 6333285 oFupFF : 3:. was: N N N 633L852: mFuuFF < .a: 92.75 3 .233 .8 anmfooczouc: 3.5a an 2.323 F F F anmfowEouc: .o a: .E 5.23 numb: m mt: mF .zFchomLmn 50950» .8“. Amcwfio :25 .33: szu 3 anSLanoo 88. :5 .553 we: 3ch 88.». ”FE; mus: an:8 >65 33323 C as :3 8a m m m Va m m m 952: E: e e z 8:2: .m $.52: .53; co 6.52: umgzmeom .33 «222.8 N N N 83 .m .83 23> :oz ocz 38.33 «E» :0 $53 25?; Box: F F F 83 .2: Scan: :33 Motifvcmnmg .50 ”Scan :0 .95 23 xo: 3 33 :3 at 225355 FF ID r—NVLOO’) r—NMQ‘LDMDNCD Q FF FF o ONQO FF Ln wamKDNm m I—NMV FF F. Fva :aLvFFgu :>.2:J!!J cu 00 (2 FF 6’ m n—NMV FF vamwmm P. Fem: UnawLLmE mzo ..D.L I 00 <3 FF f“) m r—va FF Nrn<~m‘or\m AFN: CWLUFpgo FcFmLuE :F FFFuwamv Lwcpo oxms on wcoFmFumu oz AcoFmFumuv an: F0 FF< .uchwFFFv umzszOm .LmFFEFm Fuzzmsom F on mwsou FF :wgz muqu L:o mg: pcmngFFv Lo LmFFEFm zo: o mgmmx oF ..mLsz. ..mLF m Lo N Lama n :F .,mos m ..moE m :quFz so; “:on< . NcoOm nonxme Lo mm> mF U cu ca "02 uFUv NwL:u:F ”onWHuuo FF Au LO NucmeFFFv qu:c Lo .quFucmuF aFgmmc "ham :oa uF:oz . ‘mgz oF ..sz..2».m.LoN .ggz n :F ..moe m ..moE m :Fchz Fcoom 30: u:on< ”wnaas Lo mm» NF Azo:mm Fumnm:m Fxmz OF ow .mzoFmFuma oz uF qua5 on FFFz mcoFmFuuu wcp :0; vooumgmvc: vmeFm m.uF wmamu -wn Lo mums on on mcoFmFowv 0: mg: wgmgu mm:mumn was» mF ”oz FF NwL:F:F an» cF szu u:onm ”zoFmFumo oz NF NonFmFuwu Lo mcFucmumeuz: :muuom Lo .xmz mg» Fe ugaa uoom u m :wuuom .uwugmum um:n uo>.:o» me :ox uF:oz -quca so :oFmFumu a mchume vgmzou cwuuo N . .mL:o: oF Lo>o ..mL; oF op o ..w;; m Lo v ..m»; m Lo N .L:o; :m .L:o; :aIFFa; Ann :o» uF:o3 chucoe me ummF mzu :F FF “:onm vwamu :o» m>mg mg:o; cos 2:; u:on< FmNF uF cc vgmzouv an: use Fo FFn :uuuom .xm: mg“ Fe wFuuFF chFucmum :o m>mc Lam so: A: OF ow .¥o\oz m“: Fa F: 296 m m m x: F. e v 23.895 28:25 ..oF 558:3 Lo 3:823 .323 «an: 3 3.33955: uoz m m m 23:85 .m 28:25 N N N $3.89.: 39: Lou. 355 Sin gong FE: .wmmthE F F F acSLoQEF 6me 335.5 m EPC 5.67:3 Em «.35 2328:: F £m< .moSmxs: 25:2: :9: 5%:qu ”: 3 at :5 at G 55 .mm: 3:: mo: B: 3.75 3 anSLoFEouc: m m m anmuLEEoozz 2: an no: FF: pm we: FF WF .6 .anSSFEouc: 23.: a 3:. made N N N 253328;: wFFuFF < .q: 9;; 3 :33 Lo 33:85:: 33: mn wEFFmEom F F F wESLOFEouc: .o as Fe FEE: numb: a $5 mF .FFchome: F3950» Lo“. 3350 :23 5:3: i3 3 anmtoFEou 985 ES 333 0.3 35.5 2.3 umfi «EC 333 Pw>w 3:323 C 33 :8 83 m m m Va m m m 3:2: >5; v F. v 3:2: .m NwLamc: fig .5 6.32: umczmsow 6.5m “229:3 N N N 33 .m .33 Em; :95 PE $3.33. mFfiF co mucmuw wngw: Ems: F F F 93” ~22, “:2:leme Pet: FEEDS; .So «:23 :o 9.: 23 so: 3 3.: :3 $3 297 FF co r—NMQ’LO‘DNG) V ~No OD n—NMV FF GD 1") FF F-r— (Viv wwnxo N42 33 CO r—NMV FF FF FF v—r— NMVLD‘DNW 01 :NF Q3 r—Nmé FF \0 m FF -—-v— vammkm 3N: FcFmLmE :F FFFumamv szuo uooumgwuc: FumoLF< Fuownn:m «xv: on omv wxme on mcoFmFowv oz FcoFmFuwuV Fax Fo FF< mmxmu Osman—F: mwxup UFFS ”FF-D; U. .ucmLmFFFu nozzweom LmFFEFm umzzmsom ou mmsou FF :mz: mmva L:o wLm uchwFFF: Lo LmFFEFm 3o: F wgwmz oF ..wLme ..mLF m Lo N .mex a :F ..moe o ..moe m :quFz Fcoom 3°; p:on< ”means go mmF FFHV u on ow noz FF uzonFu o FF FF U LO NucmeFFFv muF:u Lo .quFucqu FFmec .amm :ox uF:oz . . .mgz oF ..mgz m ..mgz. m Lo N .mex n :F ..mos m ..moe m chpF: Ncoom so: u:on< .mnams Lo mmF m F4 Nmums ma FFFz mcoFmFuw: mzu 2°: voomewuc: umeFm m_uF mmamu uwn Lo mums ma cu mzoFmFow: o: wLm wLmzu wm:muwn szu mF "oz FF FwL:u:F mzu cF szu u:onm :o co .zoFmFomo oz FF NFcoFmFumu Lo mcFvcmumLmu:: cwuuom Lo .xuz wzu Fo pgm: voom a m :wuuom .umugmum um:n -quc: Lo :oFmFuw: m uw> :o» xmm :o» uF:oz chFvcmuw chume ugmzou cmuuo :o m>mz LmF 30: N . .mL:o; oF Lu>o ..mgz oF on o ..mgz m Lo F ..mL: m Lo N F:oz cm .L:oz cmuFsz mam :o» uF:o3 me ummF mzu :F FF “:onm vmemu :o» w>mz mg:oz cue zoz u:oa< ”WFF FF :m vgozouv Fm: wzu Fo FFm cwuuom .sz wzu Fo wFuFFF Nmzucos 3 OF 8 £8: E F: E NwL:u:F mgu :F xgoz uszE Lo mxgoz wFF: as» cu:s :0: Fa =:_Jnu:r ,:- xx m w m Va 3:33 23 :F 9.3 F F F “:33qu 235 go we?» FEES; 2: no. Fo w u m5 “3% F23 3 37.2353: uoz m m N 23395 .m N N N 28135 ....EBE 2: :F F F F ucmuLoQEF .prm 8.3 £35 .8 48.3 wFFz 2: no. Fe m F 23 30% ans: 3 at :3 G: 8 S we: .mm: EFF: mo: 9 2 3: 9:5 3 3253285 m m m 233L852: FFm Fa uoz FFm pm we: FF mF Lo .anSLoFEouc: wFfiFFF m 3:. 2me N N N anmtoFEouc: 23: < .n: 9.15 3 20:3 .8 3333285 35a mg @5353 F F F 2:33:62: .0 2 FE FEE: uuuhfi a wFS mF £22823 torso» .8“. 3.550 :2: 5:2” meu 3 anotoFEou 985 new $.53 9; main» 25» $5 mucFF mFa:ou Pm; F3383 C 33 :3 33 m m m Va m m m 3:95 bw> F F F .952: .m N932: >5; go .952: 3:553 .23 «2353 N N N 23 .m .83 b3 :0» m; A833. mFfi :o 2253 23?; Buzz F F F 95” f2, Scamlemqu m_wFF:Fm_u:Em:; So ”53m 3 9; 23 3o: Fm 83 :5 $3 2539 r—NQ‘LOM r—wamtohw O Fwd-mm v—NM¢|-DlDV\m C A CD r—Nle-DSONUD Q“ v—Nvmm v O xo chLmFFFF .o ucmngFFu . LmFFEFm . quFucqu xFmez (I'll/1 xo mgmw» oF Lw>o co p—NM‘!‘ NMVLD‘DNG) co r—NMV NM‘rl-nkDN‘D A C) M v 03 v—NMV NMVLDVDNCD FcFmLmE :F FFFuwomv szuo uooumxmucz FomeF< Fuomnnzm uxoc o» omv wme o» mcoFmFoou oz FcoFmFuva Fm: Fe FFF chwLwFFFu muFau Lo .chLoFFFu “answEom .LmFFEFm umgzqum .FmoFucqu FFmec ”mom so» quoz F o» mmsoo FF con: mmmoF Lao mgm chngFFo Lo LuFFEFm 3o: Fu n EgzoF..Kxf..$z ...flon m Lo N .mex a cF ..moe ..mg» m ..mLxN m :0 N Low» m ..moE m :Fgqu Fcoom o :F ..moe o ..moE m :Fgqu so; uson< uwnzms Lo mm» FF Ncoow so; uson< ”mnFuE Lo mmF FF v 0» ow uoz FFHV Fence we FFFz mcoFmFuwu mzu so: oooumgwozz ommLFm m_uF wmzuu non Lo owns an op mcoFmFuwo o: wLm wgwga wmamowo pun» mF "oz FF N¢L3F3F FagzuaF mzu =F mFFF "scan .chqu on on Fow.xw so on .zoFmFomo FF Au ”onmFumo oz FF An .Fo FonFmFumu Lo msFucmumLmvca :m uLmzouv Am: as» Fo FFo copuom Lo .Fm: ecu mo uLoo coca m cmuuom .Aoz on» F0 wFuFFF a coupon vmuumum umzn .w>_:oa Amm so» quoz Nszucowm Iquca Lo :oFmFumv a mchouoL oFozou coupe :o m>mz LMF zo: N. dtssoFLgo..EgoFoum..2;mLo F ..mF: m Lo N .Lao; cm .Lzocucmume: Foo so» quoz Fmguzos me ume mgu :F uF uzooo owamu no» w>mz mgzo; :m5 :0; u:on< umm> FF Fa 0F oo .¥o\oz FFV NLMF cm a: wEoo uumnnam szu no: Aw 2539 r—Nfi'lnm v—NMVLDKONW O 0““)va mevmwkw O m m F N F A O‘ FNMVLDWN‘D V O V go chLwFFFv .a quFucva .FFLowz Fa mgmwF oF Lm>o 03 [—va NMQ‘LDKDNG) ,_ co r—Nmfl' Nmfi’mlDNw 03 A O M r—NMV v NMQ’LDQDNQD FcFaLmE :F FFFomamF Locuo noopmgwocs FumeF< Fuuannzm uxm: cu omv oon ou mcoFmFomu oz FcoFmFuwuV Fm: Fo FF< NpchwFFFu quza Lo .ucwgmmmFu umzzmeom .LmFFEFm nozzosom .FmoFucqu FFLom: "Fom :oF vFaoz N ou mosou FF con: mmva Lao mLm “:oLmFFFv Lo LmFFEFm 3o: Au n ESFOF?EFW..€F ...2FoF m Lo N .meF a :F ..moe ..mLF m ..mLF. m Lo N .meF o ..moE m chuF: Ncoom a :F ..moE o ..mos n :quF: 3o; usoo< ”onFME Lo mm> FF Ncoom so: uzon< "wnFmE Lo mmF FF u ou ow uoz FHHV Noose on FFFz mcoFmFumo wzu 2o; uooumgoocs umeFm m_uF wmsmu -wn Lo moms mo on mcoFmFomu 0: mL» mgwcu wmzmomo umzu mF ”oz FF \/ ngsusw wsu :F szu usonu nzoFmFomo FF Au nzoFmFumo oz FF Fa Lo NFcoFmFuov Lo mcFocmumeocz cm ugmzouv Fm: asp Fo FFm :wuuom Lo .sz wzu Fo “Lon uoom m :muuom .sz mg» Fo wFuuFF m :wuuom umugmpm umzn .w>_:oF me :oF quoz omcFucmum -Lwocz Lo :oFmFumo m uchuqu ogmzbu :ouuo :o o>mg FmF so: . .mgao; oF Lw>o ..ML; oF ow w ..mL; m Lo F ..m;; m Lo N .Lzo; cm .Lzo; :o-FFm; Fun :oF quoz ngucoe me ummF mzu :F FF Fauna oomeu :oF w>og mgzo; cue zo; onn< "mmF FF Fa 0F ow .¥Q\oz FFV NLMF ow a: wEou Fowwnam chu mm: Fa v00 v—NM wa r—Nm l\ l—Nm QC!) v go “:upgooEF FFLszuFuLoa uoz chuganF .m ucmupooEF pcmuLooEF .Luxm 8 m9: .3: 6:: mo: FF: “:5: AN . . . .choFqu uzono .mm>FumeL ;qu wEFu mcFucwom ,mmao; m mcFFan .coFumoauw onnm mchchon ;chz mmchu --ommmFE w>.F :chz «:oFmFowu Lo muumnnzm Fugue wgwcu wg< 300 "NM v—Nm r—Nm A W v I—NVLOM l—Nd’mm l—NQLDM A LO v anuuLoFEoocz FFm an no: wFamuLoFEoucz wFFuFF < mFamuLoFEoocz .0 yo ogsmcs FLu> ogzmcz .m ogam .m ugzm FLo> No: mcFLo op anaFLoFEoucz FFm um uo: FF mF go .anmuLoFEoucs wFuuFF m awn. mnFoE .n: mcFLn ou Fcuzou Lo anouLOFEoucs quzc on wEFquom F; FE ;onz poo.n:m o szF mF .FFFncowLmo FmegaoF Lou Fmgwcuo cusp .uzoom xFou ou anmuLoFEou «Foe uco Lmmew wLm mmchu wEom uozu mchF mFozou FFw>m Fannogmv NmLsmca FLo> so .wgzmcs umczwsom .mgzm uozzweom .wgzm Fgw> :oF wL< Nuuomazm wFsu :o mucoum w;m\w; mLmzz uzoom--NmomuF m.mFF: m.ucmom=: Lao uzonm :o wgm mgzm 3o: Aw u. 301 ne decisions will obviously be more important to you than others-- it is, the results will make more difference to you (than the ;ults or outcome of other decisions). I'd like to quickly read "ough the list of decisions again and ask how important each is you personally: extremely important, important, somewhat )ortant, or not particularly important. Please just speak for Arself and not for both of you. it about gbuying large items} ? Are these decisions 35; tremely important, important, somewhat important, or not particularly important, to you personally? REPEAT FOR EACH SUBJECT LISTED ON PAGES 2-5. MARK ANSWERS AT BOTTOM OF APPROPRIATE COLUMN ON THESE PAGES. OMIT SUBJECTS WHICH ARE IRRELEVANT (Where Tower portion of column is blank) w L158 43...... .14.... -. 3-..... ...-.. 1.1:} - 302 FF .0 op omv zoo: \mLc ”new ULm zoo: \mcz zoo: omzcoz "ooh oCN mgzoz FmFoF Nxmwz m mgzoz chE 30: FwNV FFFSF mF Fee: F0 och one: FFanoF canoe EsoF Fsonm Fae: F.02V Fm>me 30F FFF3 mnoF Feaa 30; .LaemeopFF ”mMF .moma mo mm> FF FFF .o 0F omv zQIII .F FFF .o oo omv oz .F mmF FFomnocm moF Nmzpcoe m Fxm: ozp mchzu .oco szF op :oFFFvom :F moon Lszo o>mz :oF FFFz Fov zoo: Nzgoz FFoonoca :oF FFFz zoos o mgsoz FcoE 3o: Fov \2; / F.on Nmmwcszn czoF cF oFaowo Lozpo FoFoEw :oF oo Nzooop :oF on Fo>oF Fuzz No>Fo>cF FF moon Fmezv FFFFF mF xeoz Fo UCFF ome: ummozm .zoo mum F . o mcFoo on :oF FFFz Fuzz o mm zcoz FFFz :oF oc< Fmv mooo .ooo AFN-mNF .mmzo: mcFooozv SNF m F m u l i N N 303 Fomv FmNF I zo oz mwF NszpcoE m wac asp FOFV mucoFLF Lo mo>FFmFmL zFFz :ozF Lozpmc 0:30 Lao» Fo mooFm o :F mcF>FF on :oF FFFz FmF I zo :oFFmLoowm zpcos F Lo>o .moF coFFoLmoom mmoF Lo :pcoE F .mmF oz NszpcoE m Fxo: ozp Fo onv ;pcos wco cog“ oLoE Lo; ”mNF FF Nmzucoe m wac asp mchzo wEFF Fo quLma ch LOF--Lszo zoom EoLF pcooo mcF>FF--owpoFmomm on :oF Fo 02F ozp FFFz FmF Foogom FmoFcean m amwFFoo .cn N amaFFoo F If qumc FFFFSF mF Fooeom F0 man one: Homemav Fmep mF Foozom Fee: FmF .0 OF OFF zolluw FmF .0 OF omv oqlllm pcoozpm wEFFnFLmo .moF N Fewoapm «EFF-FF=F .maF F 304 FOFF FONF meF FFmF FOOF meF FFmF FOOF FNNF FFmF FwF .0 OF OFF mHHF zomzu .Foozum mom Iwzozm oFo Foz FF F F _ F F N N F NFLOFOEO; FszF zFFz :wFoFon ozF FFFm: .oF N F m F m N F NcoFoFon mcF zFFz «Fgommmowc FF .Fsz: Fm a: Fog .m NoF ocm m zm< IFFz wzF>FF zmzoFqu FF N F _ m F m N F FNszmmzszo ozF F nooF FoF mmcmFu Fmaazm wcF on .m N F F m F m N F NFmaazw mgmowga .F F F n F F m N F NFOO OFF FO OFFO OFOF .O N F F m F m N F Nmmso; ocF oF wFF>cF _ FFFz :oF wFaowa szz woFomo .m N F m m F m N F Nmo>FFFFo.F zFFz zozoF cF awmz FF N F m m F m N F NoFoFocoF Fo “cwEmmme F oF .Fgmmmoom: FF .chFmFano oxwz .m N F m m F m N F Nomzoz wcF oczoFm FmFOF .wFsF F -FCF3F .mwocmFFaamF mmchF FFmawz .N o F m F F m N F FFOFOOOFO FFOOOFF OFF OO .F mm mco m mxszo mFoE wEom oFoE mFozFO FocFFoz F FmoEFm wFFz onoo .nmzz FmoEFO F F OFF: ;FOm .OOO: "on FFOFFO 304mm ZFFF IOFF «OF FFFFFN cho LozFFo: Fo-umFO3FO FmosFO OFF: .ucmnm: :ono zFon .oFFz csz mFoE ocmomzz .mFozFo Fm NszFcoE 3wF Fxm: sz :FF : cwzF oFoE mFFz .wsmm wzF oEFm ucmnmsc Fom :oF onoz #4 ”mmozm illil. illill l (Fri...!)( {EU USU!T.L\ ‘..r~r..\ Kira; WIT... flir‘l ~..;\;... lr ...... erJ .\!.p\... ..\;i. ...i\..\ ...;l.\.1: 305 N FFO FO FOz szo MFuzFo mcoFFm fiflflw zumzo .Foozum mom Iwzozm Q40 Foz FF m F m NxFoszoz LFof :FFz coFoFFso sz onz N F m NooFoFFco asp :FFB «Fgmmmmoo: FF .Fng: pm a: Foe ”2 com m zm< IFFz wzF>FF zmzoFon FF N F m FNngmozcmFu wzF omoF gov mmcva Foaozm mcp on m F m Ngmaazm ogoome m F m NFmo sz Fo meo meF m F m Nmmzo; wzF 0F oFF>cF FFFz :oF mFaowa Fuzz moFowo m F m Nmm>FFmFOF ;FFz zo30F :F aoox N F m NoFoFocoF Fo .cmEmmme oF .chmmwooc FF .chFmFano oxoz m F m Nwmzoz wzF ocsogo FdeF nwFFFFzeF .mwocmFFanv mmcF;F FFonz N F m chFoaozw FFwUOLm mcu on mcoLFm mcoFFm mcoFFm chszom Fgo> .mcoFFm ch> me :oF o_zoz ”unazc oF 306 Ncoom so; onn< on ”moFFommo n:oFFmsFFm FozFQII: now chmocg 0F :oFFFooo :FIII wEFFIFFma FFsklll NOEFF-FFOF FOF [ll NOOFFFFOB FO FOFF FOez FOF .ocFE :F m>mz 30F on Fogz FFN .0 OF OOF anu. FFN .0 OF OOF Mann: OOFOEIII FOF .mzo Fomzo NFcFoa wEom Fm xFoz OF mcFom . FLO: Fo mcFF FcoFoFFFo o oFCF mcFom onoo mcFFCFzF :oF oc< FON FoczFOF sz onnm mcoFFmoso ZOF m .302 FFmoz FF FamEFFoaO omes; so common FF oFo Lo ”mmozmv mav—OOU QSu. UwU «mm—.LGUOLE DFFJ 3:33)) )..: It.-. MCHLOH Om 3:6 DUFF—0.3 )2) 5:... 307 FFFF FrtF mwog 0F ago: :oF FFF FMN .0 OF OFF Oz .FOqlnnm FNN .o ofilln OOUF ofill: III: omv oz .QOFF F “mm .o op omv .xo w wnhmz m mOF mOF .OOFFTIIN .FOFIIIF ”on MOFIO FNFFOFFOFFOO FO FFOOOOFFF NCOFUFFLQ FOFoEF quom Lo O>Oz oF mchcoFa :oF OFF NwFssz OcF onnm Fuzz FNN l FNF.IFF FO- FFIO FOOFOO ;mFF FO FOO m ;mF; OFOOFE :F .Omo FoogomwFa Fo FUFFLQ FOLFV :oF :FFz EOzF ;FFz mcF>FF FFFcomwFQ COFUFFFU .oz FLOOFOF FOOOFO OFF FOOOO Fng ”zFIF IFFz OZFFFF OFFIO on zFIF Fmoz FF FOF O...F.N womcm FoogommFa F FFOOF ;FFz OOF>FFF OFFFO O .1..FMFF..FF FO-FF.. SFIF.N Foosom ; F; Fo Fso m s F: wFooFE m oomcm FoozomOFa F memcsoF OFoocom ;sz Fo Foosom mFooFE .Foocom momgm Fo woo mcF>FF UFFco Fmomcsox OFF mF FQF UL.)CH. U .+L.> .‘.+r\~h\ .¢.\\u n W fimqv / 308 FFN go OF OOF Fuzm ncm csoFm mFO chOFFgo mgp FF< Fm H:wFFFFCF N F N m NFoogom ;mF; FO OFOOFE OF OFFFO < FF N F N m Foogum wanm FFF OFFso < Fm N F N m Fawn m uo: Fan FwFoosommFa < AN N F N m NFOOO < FF mm OOOZ OEFF-FFOO .OEFF-FF=F "mF vFFco memczoz msw mFF;3\me3 mFmLF FF uxFoz anOQOFQ 30> quoz Lose 30: .wzeszNm NIF F< FN Lo Nw< :FFz Fm420 zmmoFFIQ FF umFoz ”FNNOFOF No FzmmNNN zFF ZNNOFFIO mmm FF FOF FON .o OF Omv 309 NcommwF.Fm:3 Lon NchF mF cwcz FON .0 OF OOV Faun] FON .0 OF OOV Oqunu woF ngzeF anmwmeoF wsp :F mEFFIFFzF cusp mmwF FFoz Aannoga :0» quoz H>._zo zmz FON .0 OF OFFOF xIIIIIIIIIIlIIIIIII1I.IIII.IIII..I..I......u................. FOmF ... F _ . .-....l... NFOFF FF O53 an I, I, Fa w o: m maxgazllN mma F «Op wsz 33>. 3.53:, :35 >1 . .uui » McmEoz waLuu u nu. 200.535 .> 310 $3 $3 mev 33 FE w m m N F am a xo < wc on; wFaowa cmgF FFwFaam; wgmv mFFF Fo on wFoE me mFaowa uwFFFOE FmgF me 0F mem mF FF .Fmecmm :F .FFoz EoFF mac; mme m: cmcz XmFmL 0F waF meF 0F Fsmzo cue < .vwFumaxm ma quosm cwsF wFoE mcFou mF chUFFF w;F uc30Fm maFo; 0;: gas < .m_Fm;FOF O COSF now m.Fm:FoE a mFoE mF :mFuFFco mcFmFom .choE ;FFz AFFEFF mgF mcF>OFq quo;m ucmnmzc msF Una mchmwFo mmzo; new mchoou wzF ow quocm mFFz mcF .mwmmo FmFomam :F FawUXN .wEo: Fm m_w; cos; xFoz 0F EF; Fumaxw oF FsmFF 0: mm: mFFz ch .AFFEmF mcF FFoaazm oF mcFFFoz mF :wE wcF FF .cuas FmgF FFO uwvwwc Foc mgmz Faces MFwa 05F FmgF mcFUF>OFa “new m w>mc mmq FwFFz as szFV LwcFMF szoz F .zmz Fm; cF ucmFm OF chFF 0: mm: unwamzc Fm: nxgoz 0F chmz wFFs m FF .mmchF FmFomam FOF Faces m_wFF3 wcF mume me vcm mwFFmewuw: FOF Faces m_ucmnm:: mcF mm: 0F men mF FF .xFoz mFQoma ;Fon can: .chqm mF FF :0; onnm =de: OF mFoE w>ms stozm Faces «Foe mcme as; :omea w5F .mmmFFFmE CF Fm 311 FFOV III. .wEmm mgF ;FoFIII Fe m \Fmmew FF mm; chFch N Fmmem FF w>mc :wEoz Fmemw FF w>mz cue Nmesoam coom a szucFFv mEFF Fmmem msF m>mc :mEoz Lo cmE szF Ewwm FF mmou .mocmFFwaxm :30 F30» :F oFmmem FF m>mg :wE FmgF zmm mezFo vcm mmzoam voom m mchcFF wEFF Fmmew cm w>m; :wEoz FmgF me mFaoma meow Fe m o: N mwx F .11. NmuFoco an umFFFmacs mchFmeF --nFcmemv mcFFFFoE Fo: Fwnchoo FszoFgwm Fw>m 30> UFU .mmez 3mF meF mSF mcFFso Fe m FwFFMIIIN wFoFmAIIIF NFw;\EF; FmE so» FmFFm FFFc: :02 0F onoo cw>w FF F_:uFu Fo n:anEFFmE :chFFO wzF FwE so» wFOFonllnngmEFva 0F mFFF FsmFE 30> FagF 30% 0F quoo FF UFU cmss .FmeFso» Fog nmzmz chngFFn cF vwFFLmE me oF wcFowv mFQomm .Fcmewv cmFFFmE me oF coFmFomu wcF mF mco FmN FwN AAA/\AAAAAAA KONmel—‘NMVLOKO KDKOKDLONNNNNNN vvvvvvvvvvv w v m N F m w w m N F N w v m N F F ”wzFZOFFom NIH onm< om4< ¥m< .mZMFF m>om< NIF 20mm FzmmmuuFo FF OOFE FO;\OF; OO m.FO;3 FFFOOXO mcFFOm uFF F F OEOOOF OFOOOOOO CO OFOO OF FFFFFO< .oF F OFOOOO FOFFO FO :oFFOFOOchou um F OFOOF OFOFOOOO; mcFOO FOOOO mmOOFFFF< F mOEFF FO OFFFFFFFFFPF OO OF F4FFFFO< F FOLFOE\FO;FOF OOOm O OO OF FFFFFO< F F F F O O u I\ oo OOOOOFFFOFOF NOW so» FOO; OF mcFOOFmFF u r-NmVI-DKD OOOOFOOOOO FOOFOFSO o mmmmmmmmmm NNNNNNNNNN :OFFOOFFO mwOFOxm OF FFFFFO< FOOOF :0» UFO ”.0FN .wszmzz FF NchF m0; Foosom F0 Uch F053 ummomm +FF oF mF 0F mF Foocum mmOcFOOO \FOOchowF NO>0£ 00> 00 OOOFOOO F003 ”NOON; +FN 0N OF NF FF OOFOIFOOO 0F mF 0F MF 000:8 NF FF OF O FOOOOO ;OFI ”ONFMFOZOO FN>NF meIwF: NFONFO FOOFOFOEOO 00> 0>00 mcFFOOLOm F0 OOFF0OOOO F0 OFOOF >005 30: me NOEFF FOSF F0 00> OFOZ OFO 30: N00 wzm UFO FLO: F0 OOFF F053 mw> xo oz NOEFF F0 OOFLOO 500 FOF OEO; 00F OOFOFOO Ooh 0 O>0c FO>0 0:0 OFO .OFOO OFOB 00F FOFF< ”MFFzmmzoz FF FO0FOF0Om OF0EOF F0000 F00 .FOSFOe 0: FFV FNFoogom 00F; :F 0003 :0» OFFsz OOFFOOOOOO F000: m_FO;FOe onz O03 F003V NO.FO;FOs F00» Oc< FOV 314 0FmF . ”0000F0 “00F0003 F00> F0 0F00 00F 05 FF0F 00> 000 n>FF00FFF FFm 000F 00\¥q1|lw F003 0 0000 c0gF 0000fl|10 F003 0 0OOQIIIm :F005 0 0OOQII|0 >FF000Fm0OOQIIIm N 00 mIIIN F0>0QIIIF NF003 0 00:0 000F 0F05 00 .0003 0 00:0 nFFFFFOFFF 0 0000 n>FF000FO0OOO .0OF3F 00 0000 .00>0z Nm0OF>F0m m0OFmFF0F 00NF00000 OO0FF0 00> OFO 00FF0 30; F00> FOOF 00F 00FF00 Fem N00 OF FF0mF00> F0OFmOOO 00> 00 .>00 FF .OOFOFF0F F003 ”mmomm OOIII. Fall: 00> FF0OF00> F0OFOOOO FFFFm 00> 00 me NFOOF O03 OOFOFFOF F003 ”00000 OIII.OOFOFFOF FOFFOIIIO :OFOFF0F OOIIIF FmFF0F0050000Flcoqlllm FmFF0Fc0E0OO0HIIIm IIIIIIIIlllrllllnllllllnnxalllx ”FzF OFFOZV NFOFF O03 OOFFOOFEOOOO FOFF FOOOOO FIFOOFOOFOFFIII XOOOOFFQIIIF 0>FF0>F0mcoqlllm O0EFOF0m1IIN NxouocFFo LO .0>FFO>F0m:OU .O0EFOF0N FO0F O02 "Nmoma All: ;mF30mlll OFOOOEIIIF FOOOFFFOOFFIIIO NOFF00F00 000005 LO OFFOcF00 _0:O_0_:O_O 03:) as: .1--: . III 315 FOF00 000> FFO 0F0F000mm0 >FF000 03 “00> 0FF3 00F0F0F 000000F0 0 0000 0.FF D0000 00> 00 0000F 0F F003 F nF070 OFV 0F200000 00111 00>0EII: 00> NF30F>00F0F F0000000 000F-OF-000F 0 0FV 00F0000F0F 00 F0OF0 00> F00F 00F0F 00> 00 “F00 00003 0F0F FF U00F O0 F00F0m0 00 0000 0>_>00F 000F0F000 00F F0000--000F0000 FO 00F000 30F F00FF 00F 0 0O00F 000 F000 0>_O03 0F0000 0FF3 000FIOFI0O0F 0F0F 0F000 03 00003 .>00F0 000 OF F00m 000000 0 00F000 FO 000F 00F 00F000F0000 00.03 .00OFF0FF0FF 0FF 000 30F>00F0F 0000 0F0F FO 00F0 0F0F .000000 Fo F00000000 ¥QIII 00> 0>00 0F003 003 0000000 0000F 0F N0F m0F>FF 00 00> FFF3 >FFO F003 ”mmom0v ml: N00> 0FF3 0000F 0F m0FFF0m F0000 00 03 F0mF0 300 .0000000 000 0000000 000> FF 0:111 “mmmmoo< 000F>FF 00> 000 00003 00000000 FO000OO 000> 0>00 03 0000 F00 0_F ”000 .FF .00 00 NF>F00 >F000 00 0000 0Fv 000F 0000000 0000 00F F0 00 00> FFF3 .00F000 m F0000 0F >0>000 00F 00 F00m 0F0F 00 F0Om00 0000m00m 0 00> 00F0000 00 FF_03 F0 F0 F0 F0 316 a/ a .00 00Fm000 00FF0 NF0F0F0000F 0000 Fx00 F00 FFFm "003FF>00F2F 3mF>mszF no azm .0>00000 .......:o>. 0.00an .cmeFFF C. 000000 000F0 FF0O OF 00FF NF0F OF 00FF 0000 0 00 0F003 0003 ummmIF Foz 0F N000\0F0 FF0O OF 0000000 000 00FF 0000 0 00 0F003 0003 000 0F0OO F 00 .0000F 000\00 0F Fm .NFF000000N .0000F F000F>FO0F 000> FO 0F00 F00 0F 00F003 03 N000000O000 000FEF0 0FF3 OOOFFOOOO OFFFOOmm 00F FOOOO 0F0F 000 OoF FOFF 000 F OFOOO FN FF .OOF .0FF3\0000000 000> 30F>00F0F OF 00FF 0.03 «OFOFOWOO FF 317 IIENER EVALUATIONS & IMPRESSIONS 1AKE AN ”X” ON EACH LINE: tespondent's attitude toward interview: Ixtremer cooperative I + - .suspicious, nterested, eager ' lhostiie lespondent's frankness/honesty: Did not give the xtremeiy frank . + — [impression of . ' Ifrankness espondent's f1uencx (ease of expressing ideas in words): Not f1uent/wou1d ery quent 1 + - .be impossibIe to 1 interview 1ater espondent apparentIy was: ___a10ne ___with spouse ___with chi1dren ___with others id respondent indicate that coupIe had cohabited before marriage e.go, a1ready had a routine of hand1ing tasks, etc.)? __Definite1y cohabited ___Definite1y did not cohabit (said so) (said so) _Not entirer sure: Describe any clues or comments in fuII. See especiaIIy Q. 19. TAL LENGTH OF INTERVIEW: minutes Y OTHER COMMENTS ABOUT THE PERSON INTERVIEWED OR THE SITUATION: DENTIAL INFORMATION ’“ ' 74 New1ywed Study COUPLE NUMBER: partment of SocioTogy chigan State University . st Lansing, Michigan 48823 Husband Wife __________________________________________ PHASE II - PERSONAL INTERVIEW, SEVERAL MONTHS AFTER MARRIAGE Part A — INDIVIDUAL INTERVIEW Part C - JOINT (COUPLE) INTERVIEW (P1ace address 1abe1 here) interview , 1974 . . AM ing time :“ pM viewed by: 319 In your te1ephone interview we spent a Tot of tim ' . . , e ta1kin about 62:1: ggcliionsfi HANDfLIST #1 T0 R, ALONG WITH FELT PEN. gWe roug a 0 these a ain, b ' a coup1e of them. 9 ut I wou1d Tike to ask about Ta. Before I do that, wou1d you pTease cross out the ones which don't apply to your situation. PAUSE. Now wou1d you a150 cross out the ones which you consider to be fairTy unimportant. 1b. 0f the ones which are Teft, pTease te11 me the one subject where you and your spouse's ideas are the most deferent. (PROBE: If you had to choose one, what wou1d ou sa ?) (GO TO Q. 1c). y y NAME OF NO. MOST DIFF. SUBJECT: 1c. From the ones which are not crossed out, wou1d you a1so pick one where your ideas are somewhat different--about ha1fway between the ones where you have the most simiTar and the most different ideas. (PROBE: If you had to choose one, what wou1d you say?) NAME OF No. ___ SOMEWHAT DIFF. SUBJECT: 1d. Let's ta1k a 1itt1e about (name of “most different” subject): 1. Since your te1ephone interview, has No 1 this subject come up at a11? UK I 1a. IF YES: Since then, have you YESL'ttT 2 ta1ked about it just a 1itt1e, Fair e 3 a fair amount, or a great dea1? Great 4 DK 8 2. How far are you toward reaching a decision or understanding? Wou1d Just you say you've just started, gotten started 1 a 1itt1e of the way, gotten a good Litt1e 2 part of the way, or gotten a11 of Good part 3 the way (toward an understanding or A11 of way 4 decision)? DK 8 3. How sure are you about your Very sure 1 spouse's ideas—-about where he/she S. sure 2 stands on this subject? Are you S. unsure 4 very sure, somewhat sure, somewhat Very unsure 5 unsure, or very unsure? DK 3 4. For yourseTf personaTTy, is this a No 3 subject which might sometime be DK 8 just a 1itt1e uncomfortabie or YES: touchy to bring up? Quite 1 . A Litt1e 4a. IF YES: Quite uncomfortabTe uncomf'T 2 or just a 1itt1e uncomfortabTe? ' 320 e. What about (“SOMEWHAT DIFFERENT” SUBJECT) ? 1. Since your te1ephone interview I No has this subject come up at a11? DK 1a. IF YES: Since then, have you YESfi.tt] 2 ta1ked about it just a 1itt1e, F1' e 3 a fair amount, or a great air dea1? Great 4 DK 8 2. How far are you toward reaching a decision or understanding? Wou1d Just you say you've just started, gotten started 1 a 1itt1e of the way, gotten a good Litt1e 2 part of the way, or gotten a11 of Good part 3 the way (toward an understanding 0 A11 of way 4 decision)? DK 8 3. How sure are you about your Very sure 1 spouse's ideas-~about where he/she 5. sure 2 stands on this subject? Are you S. unsure 4 very sure, somewhat sure, somewhat Very unsure 5 unsure, or very unsure? DK 3 4. For yourse1f persona11y, is this a No 3 subject which might sometime be DK 8 just a 1itt1e uncomfortab1e or YES: touchy to bring up? Quite . 4a. IF YES: Quite uncomfortab1e “”S0mf 1' 1 ' t a 1itt1e uncomfort— A L1tt]e or jus uncomf'1. 2 ab1e 321 o N _ N m N uA>z< LHV mHzszoo omzom wcp uczocm mmou :0» mo Loam nose :0: o m _ N m 4 ufl>z< LHV mczmzzou uw_ppmm pom mwpzamwu go: o N _ N m a US: LHV mhzmzzoo :pgoe om Ugo Lou .mmspo_o .mmzo; mcpuum:w>w_ do ucmucmpm Lzo> o N _ N m e ufl>z< aHv mezmzzoov m3m_-cT Lao» :wwz awcmcowmeQL Lzo> o w P N m e N:72 LHV mpzmzzou UQFvcm; mw chos mcp so: cowomooom mm __m,pa w_ppw_ asom _mae ”psoam was: -xm oz uoz m umzw pmmxw Amzo NNQNHQV Nuwpuw xm so was: Eogm mwcp m? “cwcwtkwu nose :0: Amy .mzNHH NN< Noe HH4 zmeJHIQ mH nCz< aHv mezmzzoo mco_w pom pcwgmalampm 3m: mcp ucm :vacho wcp 30: "IHHz wzH>H4 ZMmQJHIQ mH ufl>z< aHv mHzszou AFMLwcwm cwv mmmwxgme :w my wgwgp pcmszohcw use =::w= 5035 so: ufl>z< aHV mHzszoo AFmmeom :wv wmmwgcme :w mw mewsp apWFWDVmcoamm; :xgoz: sons 30: nA>z< aHv mhzmzzou .upw .mmzo; asp uczogm mxmmp co gmgpwmop mcwxgoz vcmam so» wavy £035 20: ufi>z< aHV mhzmzzoo 30% com :owpowyem use m>oF we :owmmmgqxw m_wm:oam Lso> .m— .o_ .m .w 323 N _ N N N ufi>z< NHV NNzszoo wmzoam Lzoz ;pwz ucmgm :oN mswp mL2NNNF mzk N _ N N e uA>z< NHV Nszzzou wmzo: 0N3 vczogm Nmoc so» No comm cogs go: N _ N N N . n:72 NHV NNZNZZOQ to_wpom pom Nopzamwv 30: N _ N N N ufl>z< NHV NNZNZZNN Lugow om can Leo .mmgpopo nmmzo; wcplumcw>NF to ugmvcmum Lzo> N _ N N N ”:72 NHV NszzzoNv ‘ N3N_-:w Lao» ;pwz szmcoNpNFNL Lzo> N _ N N N HSE NHV szmzzoo vavCN; N_ choE ms» 30: mm Nem> poz pmczmaom Nppwgm NFNUNwme ”pzoam “N53 ANzo NNNNHNV Ngmk om “so uwcgzp mmcwcp Nu: mzp ;pwz 30> wgw nmwmmNuNN 30: ADV .NzNNH NN< NON NSw> no: go n52.235 Ja::)§>7 u!)..7.)$) 5):).1. nll. i 3 324 H:72 NHV NNZszoo mowoo oLN co;o_w;o asp __oz 3o; .Nogocom CH ”IHHz wzH>H4 2me4HIQ LH ”:72 NHV Nkzmzzoo mooFN pom p:o;oo-ompw =zwc= ago new :NEN_NNU ago 30: ”IHHZ wzH>H4 ZMmQAHIo NH HA>z< NHV NNzszoo ANNLocwm :Nv omNNLLoE cw NW «Logp pcoezowco ocm :c:$= 5035 go: HA>z< NHV thmzzoo “_Ngocom :NV omowgcme c? N? woos“ ANNFNowmcooNoL =¥Lo3= nose 3o: Hso: NHV thmzzoo .opo aomzo; ogp ocooco Nxmmp co Nocpomow movxcoz ocoom so» oswp cogs 3o: .NF .0— .m 325 N F N m w N F N m v N F N m v N F N m w N F N m w N F N m w mm “CNNLoNEF poszosom pcopgomEF poopgomEF Ngo> uoz Appogo NFmeLuxm Amzo NANNHNV N=xLo3= o NFLLNE N :FNNE :F .NFcp NF Ncopgo EF 3o: on uA>z< NHV Nszzzoo mmzoom Lzox :sz ocoom so» oEFp ogszmF wcF nfl>z< NHV NNzszoo omzo; asp ocoogo Nooo so» Fo zoom zoos go: HS: NHV szmzzoo owproN pom mopzoNFo 3o: ufl>z< NHV thmzzoo 59Lo$ om ocm .Loo .mocpoFo .omoo; ocpuumcF>FF mo ugoocopm Lzo> u:72 NHV thmzzoo szFucF Loo» ;sz owcmconoFmL Lzo> U:72 NHV Nszzzoo ooFoco; NF Nazca osp 3o: If] ”ozonm Nag: 304mm mimFF aszo OF ammuomm -?’._r ,, ‘lfifica 1.10% L 3:. 7:14 IN..J~J..3 .. UWJNMQM in I C) 326 USE NHV NNZszoo mcFoo oLN :oLongo ocp FFoz 3o; .Fogocom :H ”:FFz wzH>H4 ZNNQAHIQ NH ufl>z< NHV NNzszoo acoFN pom “cocoonowpm =3oc= osp ocm :oLoFFco asp so: ”IFHz wzH>H4 zmeAHIu NF HA>2< NHV NNZszoo .opo nomzo; asp ocsogo Nxmmp :0 g 9.5783 ccwam 33>. 3:2,. :32: :5: .NF 327 1hen something doesn't suit you, are you the kind of person who ust comes out and says so, who drops a hint, or do you usua11y ot say anything at a11? 1 come out 2 hint 3 not say anything .uppose you wanted your spouse to start taking more responsi— ti1ity for doing something--such as getting the car serviced or loing the dishes. What wou1d you probab1y do? 1hat are the chances that your spouse wou1d go a1ong with your (ishes, after you did that? Very good, fair1y good, so-so ‘about 50%), fair1y poor, or very poor? 1 very good 2 fair1y good 3 so-so 4 fair1y poor 5 very poor 8 DK COMMENTS (IF ANY): f that didn't work and it was important to you, what wou1d you robab1y do next? 328 1 genera1, when you have a spat or quarre1, which of you ua11y brings up the subject: 1 husband 2 wife 3 both same 8 DK 7 never quarre1 (go to Q. 6) ___- :at usua11y happens then? IO usua11y comes around to the other's point of view? 1 husband 2 wife 3 both same 8 DK 7 OTHER ANSWER, DESCRIBE _— genera1, is your spouse an easy or hard person to 1k to? Why? 1__easy 2__hard you think it is genera11y hard for a woman to under- and a man and for a man to understand a woman? 329 Now would you please tell me who usually does the following things--husband almost always, husband more than wife, both about the same, wife more than husband, or wife almost always. (CIRCLE ONE ON EACH LINE) Husb. Both Wife more about more Wife the than Who usua11y: appointments doctors, tax Takes the to make up, when Straightens up the living room, when Decides how much spending money each of IF LIVING WITH: Decides about Has more to say whether or not to Has more to say about what kind of birth control to use, if any 1 2 3 4 5 8 7 Takes the initiative in expressing mutual affection (sexual) l 2 3 4 5 8 7 does (or probably would) make the final decision about the following: What house or apartment to take 1 2 3 4 5 8 Z L What car to get 1 2 3 4 5 8 i 1 Whether to buy some life insurance and how much 1 2 3 4 5 8 7 Whether the husband changes his job or not 1 2 3 4 5 8 79 Whether or not the wife should go to work A or quit working 1 2 3 4 5 8 7 How much money you can afford to spend per week on food 2 3 4 5 8 Whether or not a doctor is called when someone is ill 1 2 3 4 5 8 Where to go on a vacation 1 2 3 4 5 8 l7 330 At what age did you begin to date regularly? years old About how many people have you datedm more than just casually-- where you were real friends as we11 as dating partners? Would you say around 1, 3, 5, 7, 9, or more than 9? 1 or 2 3 or 4 5 or 6 7 or 8 0301-9me I 9 . more than 9 Of these, about how many would you say were ver close or serious relationships? (PROBE: including your present spouse} 1. _1 or 2 2. :3 or 4 3. :5 or 6 4. :7 or 8 5. 6. :9 :more than 9 331 THIS PAGE ONLY FOR PREVIOUSLY MARRIED PERSONS: 11. Think for a moment about your social life after you were divorced. In terms of dating the opposite sex, was it something of a disadvantage if people knew you were divorced? (IF YES. Did you con— sider it to be a fairly big disadvantage or just a small one 1 yes, fairly big 2 yes, small 3__po 8___DK 7___OTHER ANSWER, DESCRIBE: COMMENTS (IF ANY): After your divorce, did you have chi1dren living with you? l2 1 yes 2___no (go to Q. 13) 12a. IF YES: Was this a disadvantage, in tenms of dating? (IF YES: a fairIy big disadvantage or just a small one? 1 yes, fairly bi yes, small 3___po 8___pK 7_ OTHER ANSWER, DESCRIBE. COMMENTS (IF ANY): l3. know, we' re interested in some of the ways that people' 5 second marriages differ from their As you first. HAND LIST #2 to Please te1l me which of these are guiter different now, than they were in your first marriage. Just tell me the numb WRITE NUMBERS BELOW. IN LEFT-HAND COLUMN. b. What about number ( 1st tnumber)? Is this different now because you made special effort tto be sure it would be different, or did it just happen to work out that a. IF SPECIAL EFFORT: A special effort to find a different kind rson or a special effort to make sure that an understanding of Be could be worked out? c. How old were you, when you got married the first time? years old d And how long did you and your first spouse live together after you got marrie ed? years 332 Sgfle people think that married life is easier on women, while 9 ers say 1t 15 ea51er on men. In your own experience, does 1t seem that married 11fe is easier on men or women? 1___own sex 3___pther sex 2___neither (both the same) COMMENTS (IF ANY): What about your own marriage? (Would you say the same thing? Right now, who probably has it a little easier, you or your spouse?) l____I do 3 spouse 3 neither (both the same) 8___DK __ _ 15a. Why do you say that? Everything considered, is your life very different than it was before your wedd1ng, somewhat different, or not at all different? 1 very different 2 somewhat different 3 not at all different DK _— 8——. Does that mean that you did (didn't) live together before you were married? 1 lived together 2 not live together (go to Q. 18) ___. 9 no response to question (90 to Q. 18) IF LIVED TOGETHER: How long did you live together (before your marriage? 1___less than 1 month 6___J-1/2 years 2___1—3 months 7___2 years 3___4—6 months 8___3 years 4___7-lO months 9___4 years or more 5 a year COMMENTS (IF ANY): Since you were 18, about how many years have you worked full—time? years 333 Since you were 18, were there years when you only worked part-time? Yes No (Go to Q. 20) _— 1 19a. IF YES: About how many? years ———~— At the present time, do you work either full-time or part—fime for pay, out51de your home? (PROBE: IF YES: full-time or part—time?) 1 Yes, full-time 2 Yes, part-time 3 No (Go to Q. 31) IF A TEACHER, ETC., WHO USUALLY WORKS BUT HAS THE SUMMER 0FF, CHECK HERE [::I AND GO TO Q. 21. Does your spouse work the same hours you do, work different hours, (or is your spouse not employed)? 0 not employed 1 same hours 2 different hours _. COMMENTS (IF ANY): About how often do you bring work home, to do after work hours? Never, once in a while, about half the time, or very often? 1 never 2 once in a while 3 about 1/2 the time 4 very often 7 OTHER, DESCRIBE: When you get home from work, are you usually just exhausted, fairly tired, just a little tired, or not at all tired? l___exhausted 2___fairly 3 little 4___not at all 8___DK What do you usually do for the first hour or so, after you get home from work? Can you forget about your work once you get home, or is it still on your mind? 1 still on mind 2 forget (GO TO Q. 26) 25a. IF STILL ON MIND: What helps you get your mind off your work? 334 How much do you talk about your work at home? Very little, a fair amount, or a great deal? 1 very little 2 fair amount 3 great deal _— Do you tell your spouse only if something really interesting Eggpened, or do you tell about all the little things that happened 1 only if really interesting 2___Jittle things too (GO TO Q. 28) 1 27a. IF ONLY TELL INTERESTING THINGS: Why is that? Many people tell us that their work makes it hard to find enough t1me to do all the things they want to do. Do you find it difficult to find enough time to spend with your 5 ouse? (IF YES: Quite difficult or just a little? SPOUSE: l___No 2___Yes, little 3 IVes, quite difficult 28a. What about finding enough time for the household tasks? (PROBE: Is that difficult?) TASKS: l___No 2___Yes, little 3 Yes, quite difficult 28b. IF CHILDREN LIVING WITH: What about time to spend with the children? CHILDREN: 1_No 2__Yes, little 3[:flYes, quite difficult 28c. What about finding time to just be by yourself? SELF: 1___No 2___Yes, 1itt1e 3[:])es, quite difficult 28d. IF “QUITE DIFFICUL ” TO ANY OF THE ABOVE: How do you manage to find the time, when it's quite difficult to do so? (PROBE: What ways do you have of finding time for your spouse/children/ yourself/household tasks?) Do you tell your spouse about the more complicated or technical things you do at work? No 2 Yes Do you sometimes wish you could talk to your spouse more, about your work? _ 1 No 2 Yes 30a. Why is that? 335 1. ASK NON—WORKING R's: How much do you talk to your spouse, about what your day has been like? Very little, or a fair amount, or a great deal? l___yery little 2___fair amt. 3___great deal 31a. Do you tell your spouse only if something really interesting happened, or do you tell about all the little things that happened too. 1___pnly interesting 2___1ittle things too (GO TO Q. 32) 31b. IF ONLY TELL Why is that? INTERESTING THINGS: (GO TO Q. 32) K EVERYONE: How much does your spouse tell you, about his/her day? Very little, or a fair amount, or a great deal? l___yery little 2___fair amount 3___great deal Do you sometimes wish that your spouse would tell you more or less, about his/her day? 1___no 2 yes, prefer more 3 yes, prefer less 8___DK COMMENTS (IF ANY): Does your spouse help you with your work at all, either directly or indirectly? (PROBE: In what ways?) Do you help your spouse? (PROBE: In what ways?) 336 How often do you wish that your 5 ouse could 5 ° ' . _ p pend more t1me w1th you? Never, once 1n a wh1le, fairly often, or very often? l___pever 2_8—anK in a while 3___fairly often 4___yery often 36a. IF FAIRLY OR VERY OFTEN Have you talked about this at all? Do you think there's a chance to change the situation (so you can spend more time together?) ASK EVERYONE IN Do you consider that your own field of work and COUPLES WHERE your spouse's work are very close and similar IEEELWQB£§ to one another, overlapping to some extent, but with major areas of difference, or quite dis— tinct from one another? [___very similar 2___§ome overlap 3 quite distinct, different Which of the following comes closest to your opinions about women working: HAND LIST #3 TO R. _l Women ought to be able to work a bit but not so as to allow it to interfere with home and family obligations (SKIP TO Q. 39) Women ought not to work outside the home (SKIP TO Q. 39) Women ought to have the same opportunity to pursue an important career as men 8 DK (SKIP T0 Q. 39) F101, People can believe that women should have the same career opportunities as men without necessarily wanting that in N": their own marriage. How do ygu_feel about that, personally? (PROBET——Would you be disappointed if only the husband has a career?) Why is that? 39. 10. 337 From the list, what do you think your spouse's o inion is about women working)? p ( WRITE IN NUMBER: COMMENTS (IF ANY): TAKE BACK LIST FROM R Here is a list of annual incomes. They are numbered from 1 to 8. HAND LIST #4 to R. Which number represents the amount of income wh1ch you, yourself, earn in a year? WRITE IN NUMBER: (TAKE BACK LIST) You probably know some couples where the wife earns more money than the husband. If you were in their situation, do you think it would bother ygu at all? 1 yes 2___n0 (GO TO Q. 42) 3 OTHER ANSWER, DESCRIBE: .— 41a. IF YES OR OTHER: Why is that? Do you think it would bother our 5 ouse at all (to have the wife earning more than the husband)? 1 yes 2___ho 3___OTHER ANSWER, DESCRIBE: COMMENTS (IF ANY): What about your own situation? Who earns more right now, you or your spouse? 1___husband more 2___both same 3 wife more 8___DK _— COMMENTS (IF ANY): 338 1. Couples have different ways of handling their money. Do you have two separate checking accounts, a joint checking account, or what? 1 _2 separate 2 _joint 3 _joint and each has separate 4 _joint and husb. has separate 5 _joint and wife has separate 6 flhusband has separate account, wife has none 8_ NO CHECKING ACCOUNT 7___OTHER ANSWER, DESCRIBE: 44a. As you recall, who suggested doing it that way? l___husb. 2___wife 3___just kept on with way we did it before marriage 8___DK 7___OTHER ANSWER, DESCRIBE: People differ in what gives them their greatest satisfaction. HAND LIST #5 TO R. Please tell me the number of the thing which gives you the greatest satisfaction. WRITE IN NUMBER: a. Which gives you the second most satisfaction? WRITE IN NUMBER: b. And the third most satisfaction? WRITE IN NUMBER: 1E BACK LIST FROM R. RODUCTION OF PART B, SELF—ADMINISTERED QUESTIONNAIRE TERVIEWER: MAKE SURE CODE NUMBER AND SEX (husband or wife) ARE MARKED BEFORE PART B TO RESPONDENT.) there are some questions I'll ask you to answer yourself, by mark— the answers on this questionnaire. HAND PART B QUESTIONNAIRE AND ;IL_TO R. I'll be checking the other questionnaire, to make sure t my handwriting can be read, but feel free to ask me whenever you a any questions at all. INTERVIEWER: NEXT PAGE TO BE ASKED AFTER SELF-ADMINISTERED QUESTIONS DONE. 339 AFTER R FILLS OUT PAPER & PENCIL QUESTIONS, ASK: 46. Let's take a minute to talk about this last question you answered about who does what. We're interested in a little about h9w_you happened to work out things the way you did, and why. a. Let's start with "who does the grocery shopping. " First, let' 5 talk about th you happened to work this out this way. To give you an idea of what I mean, read through the top part of this sheet and see if some of these come close. HAND LIST #6 TO R, AND WAIT FOR R TO READ. Do any of these statements seem to describe th you and your spouse arrived at this arrangement? Can you tell me the numbers? Are there other answers which also apply? WRITE IN ALL NUMBERS MENTIONED. b. What about ”repairing things around the house”? HLw did you arrive at who should do this? AFTER lst ANSWER, PROBE: Anything else? (REPEAT FOR REMAINING ITEMS) ITEM HEN did you arrive at who should do this? What about: WRITE IN ALL NUMBERS MENTIONED 1. Grocery shopping 2. Repairing things around the house (furniture, toys, appliances) 3. Making complaints, if necessary, to salesmen or landlord 4. Keeping in touch with relatives 5. Deciding what people you will invite to the house 6. Taking care of the car 7. Preparing supper 8. Doing the supper dishes (or loading the dish— washer) IF THERE ARE CHILDREN: 9. Getting up at night, if necessary, with the children Helping the children with their homework — i 340 c. Next yfly do you do the grocery shopping the way you do? The bottom part of the sheet gives some possible answers. Take a minute to read through them. WAIT FOR R TO READ. Do any of these seem to describew __y.you do things this way? Can you tell me the numbers? Are there other answers which also apply? d. What about ”repairing things around the house"? Why_do you do it (divide it up) the way you do? PROBE FOR ADDITIONAL ANSWERS (REPEAT FOR REMAINING ITEMS) ITEM Why do you do it this way? WRITE What about: , _ IN ALL NUMBERS MENTIONED 1. Grocery shopping 2. Repairing things around the house (furniture, toys, appliances) 3. Making complaints, if necessary, to salesmen or landlord 4. Keeping in touch with relatives 5. Deciding what people you will invite to the house 6. Taking care of the car 7. Preparing supper 8. Doing the supper dishes (or loading the dish- washer) IF THERE ARE CHILDREN: 9. Getting up at night, if necessary, with the children 10. Helping the children with their homework TAKE BACK LIST FROM R 341 PART C - JOINT INTERVIEW INSTRUCTIONS TO INTERVIEWERS: l. List "H” for husband and "W" for wife before each comment recorded. Whenever a difference appears, push gently for a joint state— ment: “What do you both feel about this?” ”What is your opinion as a couple?” Rating will be easier if comments are recorded in the order that they were made, as verbatim as possible. GOOD NOTES ARE VERY IMPORTANT. Direct your comments to both spouses so that you will not influence which spouse answers. Interviewers should take turns asking questions, the man asking all ”M" questions and the woman all "F” questions. Work out your own system for writing down answers. In some cases, one person is better at this or has more legible hand— writing and will do all the recording. Or you may want to take turns, each taking one entire question. To start with, I wonder if you would tell us a little about yourselves. For instance, how did you happen to meet and get together? How long did you go together before you got married? Whose idea was it to get married? PROBE: Who was the first to get serious? Who first hinted at or suggested the idea of getting married?) RECORD EXACT WORDS: 342 We are interested in your opinions as a couple about several issues: For instance, do you feel that a husband and wife should each have their own circle of friends? RECORD EXACT WORDS: NOTE TO INTERVIEWERS: IF ONE SPOUSE PERSISTENTLY ANSWERS ALL QUESTIONS, BE SURE TO REMIND COUPLE TACTFULLY THAT JOINT ANSWERS ARE CALLED FOR. Who should be responsible for dealing with such outside people as the landlord, the banker, real estate people, and so forth? RECORD EXACT WORDS: What kind of relationship should you maintain with your in-laws? RECORD EXACT WORDS: 343 You know that we're interested in decision—making. I wonder 1f you could take one decision you made lately, or are in the process of making now, and tell us a little about it. (PROBE: Is there a fairly big decision you've made lately? Who first brought up the subject? What happened then? Where do things stand now? RECORD EXACT WORDS: All couples disagree sometimes. When this happens, what do you usually do? (PROBE: But sometimes talking is not enough. Sometimes one person just has to ”give in.” In cases like that, what usually happens?) RECORD EXACT WORDS: IF CHILDREN LIVING WITH THEM: I know you have children. What is one of the more important adjustments this has requ1red for you? RECORD EXACT WORDS: 344 IF EITHER SPOUSE PREVIOUSLY MARRIED: What kind of relation— ship do you think should be maintained with a former spouse? (PROBE: In terms of time spent, financial responsibility, e c. RECORD EXACT WORDS: There's a lot of talk these days about “women's liberation" and, lately, about “men's liberation.“ What is your opinion on such things? (PROBE: Do women or men need "liberating”? In what way?) 345 M l2. Finally, what would you say is one of the best things about your marr1age--someth1ng that gives you the most pleasure se 5 your marriage apart from man other ? RECORD EXACT WORDS: y S CLOSING SPEECH (USE YOUR OWN WORDS): We'd like to thank you very much for all of your help. It's been a pleasure talking with both of you. You'll be receiving a re- port about the study in late fall. Do you have any questions before we go? 346 INTERVIEWER'S RATING SHEET __.________________________ TO BE FILLED OUT IMMEDIATELY AFTER LEAVING THE HOUSE INSTRUCTIONS: l. Drive around the corner, etc., before filling out. 2. Each interviewer fills this out separately, with- out discussion. THIS IS VERY IMPORTANT. 3. You are rating the joint interview, where the couple was interviewed together. 4. Do not use ”both the same” any more than absolutely necessary. We are looking for differences between spouses, even fairly small differences. (CIRCLE ONE NUMBER) Husb. Both Wife Neither INTERVIEWER JUDGMENT OF: much Husb. the Wife much one does 7? more more same more more this " a. Who talks more? l 2 3 4 5 9 8 b. Who has more '3335' in joint decisions? l 2 3 4 5 9 8 c. Who contributes more ___ideas? l 2 3 4 5 9 8 d. Who shows more initiative (is less passive ? l 2 3 4 5 9 8 e. Who does more to smooth over differences? l 2 3 4 5 9 8 f. Who uses ”we” more? l 2 3 4 5 9 8 9. Who directs more questions to the other, such as ”what do you think?” l 2 3 4 5 9 8 h. Who points out more differences between spouses? l 2 3 4 5 9 8 i. Who ingerrupts the other more? l 2 3 4 5 9 8 j. IF INTERRUPTIONS, CHECK ONE: l___jnterruptions seemed to facilitate conversation . 2 interruptions blocked the express1on ___. of ideas k. IS THEIR CONVERSATION: 1 parallel, each going own way almost 1nde- pendently of the other 2 coordinated, each picking up where the other left off 3 OTHER, DESCRIBE: t. 347 FROM THE DISCUSSION, does the marriage seem more one of ”companion- ship“ or more “traditional” (segregated roles)? l___companionship 2___segregated/traditional CONFLICT (CHECK ONE): l___ppen conflict frequent 2___occasional conflict 3___suppressed conflict 4___ho conflict AGREEMENT (CHECK ONE): l___agreement after discussion 2___agreement without discussion CONSENSUS (similarity of spouses' ideas to start with): l___high consensus 2 medium _— 3 low ‘— Were gestures of hostility or affection made? CHECK ALL OBSERVED: Husband: [::I hostility ] (affection (__J ignored wife Wife : | | hostility [:1 affection ] :1 ignored husband l yes 2 no * Do they seem happy together, as far as you can tell? Compared to other interview couples, do they seem: l___happier than most 2___about average 3 less happy than most 7 8——— IF CHILDREN LIVING WITH THEM: . ‘ I. Who ”checked on“ children? T___husb. 2___w1fe 3___both 4___neither 2. Who did children come to or ask questions of? l___husb. 2 wife 3 both 4___neither 5___pn11dFEfi didn't come in 3. Who was irritated by children's noise? l___husb. 2__Jwife 3___both 4 neither 5 no noise Who answered the telephone/doorbell? l___husb. 2___wife 3 both 4___neither 5___no calls __—' or visitors 348 THUMBNAIL SKETCH: Briefly characterize the couple, with special attention to aspects of the husband-wife relationship which are not captured by the above ratings. INTERVIEWER EVALUATIONS/IMPRESSIONS ABOUT THE INDIVIDUAL (SEPARATE) INTERVIEW Respondent's attitude toward interview: (PLACE AN "X“ SOMEWHERE ON THE LINE) extremely cooperative,l + -| suspicious, hostile interested, eager Respondent's frankness/honesty: extremely open, frank + - did pg: give the l (impression of frankness Is the wife pregnant now? l yes 2___§eems to be 3___apparently not IF YES: Was she pregnant before the wedding? ANY OTHER COMMENTS ABOUT THE INDIVIDUAL INTERVIEWED OR THE INTERVIEW SITUATION: APPENDIX D LISTS ACCOMPANYING PHASE II QUESTIONNAIRE 349 #1 Buying large items (like a car or furniture or dishwasher) Fixing up your house or apartment Dividing household responsibilities Having or adopting children in the future Arrangements for children from a previous marriage (if appropriate) Schooling or education for children How much the wife works The type of job the wife takes (or might take in the future) The type of job the husband takes (or might take in the future) Moving someplace else How How 350 #2 the money was handled easy it was to talk over problems Similarity in goals, values How disputes were settled Maturity of both people How How How Who How How How How well you knew each other before marriage much the husband did around the house much you talked over practical things pefgpe marriage took more of the responsibility for making decisions much time you could have to yourself much the wife worked, outside the home well you could understand each other's work "practical” or ”romantic" your approach to marriage was Relationships with in-laws How heavy the pressure of financial problems was l. 2. 3. 351 Women ought to be able to work a bit but not so as to allow it to interfere with home and family obligations. Women ought not to work outside the home. Women ought to have the same opportunity to pursue an important career as men. #3 l. 352 Annual Income Under $3,500 $ 3,500 - $ 6,999 $ 7,000 - $ 9,999 $10,000 — $14,999 $15,000 - $19,999 $20,000 - $24,999 $25,000 — $29,999 $30,000 and over #4 353 #5 Your career or occupation Your relationship with your spouse Religious beliefs or activities Leisure—time recreational activities Participation as a citizen in the affairs of your community Participation in activities directed toward national or international betterment Running a home (If appropriate) being a parent Something else .wei_iv—W. _ __ .7. ii..‘-—. ~ \ 354 #6 HOW DID YOU ARRIVE AT THIS ARRANGEMENT? (SEVERAL ANSWERS MAY APPLY) l. I didn't give it any particular thought—-this way just seemed natural 2. We talked about it once or twice 3. We talked about this quite a bit 4. When I'm unhappy about this, I've let my spouse know 5. When my spouse is unhappy about this, he (or she) lets me know 6. We had a good quarrel about this 7. I had to do a fair amount of ”convincing” 8. My spouse had to do a fair amount of ”convincing“ 9. I just ”put my foot down“ l0. My spouse just ”put his/her foot down" ll. We decided that one of us would do it and that the other would do something else, in return l2. I've given it some thought, but it wouldn't do any good to try to change things l3. I thought I'd wait until later for a better time to bring this subject up l4. Some other way (what?) WHY DO YOU DO IT THIS WAY? (SEVERAL ANSWERS MAY APPLY) l. This is a IIman's job” or a ”woman's job“ 2. It's the only practical way, under the circumstances 3. The person who's pettgp at this should do it 4. The person who ljk§§_it more should do it 5. It's more fgg_this way 355 #6 (continued) It's easier than trying to change my spouse It's faster or more efficient this way This seems like the only fair way to do things The person with more time should do this The person who's more unhappy if it's not done (or done well) should do it This is only a temporary thing—-it will be different later I refuse to do it any other way My spouse refuses to do it any other way Some other reason (what?) PHASE II. APPENDIX E PERSONAL INTERVIEW, PART B 356 CONFIDENTIAL INFORMATION CODE: l974 Newlywed Study - Phase 2 Department of Sociology Michigan State University Husband Wife PART B - PENCIL-AND—PAPER QUESTIONNAIRE TO BEGIN WITH, TWO QUESTIONS ABOUT EMPLOYMENT: l. Whether or not you are employed right now, please characterize your level of career ambition over the long run: CHECK ONE: l___Very high ambition High __ Medium or average —. 4___Pretty low; no particular long-run ambitions I don't plan to work _— U‘lOJN 2. IF YOU HAVE A JOB NOW: In general, how do you feel about your present work situation? (CHECK BELOW) IF YOU ARE A HOUSEWIFE: In general, how do you feel about your present situation (being a full-time homemaker)? .; CHECK ONE: l___Extremely satisfied 2___Very satisfied 3___Fairly satisfied 4___Not very satisfied 5 Not at all satisfied AND A QUESTION ABOUT COMMUNITY SIZE: 3a. How large was the place in which you spent most of your high school years? (If you lived in more than one place, please check the 93g where you spent the most time.) CHECK ONE: l large metropolitan city (500,000 or more) __T2 Suburb of large metropolitan city 3 SW1 metropolitan city (100,000 - 499,999) __—7L___Suburb of small metropolitan city 5___City (25,000 - 99,999) 5__;Town (under 25,000) 7 On a farm 8 In the country but my family did not farm 3b. How large is the place you're living in flgw? WRITE IN NUMBER (SEE ABOVE): . How often do you and your Some questions about communication: 357 PLEASE CHECK QNE_ANSWER ON EACH LINE: Very Often Often Occasion— ally Seldom Never spouse talk over pleasant things that happen during the day? . Do you and your spouse talk about things in which you are both interested? . Do you and your spouse avoid talking about certain sub— jects? . Can your spouse tell what kind of day you have had without asking? . Do you and your spouse use words which have a special meaning not understood by outsiders? . Can you and your spouse dis- cuss your most sacred be— liefs without feelings of restraint or embarrassment? . You and your spouse are visiting friends. Something is said by the friends which causes you to glance at each other. Would you understand each other? . How often do you and your spouse talk with each other about personal problems? . Do you feel that in most matters your spouse knows what you are trying to say? . In general, have you and your spouse talked most things over together dur- ing marriage? 358 Looking at your circle of friends as a whole, how would you characterize some of their main values or interests? CHECK ONE ON EACH LINE: Dominant value or character- istic Secondary character— istic Not a character- istic . “Home—oriented”—emphasis on home, kids, garden, casual recreation at home, etc. . “Experimental“-emphasis on new life styles . “Sophisticated”-emphasis on cultural activities, more formal entertaining . “Community-oriented” -community and civic activities . ”Religious”-emphasis on spiritual values, reli— gious activities . "Intellectual" or academic I I l l l l I Among your circle of friends, how do most of them feel about women working? CHECK ONE: l___Women ought to be able to work a bit but not so as to allow it to interfere with home and family obligations. 2 Women ought not to work outside the home. 3 Women ought to have the same opportunity to pur- sue an important career as men. 4 I can't really say, since some friends have one opinion and others have another opinion. 359 Husbands and wives are usually different from each other in many ways. For example, even if both people are quite talkative, one is usually a little more talkative than the other. How do you and your spouse differ on the following? PLEASE CHECK ONE ANSWER ON EACH LINE WHICH ONE OF YOU USUALLY: HUSBAND EXACTLY WIFE MORE THE SAME MOREV . talks more? . has the quickest temper? . expresses appreciation for the other more often? . has more ”ups and downs” in moods? . is more likely to be ”on time?“ . is more likely to give in, when there is a disagreement? . has more energy (tires l—‘L‘1_fi less easily)? . dislikes arguments more? . tends to take the lead, in making decisions concerning both of you? . is more likely to say something, if unhappy with something the other one does? ll. shows more initiative? 12. does more to smooth over differences? . is more outgoing around strangers? . is more level-headed in a “crisis” situation? . has stricter standards for ”tidiness“ around the house? 360 xcozweo: gems“ new: ewee_eeo one me_aemz . O ,— cweu_w:u mzu new: .xgmmmmuw: we “gave pa a: meeeema .m “ZMmQAHIU mm< mmmzp mH flea; mzcuwu wzp mew inmo~ gov mwcmvu cwaazm wcu mcwoo . w Lmnazm mcwemawga . |\ Emu ms» we mLmu mcrxmh . KO wmzo; may on mu_>:w __wz so» w—aoma “on: mcwuwowo . Ln mw>wum_me :u_z :oaou cw mzwawmx <§ ego—w:c_ Lo :wEmm_mm op .acmmwmom: 4w .mucem—anu mcwxmz .m Amwucmw_aam .mao“ .mczgwcezwv wmzo: as» mzo xomzo ~mHIH muoa 0:: :kHz 30> mx< amHmmHk44<2w= oz: .w m nczocm mmcmzu mcwcwmamm .N H _ I iii- ii- mcraqozm xemuocw .— ” . umwemeumm vwewmwumm uwwwmwpmm umwmmeumm mco mzmz_m econmzz wEmm , wee; m>m3_m ran an poz umzszom »_wEmprM emgpwmz awoe_m coca wcu . cusp meEFw mew: _ egos . uaonm , mace ucmnmax mew: ;uom , vcmamsz ililiiiiiiIL xgozwsoc Leona :pwz eeee_eeo wee ueee_ez .o_ cmcw_w;o ago 59?: .xemmmmom: m_ .uzme: pm a: mewpawo .m ”ZMmQJHIu mm< mmmxk mH ALmzmmzcmwu ecu me? . iveow gov mm:m_u quazm mgu mewoo .m Lwaazm mcwcmawga .N 361 emu mgu mo memo mcwxmh .o wmzo; as“ op wu_>:v __wz :ox o_aoma “as: mcwvwoma .m mm>wam~wg cue: cuzou cw mcwamwx .v vLo—ucmp Lo :wsmm_mm op .xgmmmwomc em .mucwm_aeou mcwxmz .m . Lumzuv Hz o4203 or: .o_ newsman mEMm ease we“ wLoE #3030 wmwz cuom Ammucmw_nam .mwou .mezuwccawv wmzo; ozu uczogm mmcwcp mcweemawx .N L A ucmamaz mamz_m pmoe_m mcwanozm zemuoxu ._ BIBLIOGRAPHY BIBLIOGRAPHY Books Bernard, Jessie l972 The Future of Marriage. New York: World. Blau, Peter M. l955 The Dynamics of Bureaucracy. Chicago: University of Chicago Press. Blau, Peter M. T964 Exchange and Power in Social Life. New York: John Wiley S & ons. Blau, Peter M. and Otis Dudley Duncan.‘ l967 The American Occupational Structure. New York: John Wiley & Sons. Blood, R.O., Jr. l967 Love Match and Arranged Marriage. New York: Free Press. Blood, Robert 0., Jr., and Donald M. Wolfe. l960 Husbands and Wives: The Dynamics of Married Living. New York: The Free Press. Festinger, L., S. Schacter, and K. Back. l950 Social Pressure in Informal Groups. New York: Harper. Habermas, JUrgen and N. Luhman. l97l Theorie der Gesellschaft oder Sozialtechnologie. Frankfurt am Main, W. Germany: Suhrkamp Verlag. Hoffman, Lois Wladis and F. Juan Nye. l974 Working Mothers: An Evaluative Review of the Conseguences for Wife, Husband, and Child. San Francisco: Jossey-Bass. Homans, George C. l96l Social Behavior: Its Elementary Form;. New York and Burlingame: Harcourt, Brace & World. Hopkins, T. 1964 The Exercise of Influence in Smgll Groups. Totowa: Bedminister Press. 362 363 Kohn, Melvin L. l969 Class and Conformity: A Study in Values. Homewood, Ill1nois: The Dorsey Press. Komarovsky, Mirra. l967 Blue-Collar Marriage. New York: Random House. Leplae, C. l968 Structure des taches domestiques et du pouvoir de decision de la dyade conjugale. In Pierre de Bie et al., La Dyade Conjugale: Etude Sociologique. Bruxelles: Les Editions Vie Ouvriere. Michigan, Department of Public Health, Center for Health Statistics. l97l Michigan Health Statistics: l968—69. Lansing, Michigan: Department of Public Health. Mishler, E.G., and N.E. Waxler. l968 Interaction in Families: An Experimental Study of Family Processes and Schizophrenia. New York: John Wiley. Robinson, John P., Robert Athanasion, and Kendra B. Head. Measures of Occupational Attitudes and Occupational Characteristics: Appendix A to Measures of Political Attitudes. Ann Arbor, Michigan: Institute for Social Research, Survey Research Center. Rosenberg, Morris. l968 The Logic of Survey Analysis. New York: Basic Books. Steinmetz, Suzanne K. and Murray A. Straus. l975 Violence in the Family. New York: Dodd, Mead, and Co. Thibaut, J. W. and H. H. Kelley. . l960 The Social Psychology of Groups. New York: W1ley. U. S., Department of Commerce, Bureau of the Census. l97l U.S. Census of Population -ylg]0: Number of Inhabitants — Michigan. Vol. PC (l) - A24Mich. Washington, D.C.: Government Printing Office. July. Waller, W. and R. Hill. R . d d't' l95l The Family: A Dynamic Interpretation. ev1se e 1 ion. New York: Holt, Rinehart and Winston. Warner, W. Lloyd, Marchia Meeker, and Kenneth Eells. . _ l949 Social Class in America. Chicago, Ill1n01s: Sc1ence Research Associates. 364 Articles Back, Kurt W. l95l Influence through social communication. Journal of Abnormal and Social Psychology, 46: 9-23. Bales, Robert F. and Philip E. Slater. l955 Role differentiation in small decision—making groups. Pp. 259—306 in Talcot Parsons and Robert F. Bales (eds.) Family, Socialization and Interaction Process. Glencoe, Illinois: The Free Press. Berger, J., B. P. Cohen and M. Zelditch. l966 Status characteristics and expectation states. Pp. 29—46 1n J. Berger et al. (eds.) Sociolo ical Theories in Progress, Vol. I. Boston: Houghton Mifflin. Berger, J., M. Zelditch, Jr., Bo Anderson and B. P. Cohen. Pp. ll9-l46 in J. Berger, M. Zelditch and B. Anderson (eds.) Sociolo ical Theories in Pro ress, Vol. II. Boston and Palo Alto: Houghton Mifflin. Blood, Robert 0., Jr. l963 The measurement and bases of family power: a rejoinder. Marriage and Family Living, 25 (November): 475—478. Blood, Robert 0., Jr., and Robert L. Hamblin. l958 The effect of the wife's employment on the family power structure. Social Forces, 36 (May): 347-352. Blood, Robert 0., Jr., R. Hill, A. Michel and C. Safilios-Rothschild. l970 Comparative analysis of family power structure: problems of measurement and interpretation. Pp. 525-535 in R. Hill and R. K0nig (eds.) Families in East and West. Parisi Mouton. Buric, Olivera and Andjelka Zegevic. l967 Family authority, marital satisfaction, and the social network in Yugoslavia. Journal of Marriage and the Family. 29 (May): 325-336. Caputo, D. V. . l963 The parents of the schizophrenic. Fam1ly Process, 2: 339-356. 365 Dahlstrém, Edmund. l966 Exchange, influence and power. Acta Sociologica, 9: 237-284. Emerson, Richard M. l Power-dependence relations. American Sociological Review, l7 (February): 3l—40. Farina, A. l96O Patterns of role dominance and conflict in parents of schizophrenic patients. Journal of Abnormal and Social Psychology, 6l: 3l—38. French, John R. P., Jr., and Bertram Raven. l960 The bases of social power. Pp. 607—623 in Dorwin Cartwright and Alvin Zander (eds.) Group Dyngmjcs: Research and Theory, 2nd edition. New York, Evanston and London: Harper & Row. Gillespie, Dair. l97l Who has the power? The marital struggle. Journal of Marriage and the Family, 33 (August): 445—5l6. Harsanyl, J. C. l962 Measurement of social power in n-person reciprocal power situations II. Behavioral Science 7: 8l—9l. Heer, David M. l958 Dominance and the working wife. Social Forces, 36 (May): 34l-347. Heer, David M. l962 Husband and wife perceptions of family power structure. Marriage and Family Living, 24: 65-67. Heer, David M. l963 The measurement and bases of family power: an overview. Marriage and Fgmily Living, 25: l33—l39. Hoffman, Lois Wladis. T960 Effects of the employment of mothers on parental power relations and the division of household tasks. Marriage and Family Living, 22: 27-35. Kandel, D. B. and G. S. Lesser. l972 Marital decision-making in American and Danish urban families: a research note. Journal of Marriage and the Family, 34: l34-l38. Katz, I. and L. Benjamin. l960 Effects of white authoritarianism on bi-racial work groups. Journal of Abnormal and Social Psychology, 6l: 448—456. 366 Kenkel, William F. l957 Influence differentiation in family decision making. Sociology and Social Research, 42: 8- 25. Leik, Robert K. l965 Type of group and the probability of initiating acts. Sociometry, 28 (March): 57—65. March, J. G. l955 An introduction to the.theory and_measurement of influence, American Political Science Revigy, 49: 43l—45l. Martin, Roderick. l97l The concept of power: a critical deference. The British Journal of Sociology, 22 (September): 240-256. Michel, Andrée. l967 Comparative data concerning the interaction in French and American families. Journal of Marriage gpd the Family, 29 (May): 337—344. Moore, James C., Jr. l968 Status and influence in small group interactions. Sociometry, 3l (December): 47-63. Pearlin, Leonard I. l Status inequality and stress in marriage. American Sociological Review, 40 (June): 344—357. Raven, Bertram H., Richard Centers and Aroldo Rodrigues. l975 The bases of conjugal power. Pp. 235-240 in Ronald E. Cromwell and David H. Olson (eds.) Power in Families. New York: John Wiley and Sons. Rodman, Hyman. l967 Marital power in France, Greece, Yugoslavia, and the United States: a cross—national discussion. Journal of Marriage and the Family, 29 (May): 320-324. Safilios—Rothschild, Constantina. l967 A comparison of power structure and marital satisfaction in urban Greek and French families. Journal of Marriage and the Family, 29 (May): 345-352. Safilios-Rothschild, Constantina. l969 Family sociology or wives' family sociology? a cross- cultural examination of decision-making. Journal of Marriage and the Family, 3l (February): 290-30l. Safilios—Rothschild, Constantina. l970 The study of family power structure: a review of l960—69. Journal of Marriage and the Family, 32 (November): 539—552. 367 Strodtbeck, Fred L., Rita M. James, and Charles Hawkins. l957 Social status in jury deliberations. American Sociological Review, 22 (October): 7l3—7l9. Tallman, Irving. l969 Working—class wives in suburbia: fulfillment or crisis? Journal of Marriage and the Family, 3l (February): 65-72. Torrance, E. Paul. l954 Some consequences of power differences on decision-making Turk, l in permanent and temporary.three—man groups. Reprinted from Research Studies, State College of Washington, 22: l30—l40. James L. 975 Uses and abuses of family power. Pp. 80-94 in Ronald E. Cromwell and David H. Olson (eds.) Power in Families, New York: John Wiley. Wilkening, E. A. and Denton E. Morrison. 1 963 A comparison of husband and wife responses concerning who makes farm and home decisions. Marriage and Family Living, 25 (August): 349—351. Unpublished Materials Berger, J., and M. H. Fisek. 1970 Status characteristics and expectation states: an extended formulation. Unpublished manuscript. Stanford University, Department of Sociology. Berger, J., M. H. Fisek, and P. Crosbie. l97O Multi—characteristic status situations and the determination of power and prestige orders. Technical report No. 35, Laboratory for Social Research, Stanford University. Berger, J., M. H. Fisek, and L. Freese. l970 Paths of relevance and the determination of power and prestige orders. Unpublished manuscript. Stanford University, Department of Sociology. Burr, Wesley R., Lois Ahern, and Elmer M. Knowles. l9 An empirical test of the normative-resource theory of power. Paper presented at the Annual Meeting of the American Sociological Association. Freese, Lee. l969 The generalization of specific performance expectations. Unpublished Ph.D. dissertation, Stanford University. (Mimeographed). Kruse, R. l967 368 The burden of proof: A preliminary inquiry into the nature of relevance. Unpublished senior thesis. Depart- ment of Sociology, Stanford University. Rogers, Lillian Edna. l972 Tress, Paul l97l Trost, Jan. l969 Dyadic systems and transactional communication in a family context. Unpublished Ph.D. dissertation. Department of Communications, Michigan State University. H. Inconsistent status characteristics and influence processes: a replication and reformulation. Unpublished M.A. thesis, Michigan State University. Unpublished data cited in Safilios—Rothschild, l970. pedandl’rintedethe U. S. A. Professional Thesxs 011 ft and Paula Haughey 4 Maplewood Drive East Lansing, Michigan 4882 23 7. Telephone (517) 337-1527 'lIC II N 9 11111in mm L