\l‘) g lllfififlflififllfifi'fllfllfilfll 3 1293 02058 2056 1) O O O This is to certify that the dissertation entitled R; W #W [0er S+ra presented by Jean/n, kHHfi/NM has been accepted towards fulfillment of the requirements for ,$0 c‘f‘h’Weflee in SQQIUQQ-y : mat" I Major Erofessor Date M "cl”.--AL - A ~ m 11L 042771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2003 110209 11m GICIRODIt-Dtnya-pu FAMILY AND HOUSEHOLD WORK STRATEGIES: NUMBER OF EARNERS AND SELF-EMPLOYMENT IN NORTH CENTRAL LABOR MARKET AREAS By Jean Kayitsinga A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1999 mGMYJIMA‘I ~‘ mmmnom i if?" " 375‘]: . ' r4..._1('>f5.u- _ ABSTRACT FAMILY AND HOUSEHOLD WORK STRATEGIES: NUMBER OF EARNERS AND SELF-EMPLOYMENT IN NORTH CENTRAL LABOR MARKET AREAS By Jean Kayitsinga Families and households have adopted a wide range of work strategies in response to social and economic changes in the last two decades. This study examines self-employment work and employment of family and household members other than the head of the household, the spouse or partner. This study relies on the 1990 Public Use Micro Data Samples (PUMS-L) Labor Market Areas and the 1990 Summary Tape Files (STF3) data. Conceptually, economic restructuring, with its uneven impacts on families and households across spatial locations, and social embeddedness/social capital perspectives are used in order to better understand how families and households adapt and respond to the social contexts in which they are embedded and interconnected Analytically, this study uses a multi-level framework that considers a combined effect of labor market area, family/household, and individual factors. I find that household work strategies vary within and across different labor market areas, between non-metropolitan and metropolitan labor market areas, and across different households depending upon the family/household as well as labor market area social structures. The spatial location, social and economic structures of labor market areas in which families and households reside, and family/household social structures affect family and household work strategies. This study contributes to research that attempts to link micro-to-macro phenomena of family outcomes. xenon (MA ’IIIIMA"! » 3-1133 (“IA 28'1““ '10 “38%th , Jinan _. ' . - we r" . ‘ "-' -" 1.: ,j‘ | '1 i ' ,~*'- .. 3:1 ~.Q 2' , Q , Want n".f'. .‘ l ‘I ..', m), . h '7; C ('1‘ 1 A : i ‘l ‘1’." .. ' ,- I. .‘KI- . l. k '7' "Cl. . .34 .‘ F - , ‘ . 7.“; _ f , “a“. wall ablodsworl hm: amlimlfl ‘3‘ "A 1“ Q “ '.-‘, . " Q: I. it ' ain‘t-lo dialectical)": laiooa or senoqzm ‘ 4»! .3"- , . an . a _‘ I = Whoa how tmmvolqma—“tlue . ,1 g j" l ant-q to more art: Mailman 9mm band i Q W (J—ZMU‘I) Home? c1130 019M .7 chance .‘(llnmqaono'D stab l HT?) . W cums abltiliuzuort Om. . « or who in been me esviwuqmoq dim Iniooa am 01 Mona-)1 om; ' - ~ . I can (but: aim .‘(llaonvlanA - mm .8910 ram wrinl WWW»: Trev eaigeiam I. dun ml rmiloqanam hm “wash-u as Wuhan: -“u . DEDICATION For my sons, Olivier M Kayiranga and Cedric K. Kayitsinga for their love, joy, and fim times throughout the writing of this dissertation. I also would like to dedicate this dissertation to my parents, Deogratsias Sebinyogote and Alivera Mukakarangwa, who always valued, encouraged and worked hard for the education of their children, and to my relatives, Edith Kayitesi, Jean Baptiste Kayitare, Judith Kayirebwa, Bonaventure Karangwa, Jean Claude Kalinganire, Marie-Chantal Kayirere, Kamatali Jean Bosco, and Kangwaneza Anita who all passed away during the 1994 genocide in Rwanda You will always be remembered . '._, c i . - - ' A r .fir'fig‘»..-_;Q. u ' , - drawn»; . - .- ~..,A .7... 7,_ a-“ with at ACKNOWLEDGMENTS This dissertation has been successful because of the encouragement, support, patience, and expertise of many people. First and most important, I would like to thank my major advisor, Dr. Janet L. Bokemeier, for her guidance, support, encouragement in all aspects during the writing of this dissertation and for her indispensable support during my graduate student life. I would like also to thank Dr. Bokemeier, Principal Investigator on the Regional Project 5-259, funded by the Economic Research Service, US. Department of Agriculture and Agricultural Experiment Stations and land-grant institutions afliliated with U. S. D. A., for granting the permission to use the PUMS-L data I would like also to thank the members of my committee, including Dr. Stephen W. Raudenbush, Dr. Stan Kaplowitz, Dr. Nan E. Johnson, Dr. Brendan Mullan, and Dr. Thomas Conner for their guidance, suggestions, and comments on this dissertation I am very grateful to the Department of Sociology at Michigan State University for their financial assistance, encouragement, understanding, and support. To all graduate students who supported me, especially in difi'rcult times, including Jeanne Lorentzen and Elaine Marie Allensworth, Thank you. Finally, but not the least, I would like to thank my wife, Gaudence Kayitesirwa, for her love, encouragement, support, understanding, tolerance, and patience. Without her social support in all aspects of our family, I would not have the daily energy, persistence, and enthusiasm in writing this dissertation. araamtuwou'u Warm almond lu’taaaouue need ml commit: airfi‘ ., When has mfl .alqooq vnam’lo aeiqua bars 30qu ' .; urinal :Iaiameilofl ..l renal. .10 ,weivba 10mm a buMuoth'iomumuqoflEU j 103..A.(I a .U ritiwbcm‘rli'fiamohmnzni .mb .. airworoeIeazlilbluowl .‘rdmiwomwarflfizndnabmfl W . T 7 I. ___W , J 1 M.Mflm3mn r: c "7 1. I ", . 2"} .. 1M§ “$7.94" 7‘ ‘ . . .0" j ’ .'-'z' . ,5 1, I ~ ‘ -.-bmfluhmhflmramv . I44“. Eu w; ‘. : ' _ '- " memmma I. .. 13-... . - -:.'-' "V. ~ o _ . ,' T “‘5 ‘n . ' ‘r'_-’~ .. (“1"; ‘ 1" TABLE OF CONTENTS LIST OF TABLES vii LIST OF FIGURES viii CHAPTER 1. INTRODUCTION 1 Data 3 The Research Prohl em 4 Family and Household Work Strategy 7 Conceptual .‘Tr L 8 The Concept of Strategy 8 The Concept of Work 11 The Concepts of Family and I' L ' ‘ ll Methodological .‘ n ‘ 14 Factors Associated with Family and Household Work Strategies ..................... 16 Summary and Implications for the Study 19 CHAPTER 2. INTEGRATED THEORETICAL AND CONCEPTUAL FRAMEWORK .................... 21 Restructuring Perspectiv 22 Restructuring and Self-emplo, ‘ 24 Restructuring and Multiple Farmers 29 Spatial Aspects of Restructuring 30 Changes in Families and L L ‘ ‘ 34 Social Embeddedness/Social Capital Perspectiv 36 Social Capital Theory 37 Social Embeddedness Theory 46 Summary of Research Propositions and II, , ‘L 54 CHAPTER 3. RESEARCH DESIGN AND ANALYTICAL STRATEGIES 65 Data 65 Variable Construction 67 Analytical Strategy 80 Multilevel Models 82 Multilevel Models for Discrete Response Data ......................... 84 mm W30 MAT ...8’;1..IEAT iOTZU ..... ...............83510i)1"1 rorau .1 513m . . .......VI01TDI.IGO5TTKI “M mldrn‘1 dmaaesfl arIT Mara.~.....w m mar bill (tiara-l ‘ 1% W3 W.» WibgvmaDwi .M .Imjrwa :4!” NWT Wm! ‘ WWI has Mime S m MW GETAflOEI‘mI ”W CHAPTER 4. RESULTS Descriptive Statistics Family and Household (‘L ‘ ixiiCQ Labor Market Area (“L ‘ Mics Factor Analysis of Labor Market Area Characteristics ........... Bivan'a .A--';~‘- Family and Household Characteristics and Self-employment ............. Family and Household Characteristics and Additional Eamers .......... Correlations of Labor Market Area f“ ‘ ish'cs Multi-level Discrete Models of Self-employment on Family/household and Labor Market Area I“ ‘ Mics Multi-level Discrete Models of Additional Earners on Family/household and Labor Market Area (‘L ‘ Mics CHAPTER 5. SUMMARY AND CONCLUSION Summary of the Results in Relation to Social Embeddedness/Capital Perspectiv Summary of the Results in Relation to Restructuring Perspective .................. Policy ' r " " Limitations and Future “ L REFERENCES APPENDICES 88 88 88 96 99 101 101 105 109 114 132 143 144 153 158 161 168 187 Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. LIST OF TABLES Summary of Propositions and II, r " Descriptive Statistics of Selected (‘L ‘ Mics Family and Household Characteristics by Non-metropolitan and Metropolitan “ “ Labor Market Area Characteristics by Non-metropolitan and Metropolitan “ " Oblimin with Kaiser Normalization Rotated Factor Pattern (factor loadings 2.60) in 112 Labor Market Areas Percentage of Family and Households Involved in Self-Employment by Family and Household (‘L ‘ Nam Percentage of Family and Households with Additional Eamers by Family and Household (‘L ‘ km: Correlation Matrix 0f Labor Market Area (1“ ‘ i\iih\ Multilevel Model Estimates of Self-employ Table 10. Multilevel Model Estimates of Additional Famers Table 11. Summary of Findings vii 62 91 97 111 114 132 156 : MAT 10 'l'dJ t, ’ "lbnaano'uizoqafl lo-{wnmua .l slle F’ _. ' ' 1 v, 7_ WAWBZ “to examine avm'rmaa .S alfi'l' " : «26 23W Nominal-l ban (lime?! C old-T ; ' - _.;...‘......... ....... . ........ “Mini! nwiloqomM ,_ { . NQW'JWmI-MMW bald-T ’. '3’" ‘ . ”mailman“ Mil miloqamM " H ’ ' ' m ”inlaid: quill-almond) MWWOW .6 SHOT ,Wbmvfim‘l Milan-110mm UNIT -~ v Whitman! rhu'lnximM memo .s dds‘l‘ - rm mum .9 aldfl‘ ‘ 710064me .01 slde’l’ awn own»?! .1: aid-T Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. LIST OF FIGURES Map of Labor Market Areas under Study Multilevel Discrete Model for Self-emplo, ‘ Multilevel Discrete Model for Additional Famers Number of Earners Within Families and 1' L ‘ ‘ Percentage Distribution of Family and Household Work Strategies ............. Percentage of Families and Households with a Self-employed Member by Income Categories (000's) viii 86 87 89 90 105 1‘ (ll (8 08 09 2m 41an H" :r!. VN'JTllIU.‘.'(I'J.1*1 _. -... x.;wm:‘l 'm'l'~'v'ibz ‘ : "10(1'3'Ab'lll”115/41“1.... u 29:331th 410W Linda/1m” tun,- vlvms ' 1; mum iii/5.11 agonuna‘i (d iadmsM b‘J‘tOkYmU-llsa IS dllu 9.5, '0 l'u~“"""’1fwir'..i "mi N» u’ru‘v 'vv‘n‘i n'rJ ‘ 1'4' {Lyn ‘ 'n ‘2”.an ,V 'Afi a.- i .flugI-l 1 3m?! 1 7, mum t— {1111in ) . 3111;." (‘5 tinsel-l Chapter 1 INTRODUCTION In the last three decades, changes in the structure of the economy have had implications for the survival and economic security of rural workers, households, and communities as indicated by escalating levels of poverty, unemployment, and underemployment in rural areas, increasing housing costs, insecurity of marital relationships, and migration within and between communities, particularly that of younger educated adults to urban areas (Fitchen 1991). Non-metropolitan families and households have adopted a wide range of adaptive strategies, such as combining work on farm and off-farm, paid homework, paid child care, self-employment, unpaid family work and housework, increasing the number of workers per household, seeking multiple sources of income, and traveling long distances for better job opportimities in other places. This study examines the different family and household adaptive strategies of non-metropolitan and metropolitan families and households with a focus on the number of earners within households and self-employment. The main objectives of the study are (1) to examine family and household adaptive strategies focusing on household and family work strategies; (2) to determine how these forms of household work strategies vary within and across different labor market areas, between nnal and urban labor market areas, and across different households given different family and household structures, and 3) to determine and l Titty-11') K011" )‘JflOfil'il MOM {mama 311130 mulomfi; um n! angina-rig 'ebnaeb ‘Jfllll wzl ed! M mm m lsun'lo '(tnuvw. Olmum‘nu bur, lam Jib mt mortamlqmi NOW .Vrra‘mq to 21') 131' gum. pa 2: {u hammer: ki. 4:; v'nurnmoo WNW $300 gnlworl gmamr 1m 4m: . imm m mamyolqmmebtw WW. -GW muted bun uulrm nmmrmm bun aqirizrroitsbt WW .0991 nsdau' balsam rusdrrv oz atlnha betaoubc ”gram m whqaba'to 339131 55m 8 Intact»: aim! eblodsmod m Nib biaq ,ahowamofl bmq ,xurv;l-tlu bna mil to Brim all! gniaammu .flrwaworl but: how W anal gnilwrm bus ,amouu; )9 names ' “We!!! minim: (but: an” .330an brim minnonom bns riariloqouem-noa was: has ablodezuod mum, momma infirm-ethic mum mm at” —W no airman? gamma wands ,, _ an. _, WmmmmhmmWa Wqfimmhm WNW assess the relative importance of factors associated with family and household work strategies. A variety of factors are associated with family and household work strategies. At the macro level, the economic opportunities and their spatial distributions, mainly due to the recent widespread transformation in industrial organization and labor as well as social and demographic composition of different labor market areas, constrain and enable the employment and self-employment opportunities of families and households. At the micro level, family and household social structures, their composition, the life course stages, socioeconomic resources, and other family and household characteristics such as the geographic mobility status and educational levels of their members influence the employment and self-employment opportunities of families and households. The link between macro and micro factors describes the interconnection of families and households with labor market areas (places in which they live and work). Two core questions of the study are (1) To what extent family and household social structures affect the employment and self-employment of members? And (2), how does the labor market area social structure enable or constrain the employment and self-employment of family and household members? A general assumption is that the family/household social, economic, as well as cultural capital are crucial factors for comprehending the diverse social contexts that permit family/household members to adopt any work strategy. At the same time, the way the family/household is situated in broader ~bow blorlseuorl bus vhmn't thiw hammer am us! ”to eonanoqm‘: evrmizn art: was Jaw 1A 39w WHOM“! bot.- ‘(lrma't dim batsmmr. urn (10135110 ‘(bnav A ofwbmmufilb Wage tied: baa aamrwmqou .ummooe ail lSVul 0mm ad! WW” We lumubni ni normanulenw bmeebm maoet ad! Intbl mm to nomaoqmoo autqargomh baa bison M mandamus; baa :namvoiqma orb aid-n larvae blurlawoll has '{lmw't .lwal oiaim all: M 7' “ ‘ - ‘ .-"A 1r. " "i ' . , '1' o. . . . -' ‘ 7"“ a. ' . ,. ‘w .M“ mummmmammmmm aw". «w. a “9., -.. it -- , ._' ”0:33.32 .1)“. 1w.” . 'L" '0 f” - , ‘ as . I, , ‘ > ;.. . ~ . ‘ ‘ V r-,-¢;ra.'»'?i~~'.ir. ‘ ..-. ’ _ .. , bum (Idrdom sumac! am am . 4.3.3.. tfi-c "JD-12$?"Hl‘ "1‘ ~. ‘ . . ~ 7,...5‘. .._ 5r ‘1. . , ‘ ‘ _ '1‘ "MA- '1 . ' I , ‘ . - ... v MAM-193598 hm muiohrmdl .'§"""'.; '1‘ .5119!" - {35" ‘ ' ‘ ‘ .dineabnona'tmarmbnamwnaeemd v ,- -mmwm'mabrodamq vacuormmthmw’hm mmhflrmmm A ruin—ammonia Jinn-pm.“ environmentI affects the “strategies” its members can do and adopt This study builds on previous research that integrates multilevel theories of social phenomena and analytical strategies. Specifically, this study combines two theoretical perspectives relevant to the understanding of family and household work strategies: (1) the restructuring perspective, and (2) the social embeddedness/social capital perspective. Analytically, this study uses a multilevel framework looking at how individual, household and labor market area factors intersect in association with household work strategies. Data This study uses two data sets: The 1990 Public Use Micro Data Samples (PUMS- L) Labor Market Areas and the 1990 Summary Tape Files (STF3). The PUMS-L is produced by US. Bureau of the Census and funded by the Economic Research Service, US. Department of Agriculture and Agricultural Experiment Stations and land-grant institutions affiliated with U. S. D. A. Regional Project S-259. Labor market areas (LMAs) are groups of counties that encompass the county of residence and the county of work. The labor market area PUMS-L data provides an excellent sample for linking individuals, households, and labor markets areas, thus permitting the incorporation of multilevel factors in models of household and family work strategies. The other advantage of using the PUMS data for labor market areas is their coverage of both rural and urban local labor markets. For this study a sub-sample of 112 North Central labor market areas (LMA) is first selected. These LMAs are groups counties mainly in the '1hetum“arvironman”isusedmmyinstanceammfatodreecomnnc,sociaL political,and cultural contextsthatconstrain, guide, orenable individual actions, practices or representations. 3 no W (hm eid'l Jqobs bun ob use madmam an murmur" ad! 2133115 'rmmrmiv. mm‘Mmmomdq lsiooaio aaimrh fwainlurr: Aftf‘f‘fltli 1.. il :lmlm‘l tum ammraavituaqzuq 18303103!“ um emid’nu'r vbuv 21m (lint-Blow}? .nigm w") Laaigaaa'.re )lrow blooms-mi burr (mm; '1' umbmum 7 oil ,laubmbm won tu gonna-l Arowm‘nmi lavelblmn I s .f': 2;... ‘ ' m hora blodaworl miw amnion?" ni tsunami m .lli’: ':,=_v '. I m a " WOOUWWI m .atumuchoeu mum .) 1‘1 mummcwvi mrbrmem‘lAbriuM 1060.1“ . 1 ~_,‘. ' if . “WWW ad: in uaarufl .‘CU ‘(dbaauboq * i .‘rr' 5:. ' H ' V ‘ I ‘- MA'm‘mmangA'to rmrnmeflfiu . ‘ .»-;‘T-"‘:;”‘r7-..«’m “ )1 ” ' E3351" ‘Kzé'mi-I" ‘Y, ":3 - ..r‘ '3' ' '1? 3.,-—~_ ' L‘ ‘ ’~ ~ "r ' W5 .A ..'(l Urtuw mama anonm‘rtaai , 1 Vfimflmltommtam) A, Ifiz‘gfi‘ .5711: . ..J.‘ .1 .‘ :r- -. ‘ I. ,5; l: . . . . "F wmmwamu . ‘ ‘ . .. hmnrmlmmm it " ' mmmmmmbm new madallmlmu- states of Michigan, Ohio, Indiana, Illinois, Wisconsin, Minnesota, Iowa, and Missouri. The boundaries of labor market areas are not limited to geographic lines of these states. Some counties are fi'om the neighboring states including Kentucky, West Virginia, Arkansas, North and South Dakota, and Nebraska. The major limitation of these labor market areas is the Census Bureau confidentiality requirement for any geographic identifier on the PUMS to contain a minimum of 100,000 people (See Tolbert and Killian, 1987, for greater details). A sample of civilian working-age individuals (age between 16 and 64) is selected from the PUMS-L data Four types of households are selected: Married families, male and female-headed households, and cohabiting households (unmarried persons living with their Mes). Excluded in the analysis are farm families, households of living alone individuals, as well as group quarter/vacant households. This corresponds to 64,356 households with about 35 percent (22,790 households) living in non-metropolitan areas. The summary tape files data from the 1990 census (STF3A) is used to construct characteristics of labor market areas. Data fi'om the summary tape files is at the county level and is aggregated at the corresponding labor market areas. Characteristics of labor market areas include the opportunity structure ( i.e., the availability and types of jobs), residential stability, social and demographic characteristics, and social inequality (i.e., poverty, unemployment, income inequality, public assistance). W In the last three decades, major changes in the economic processes, labor force composition, and state activities, known as “the restructuring of the economy,” had AmoulM ban ,svm! .3104;an mann'mw 2:3:er ,wn'Lvu wk) rmmrioaM'lo m £33832 MN) can“ niflqnscmg 0! bsimuf :03: we. . . 33;; 331mm aodsl '3” mambmmd afl' WV MW (1503:3031! gnlbulam?‘ 22.733, .394"an an H“ mm’! 3m wanmsno smog “Minnow-swims: W 32231de cm .sioinU (umk‘. hm.- de .mnlilA 9W [£8101 Wm ‘(uiwmbhr mmmd 4152233953313 23 am tom WOT-mm 000.00l '30 mummim s msmou 01 QMU‘! am no minim “WWW nnlhvw to 31mm. r. r'. (41313331: 1335913 1n: WWI ,uliui)! humane! nubJ-ZMI m :mon uuxr . a.) a. . m m M nucwtad Mum? hm mam _. whmn’l baan 711“” MIN draw gmvil manor; bomsmnu) ablodowod . ‘ .m 80h! gain! 10 ablm'uauwi ,eudims’t all} .mbtfibo m atmoqeanm an” 2le ..' MiWnon m gmvil ( awodawod ' -' .3ufim Mil :qa mmmnz at” 0 8m ”alum wads! '30 2902313333”: aim Maw 23 has W W)” M ed! shut-mi 23:13 min!!! W in hiooe ,mltdm 3333mm”: important effects on the socioeconomic well-being of a large number of workers, families, households, and communities, particularly, those in non-metropolitan areas. In the late 1970' s and 1980' s, employment in agriculture and extractive industries and in manufacturing industries underwent sharp contractions while the service sector has increased However, the rise in service sector employment in rural areas has been limited to lower-wage jobs while the urban areas have experienced a rise in jobs at both _ ends of the wage spectrum (Gorham 1992). These changes have not only exacerbated the existing spatial disadvantage of rural areas (Tickamyer and Duncan 1990), particularly the economic survival of many families and households in rural commrmities, but also they have opened new horizons for making a living, as revealed by a growing diversity of working activities. The net and profound impact of the recent economic transformation and long term structural changes on families and households have been the persistent and increasing lack of secure full-time jobs and an increase in numbers of workers with average levels of education and high job expectations (Mingione 1991). Recent research indicates that poverty, imemployment and underemployment have increased while the earnings have declined (Lichter et al. 1994; Whitener 1991; Jensen and Eggebeen 1994). Moreover, the impact of the ongoing structural changes on families and households is unevenly experienced across spatial locations (Rural Sociological Task Force on Persistent Poverty 1993; Lichter and Constanzo 1987) and depending upon their structure, class, gender, and race (Tickamyer and Bokemeier 1993). .amihow'to 'rsdmun 93ml 5 to gmsdollsu slum.» mom» at“ no men-i inmoqni nl .2801. wfloqonsmmon at 930m fighnluraahar’ your": .. my; hon ,ablorbwod ,eoillnfl 9 nibnsmmnvhmmm bus 91011114ij to ' :‘amvr Hm , ‘10! b .n :- '0\‘Ql era! alt mmmw elirlw autoimmune quilt mam-slim: eermirbni gnhumilunsat use-glam [can at lmyolqmmae soivxw m 29h em ,15‘13‘1'0H1bl8m mqums 31nd ants "arm! 31? elirlw rlcr vgaw'urxml ()1 betimil a}. ." J Mend-made out" ($09! rntuhud; mmueqe 538W Mt'lo dim Ila-flint. maxim l') «tee-w liner to agatnmbeub lrzuaqe 3mm ,W:thrl has aellkmal mam lo lmmuz mmofloas ed! Wit 8 gaiilem 10" meshed awn bemrw avail m .eemvnaa gauhow 'l eth'lo 331mm bmrolmq has ran at” M be: asthma no with linoleum mm 4;.._—...—-4.“— -. Ih'dul stub-M W10 )L'étil gmamani ‘ fitment-1mm .mmlneuathai M Jan WJweuilaeb and again” 2 cm," v-wr 17w. WV." '-- Wefl‘b swarm at: ,uvoeioM Thus, it remains important to understand how the ongoing, complex mixes of economic and social changes contribute to the levels of poverty, unemployment, and underemployment, and to financial instability of many families and households, particularly those in rural areas. It is also crucial to understand how families and households can adapt to variable social contexts surrounding them, maintain and secure their own and their children's future. Despite a large volume of research on changes in macro environments and their impacts on households and families, more research is required to understand how families and households survive and adapt to those changes. One fundamental social change in the last two decades is the increasing number of women in the labor force. Women' s participation in the labor force, particularly married women and mothers of young children, has increased dramatically’. Despite this trend, family income has continued to decline’. This change in farnily income is mainly associated with the lack of secure and good-paying jobs. As a result, a higher proportion of families and households have increased the number of earners while some even hold more than one job. In particular, a previously non-working spouse, a partner, adult children, relative and non-relative members may enter the labor force in order to cope and deal with changing environments surrounding them. In addition, there is evidence that self- zBetweml940andl990,theproportionofbreadwinner-fiil!‘ ‘ ' “ “'fmme'm-t 70pu'centtoabwt20percent. In1990,65percentofmotheiswithyoungchildrenunderlByearsofageand 53 percentofmotherswithchildren underage3 wereworking (Skolnick 1995). 3Skolnick(l995)showsthatthemedianfamilyincomehasdramaticallyincressedbetween1947ar|d I973. Hammmemedhnmwmufmaflfinnnuandnmfiedmplefimihaamaimcedmpaiodsof deelineandtwoperiodsofinereaee. Despitethesefluctuafionsandtheincreaseoffanalelaborfixeabylm, themedianfamilyineomewasoalyllpercentgreaterthaninl973. 6 W ' )9 'm comm xelqmoa .gmogno 9dr won b.’lfll21'»bfll1 or ivmnoqmv cnlufil‘fl tr ,aurfl‘ baa Jmmvolqmena .vhevoq'lo View! em 0; _.;._;.!mm :3 auras-J lsieoe baa aim .zblodeouort has aetlimu't mom in In: 4mm Irammn-‘t m but: ,memvolqmm inseam WM Wuhan OI Imam; uwlt' at Al mam. lmm n: 39'de melanin-1 mW~Wm .aled! anibmumue vii/91mm lat ,. we Julmm f-Ilalu: ram ablatieeuod email u’tmblvris nod: has nwo lid! WWW!!! ui aegnarb no rtzrmewt to errmlo I 33ml a :11th 'Qho't'apet ab claim-.1: mom .auiturw't has Mediation no 213M ‘m and! at not» has av! [we ablrxteauod bus 29th ‘ ”in!!! minimum ed: at cabin-J; ow: mil '3!!! m sands ‘ ”. wooing ,emo’t ma all! n: musmmmq a sum: .‘(iluiamb healers“: and .rmbliris 3mm 03“!“ at w ain't fault-Job m bounnuoo Mam Managua-boos has maeflo m w WI avail ablodewod W s .ialuo'meq nl .dal *mwn autism Waterman baa m mime gnignerlo --,1 n (4' L ‘ .. 1,1-';-¢,,:f.‘.;:- 'r-v; if" or; .,,.. m x“d‘."§"n Til-yer“: ’4», '~ ,erk-‘al‘ew tarewee w. ., “'2‘ . _ ”V." i n; J- “1'" “1:1 " j’i'j' ." ' ~ W” -> 4~1~W’#&xqar.¢n~g a employment‘ activities, unpaid work, home—based work, (Gringeri 1994), batters, and illegal jobs are increasing. Some families and households, especially those living in rural areas and those in and near poverty, may be engaging in self-employment activities in order to get by and improve their financial situations. Other families and households may engage in self-employment work and/or receive a self-employment income because of the changes in high technologies, especially in microelectronics or because of challenges to businesses from international competition (Steinmetz and Wright 1989). This study examines the different work strategies of non-metropolitan and metropolitan families and households, with a focus on the number of earners within families and households and self-employment as family and household strategies. We! The concept of “strategy” has been used in social science to refer to individuals and groups’ conscious and rational decisions for immediate needs and long term goals and describes the logic that people use when they allocate scarce resources such as land or time to different activities such as self-provisioning or paid labor (Garrett et al. 1993 ). In many cases, this concept has focused on families and households to explain the ways in which they confront, cope, deal with, or overcome changing structural contexts of their environment in order to satisfy their needs. In particular, it has been applied to families and households that are poor or experiencing economic difficulties (Crow ‘The ' fself-emplymentwork ' W increasedfi'o 190 to l991,d erm,£mmmmn'f933?3§'d enmtllgt lncontrasli‘, self- unploymentinagriculturehasdeclinedfi'om1970tol990,increasedalittlebitinl991,anddecreasedagain l992andl993,theninereasedagainl994. "Vi . Q has marked A“)?! negnnm ,Ailm beam-3mm 131-; It biannu aertmtoa 5mm term at aaivil 960d) ‘(llniwqae .eblud-;--.w-il but; curltmut ymod gmxjmioni no "101“ m with“ memolqm'eiloz m onuragtw ed am .I' xewq user bur. m 330d: be. 1.. abtodawrxtmuihmul ted“) commute: lui'umm‘l r: m: L “”71"“ has {d m 01 m WW-mmmuoz s D'Il'a'flt YO‘bfl! hm mmuamlqme-llee ni 9333mm ”owned‘lo ”Weslemmm m {limo-me:- .ca‘iauivflfl‘fiol rigid in manila edi'b MW 704cm how :mts'mb 9dr norm-re {bola M W” redeem at" no amt . rltm drlodemuod bus will“ .-. 1 MW hm (limit as In: trrmtnmwl re bnr chlorine-l ' hill-rather: need and "‘(gataur'to manna at” '10? misbeb Win has anorectic: ‘aquora has man-u .‘ "" “ " ‘ * [unionism most Wand or warren-do Wwdqaaquwmgolaueedi'xmbhn «Mariam mefibmmm “’M and We: anti ..'-ma (mm Ill . «‘5' N as» M00 tr!” W Ci “I“ We!“ ai museum: rind! WWummmummaum 'wfii'f—‘rr. ". f: '54er- ’7"- ’1 Home“, , , "‘ - ‘m’ » > 7 .31 sauna,» r/ , ,-.,',»_:‘ .-j' -. , I ,. ‘ w'_',:r;':' “I H . "13$ ‘ , . q u .' -I" - L." . 1989). In general, the concept of “strategy” refers to alternative ways that families use to lessen the gap between their needs and available resources. As useful as it may sound, the concept of “strategy” presents not only conceptual difficulties but also methodological challenges, mainly, those associated with the unit and level of analyses (Moen and Wethington 1992). The next section presents the conceptualization and methodological difficulties associated with the concept of family/household “strategy.” W “ ” The main difficulty in using the concept 0 “strategy” is the complexity and diversity of contexts in which it has been applied (see Crow 1989; Clay and Schwarzweller, 1991; Moen and Wethington 1992; Tilly and Scott (1978); Tilly (1979); Hareven (1982, 1991); and Saraceno 1989). Most studies of family and household strategies have focused on “survival” strategies. The term survival is deliberately applied to the segments of population that are in material deprivation and economic uncertainty, and reflects behavioral responses that allow them to deal with material shortfalls and uncertainties of their situation. Survival is a process of constantly struggling to acquire resources and include formal and informal work, welfare, and kin (Dill and Williams 1992). Harvey (1989) distinguishes between those who are poor, a situation which is temporary, and those who live in poverty. Those who are poor lack resources while those who live in poverty learn how to live with variable social and economic environments — the lower class (Harvey 1989). According to Harvey, the lower-class environment is constantly threatening with its permanent crises and future uncertainty. V l otbeu aetlrrnrii tram new 'r'm'vnurlc r-.' ”a -' ' ;4 v: v— -. I 3:. WM) mi! $513093 (II .(M ,bnuoa yarn in as lu‘ioeu 2A acumen: Jimm- .-.., l. ' um. mm nae-mad qag «lined oats Jud aeillnei'ttib rm rqz'xv- u H'ru rm 2:1 x-rrq " {gamta'"io mourned W30 level has rinrr ed? {in '4 b 1' m. . um it ..'“r'rzm ,wg'r' ilhrla lsargoiobd" ”W03 Sill 810mm am" up. 3., 6," vii." ..‘3'Vl nnrgnirlmw has new) ".WWM quuna Ail rill n ligrm‘lrfifl. nortlu'ui‘nb irroigolobodu ‘ awarttttiuum ZRPJEUL. .m' 51*]. Mid “mmt‘io "some art! gniar m (16mm mam at” . ”WM anthem need and ti fluid av m mumm'io mm ”W \{IfiT $001 WWW/J but. "GUM if“; r ,nilamm ’_ “90mm 19891 momma bar: .nwu .tavcmsvmrt w W “Iguana“ no beetroot e and 2513:?!“ ”m ai cram mursluqoq it, ememgoa ad! or will“ til! W lrzromrded Aeolian hm Mm lam anaemic nerh'm militiamen I "Mills lama] autism bus neuron! oolliagnimleflQl ‘. Willi“ (€te ‘ “I .woqrrievrlortweaodrbm .nmaqm WWmuaumm.mamummm ‘ I‘MQM mimrwmu ~mwootivne Having neither the material resources nor the political power to transform their environment, the poor must devise ways to ride out or otherwise neutralize its unpredictable nature (Harvey 1989). Families in poverty or in which the husband is unemployed or underemployed, for example, in response to the lack of income opportunities, can (1) search for better income opportunities or migrate in other regions; (2) involve other family members in the paid labor force, particularly the spouse and adult children; (3) seek for other sources of income, both formal and informal, including public assistance benefits and welfare receipts, (4) seek for social support fi'om kin relatives and friends; (5) exist in poverty and survive at a bare subsistence level, with inadequate food, clothing or shelter, credit arrangements etc; (6) use traditional "knowledge" or develop new ones to satisfy their needs. These alternatives are not exclusive. Those who can adapt learn how to live with limited cash flow or be able to produce something that can generate income mostly in the informal economy, self- provisioning, or homework (Campbell et al. 1993). Other studies have focused on “adaptive” strategies. Adaptation refers to the coping behavior that individuals and families use to resolve a crisis in order to achieve a new bflance in life. The processes under the social construction of adaptation include adjustment, conformity, assimilation, compliance, and sometimes a passive acceptance of change. Clay and Schwarzweller (1991) indicate that when a change enables the household to become more competitive and effective in dealing with the extemal environment and without upsetting existing norms for internal cooperation, that the change possesses an adaptive quality called a "household strategy" and also that some usr‘ll :ruolarunl or rmlx; immloq am 1'»; ”arm-«m 'ul mum er?! mitten“ ' . all asilnb‘uen um. P-filv. 'II;.:I(v;;bI1 v.‘ - 2'. ‘I [In/11'- ).«le tooq :Klr MW (I brmdzurl orb dairlw m to {Home m 2 ,1: my. .Iwfil {3‘er tl » muran sidearm emonno be! ed) 0! saunas; .u ulqm 3;.) ml ,iawnlr rumour; If; 1:510qu :anoinerwon‘r armimro aeilmutmqqn urn-om mud n? urns»: '. I ) {er ,asniwrm WW” (“mm .‘L‘Kil wdrri iru; cult ’tl rwdm'wi (hrmfi radio avlovni (9 mm br- lsrmol Iliad .umuum'm (sunny wrin- to) Jazz: I C) Jimmie filth It“ MW [since 10”! ale-:2 (M .zrqreosr m’tiw bns grim-red anagrams aim WWW. leavivrua bun (new-q m mm (? . Amen? has m We) ..an mm libero mum u frntrlmla w.) synapse-d w m from (mm 0! mo wen quiewb 10 "WW “View“ 01 Wed maul all). (In) ortw seat” .eviwluo aim Mr M new mil gnirhsmoa m (“l .b to llodqrru')’ how-strait! in ,gninoia'rvom ‘ -*W“ Nbouroo'i wad whim rod’O . Mmmrmr ml) travelled 3mm 'MM «IT .33“ m sorrel-d well Q strategies may generate negative effects, putting the household in a disadvantaged position (Clay and Schwarzweller 1991). The concept of “strategy” has also been used in studies focusing on shifts in marital and fertility patterns as a family-level response to broader social and economic exigencies (Easterlin 1980), migration (Kertzer & Hogan 1989; Hareven 1982), intergenerational transfers (Hill 1970), various forms of human, social, and economic capital in the kinship network (Coleman 1988), and the impact of cultrual aspects related to care, affection, and role behaviors on the choice of family adaptive strategy (Hareven 1991; Stack 1974). All of the above examples illustrate a variety of contexts in which the concept of “strategy” has been applied. The question remains what should be considered a strategy, what is not a strategy? Any behavior or actions manifested by household members, or anything less than a dissolun'on of a household can be taken as a strategy. That leaves no room for "non- strategic" change or for individual behavior that is not germane to household maintenance and development (Clay and Schwarzweller l991:7). Individual members may engage in behaviors that are passive, non-strategic, overtly resistant, antagonist, and ambivalent, or behaviors reflected in laziness, greed, selfishness, revenge, or egocentrism, or in "everyday acts of resistance" such as income retention or non compliance ( Wolf 1992: 17). In summary, the underlying assumption in the use of the term “strategy,” whether for survival or adaptation reasons, is that individuals within families and households make choices and exercise priorities when responding to external or internal needs. The 10 begmnsvbmib a m blodaauod em gnltluq ”steam 1"“683” wanna-J am new. I 1 .'-)-"i ulicmrsndua has 1819} Ironing . é [Ii with no gai'auoo'l aeibtm m oeau near! our,- a’ui 73318112." 'to fire-moo MT who! lanaue tab-Old o: aznoqezn lanai-vim“ s w mmmr; {1: hm? hm Imu- .(Mi ”NIH-1989! nagoH shunned) noimmmv “'le (“travail aim WWM’M ztmo't anoruw 1wa Mi i) ne'iamm lsnmmmegrai WWMW all) hm: .1889] nameio')» human qulenul all! m Iniqu WWW (Ems) 'toaoiod'a am no mousdad elm has .1101th .555 at _.‘ _ “ ' aniasiqmaxsevorlu :‘ntztollA 1h?! 135:8 I”! ' 5;], We!” hack”: and 2m ganglia m tqeonoo ed! ,5' , . Warm 3915an (391311231an d mm mama to rounder! {rm 1" will”!!! ad on “01‘9th {to noimlneaib. , W830i Him-dad imbivtbm wt to egrmdo' mascara A . MM“ {ID} Wave!) has moan-nun aria-q m tut: minded to 33:31:: am ”M m’ banana: minded to ,rnalavidme iomevahvmomnomm l ..J“ », V ,’ I. _.‘ ' w .fitzsm1loW)milqrmo - concept of strategy emphasizes different ways individual and family’s resources are organized to accomplish desired end states. This study intends to determine how family and household resources (economic, social, and cultural resources) influence different work strategies. In the next section, the concept of work is discussed. W In many studies, the concept of work has been reduced to official employment. Household and family work strategies from the point of view of work require a clarification of what is work, who is involved in work at the household level, and for what needs. The concept of work requires a broader definition beyond the remunerated or wage labor (Mingione 1991; Collins 1990). Work should include all types of formal employment, but also a variety of irregular, temporary or occasional activities undertaken to raise cash and various activities that produce use values, goods and services for direct consumption either by an individual and his/her household or by other individuals and households, which are more or less necessary for the survival of individuals as distributed in different household structures (Mingione 1991: 74). Given the limitations of the data, this study focuses on paid employment of family and household members other than the head of the household, the spouse or partner (adult children, relatives, and non relatives), self-employment, and unpaid work. WW5! The concepts of family and household have different meanings depending on theoretical perspectives and analytical strategies (Bokemeier 1997). Traditionally, family is defined as a domestic group of two or more people related by birth, adoption or 11 are mailman 9'vlimn’t baa ltmbl rllh'il w. a ma mm, Pssparlqm-J v1; “"230 M' vlima'l wort animate!) m abnmm {burg mtl (9:an Emu bum-so dailqmonoa or be!” Wit) mwflnumuoam lmmlm um: lmw .'..tr.omml am 10099“! htorlamodh- banned!) ai alum tn tquonm art! mince I/ an am a] asap"!!! M WWMUI “ WNW 01W” "sod and bow to tqeouou 'fl‘l) mural! ‘mam nl isfimamtommo mm m mor'l migatmre zhow (limit baa mods-nu WW 3‘ in barn at bsvlovm 6:: 0d” .110” at mlw to from Wtflndsmmpmbowlownadl mm MW (0?“ mdlo'.) lQQl onmgn. Mlmdnl can. file-mam .rdrgtmi‘m {nhsv it acts md .memvolqu ”mummy-awn" mamas-m ‘ WWWMI m vd will! 9 1108qu Wm»: 10 mom or» dantw groom ‘ V ..‘. W'WHM tnm'l‘lib né maximum Mdhwmtumm ,raa. .« r .mhmfla investor-ma marriage (US. Bureau of the Census 1990). Families live together, share economic resources, act as cooperative and caring social units, and provide environments for the emotional, social, and economic well-being of family members. Family members live in a distinctive domicile called home, and the household is defined as a unit of co— residence. While the above definitions of family and household include a variety of family forms and structures, its usefulness in analyzing social situations, especially those associated with the diversified ways individual family members respond to changes in their surrounding environments, depends on theoretical and conceptual orientation (Bokemeier 1997). The neoclassical approach assumes that the household is an income-pooling unit with shared common interests that make rational choices to maximize the utility of the household as a whole (Becker 1981). The underlying assumption is that of consensus and cooperation within the family or household Also reflected is the assumption that members of family and households work together toward a collective goal (Tilly 1979, 1987). However, co-residence does not always lead to income-pooling and mutual obligations for the purpose of survival, nor is it strictly necessary to be co-resident in order to be involved in household strategies (Mingione 1991). People may live in different households but share mutual obligations. Separated and divomd parents often continue relationships with their children. Seasonal and temporary migrant workers are not residents all the time, but are fully part of the household reproductive strategy. Additionally, the physical co-residence does not tell anything about the kinds of resourcesinvolvedandtowhatextenttheyarepooledinordertoachievethegoalof l2 OIRIOOO'J‘J 3w!“ tuning-tr s-nl mlu vrfl NINJA ”in-,1: q! t It) rvnmuil 2mm 318161 ammoni‘ are ubnmr; DU!» .amu Ens”. 54:. .r. -L~r..-;x:1..-rarroo.- 7.8138 m at evil madman: '(lim‘l nodular" {ham to gnu/Lila u .unmtna hm; .lcr'xv. ,lanoitu. 0010f!!!“ I 33MB!!!) 7.: blori' tort or ' w; , «(l he’ - slurmob W030“. '10 W I abulani blorloznorl 1m vlrmo'r to anouim‘bb evoda am alidw .m mm W ui ”army u: r ‘mou'rta blur armo'l vi“ WWW!!! (hard? Mmbm .- raw lartmavvb an; mm but!” 1 mm has new no zbnaqab ,memnm' mu anibmmra 1rd .(VQQI veronica) at tut: ma damnqqs lsalaulaoon arfl‘ .mmohntamumwoimmnmbsmtzfl'n a“ 4189i ”Mralormaublm w, ngrmwmmmrmaqmm swim but (lime) to m i ' www.mwofl .(WQI ‘WMhmammtmm Matbvvlovoradotm survival. Finally, households may or may not contain families (related members by birth, adoption or marriage) but pool their incomes. This neoclassical approach is criticized for ignoring the processes that structure family arrangements and the social contexts that embrace it. The household is embedded in various social contexts including community, kinship, friendship, and other networks contributing in difi‘erent ways to the patterns of social reproduction. Furthermore, this definition is criticized for ignoring the competing interests, internal conflicts, and struggles for power that characterize family dynamics (Wolf l992,l991,1990). Mingione conceptualizes family and household as overlapping, reciprocal, and interdependent social units. In his framework of household strategy, Mingione (1991) emphasizes the considerable importance to processes of social reproduction. He defines the concept of social reproduction as diverse conditions and organizational relations which allow human beings to survive in various social contexts and groups (Mingione 1991). He considers the household as unit of analysis, but departs from the others by emphasizing the social relations between members of the household: A household cannot be simply viewed as a statistical or physical unit of co- residence, but must be seen as a set of changing social relations which establish a set of mutual obligation (basically reciprocal forms of social organization) aimed at helping its members to survive (p. 132). —The most important social network in which the household is strictly embedded is the family/kinship system (p.133). Feminist family scholars define the family as a location in which production, accumulation, redistribution, and consumption of needed resources is carried out. 13 .-lnid '{deradmem f’filul'l" . .: ..rut (lunar/r 1m 1:..r m lam xlrlu'la writ-l idlitfli‘l . ,3. unrt' ‘r' h3:...fi;.12(-:3rat"tam10M' C momma Nd! 2922qu ‘Jlll gummy "if m .ullrv . , fl'fl'fllflf“; lam-ml nan :mfl‘ - u bobhedmai blorlaeuod or” .n awrrEv-n ml: at; w m M: m ..1 antagonist“ zmwo baa ,qrrlebamfl .qrriannl .{Ia-‘un‘nmou gamut- .2: aromas lance wound .MW newborns: laurel“ mm .11 or. .v: rm tab at 3am “Whirl!!!" atmtum .gnusqmua at: gummy rot beam. 1') 2: float!“ MJQRI£9W 'lloW) mimrnvb (lion: firm-writ) tall! uwuq wt to” M mm» blorlazuon [m virtual vs lsurqsumu anoigniM ' "I. I... V ' l ' a . . “Whale W or ammuqmr elder-alumna at: m , WMumowhnqm mee'to toast-road: ”new at sv'mnc or agn'ratl trammi walls Mu According to Hartmann (1981, 1987), the family is a location in which production and redistribution of resources are carried out — a location were people with different activities and interests in these processes ofien come into conflict with one another. No longer is it possible to conflate family or household and assume monolithic interests or access to resources of household members (Wolf 1991; Tickamyer et al. 1993; Feldman and Welch 1995). Instead, studies focusing on the household nwd to analyze household dynamics and social relations. Such an approach would include analyses of social relations within the household, especially, the relations of gender, generation, and relationships of power and authority, that are crucial for its survival and the ways in which household members interpret and respond to social and economic changes in their environments. For a more extensive review of conceptual definitions of families and households, see Bokemeier (1997). The next section discusses methodological challenges associated with the study of family and household work strategies. WW Conventional research on families and households treat them in isolation from other social institutions, such as politics or the economy. Ferree (1990) criticized the conventional conceptualization of “separate spheres”, where the family is considered as “haven in a heartless world” (Lasch 1977), where fundamentally different social relations prevail than in the rest ofsociety, and, where women are given a distinct and secondary status. She indicates that family isolation is illusory given the close connection between families’ internal lives and the organization of the economy, the state, and other institutions (Ferree 1990). 14 ~r - N..‘I‘. ‘l:§- ' i L. I. E '5 F I; i . .- .--.“yw.., , '-‘.V ..'A ' I A A‘.I' ‘ , . .2 l‘ _ ' _‘_v_.‘ .. M ..r brrarmnoubmq dnrrlwm Ir-Jlls'm . ,. ”Ia-m _. J - < Umsmrmflm inflo'fiib rhiw slqocq 719:. nuzagmi. u! 3.: ‘ m 91: anum’cm'loMI ' OW .mbormano 111M Iolf'irm mm u' '0'» mun » '1.‘ V” ”04' m alurgmrbalm 10W aidriloaom ommzw ma Hung/J ml tall)! mums 01314qu n ““7 W I“! .llbflmmbil WW 310W 1 <‘u‘:""1';rrrbluisu'uri to zurruoealotm “WNW! blrxlueuod 5A7 no gummy! .3 alums no: man] ROW MW” “hmobulom bluow (bumpy. in. my . mud-n lam». has ”I!” WMJM'M alarmist orlr ,‘Illrz. Hr“: INN" :Pm‘fl ad: HM!" w wwma an 70'! 13mm are red: , unodiuu bur. ewoq lo cam WWW or bnoqmr hm mam: mdmom biodaeuod m ,1_ r. Wlom. “mi: .- u/uu. 1,”;le mm Wmuan 3dr (rm. .mwhaliufl m was“ ‘ . W . . but (Hamlin Ma arlt dim batsman; 8W ' "re-2; ., » Wm ' hm m no dam lmorrnavnot) L.” a! ”whiten as (hue ,enoiiunum laiaoe 3‘. 5H m" bituminous-awn imam '0 The separation between the “public” and “private” spheres, between production and social reproduction, between paid work and non-pad work is purely deterministic and an oversimplified view of social reality. Family and household members construct their lives within and through the constraints and abilities of their family social structures as well as that of other social structures in which the family/household is embedded. Families and households are not only connected to, and influenced by, other social contexts, they also include these structures (Baca Zinn and Eitzen 1996). Curtis (1986) indicates that despite the increasing attention to structural components of the economy, little attention has been paid to the relationship between the larger economy and the structure of the household or family unit. An understanding of the complex interrelations between changing patterns of social organimtion derived from the sphere of reproduction a well as production is very important to the study of household strategies. This study considers work strategies as embedded in the social relations within families and households which are also embedded and interconnected to external structures, including communities, kinship networks, firms and industries within labor market areas. Moen and Wethington (1992) indicate that the repertoire of strategies of individual families both shapes and is modified by institutional, cultural, environmental, and interpersonal circumstances. Furthermore, communities and social agencies, firms and industries, and states have strategies, as do families and individuals. Changes in the organization of the economy, demographic changes, and state reforms enable and constrain the ways that families survive and adapt to such changes. At the same time, individual members of households and families are actively involved in various 15 3 . n' ”K r.— M Y. ‘ noilauborq seemed ,eswrlvy. 'ummq ' hm. " n 'm ‘s’l! ~~~~~nnd normal-tread? Oixclairmetsb '{lmuq ei alum Lyra—non hm. ,m ./ : '1 3175.190 erbubmqfl IM' Meantime: hisxhxuod bmsrhme-I u. .‘ - . .v z» 121/ ooflilqmmm-h We lain! {limit riam'tn asuilrda hm. --r .- ..'.-:n~:u', -m {lg-Mm“ hna nulrm 39:“ fl. Wei blodomod‘v'tlumm rim Mun: m ' mm mm: maro'to red: as I”. ‘ {mm-did Willi! bus ,0: huxjasnov flux. :04: ..vu' .zt-l-idaauod balm (NW 40'“ mi?! baa nan and I (can MUG new!) shalom eels wad: M J human: or ”061mm: grimwmu yd: errqmb tart! ear-a“ , fl Ila-tea named quremmam :6: m burr mad M "0m“ ~- ’ r .. I , “Wu 0A .tinu {lime} in litedtmrod am to W ' W Mimic lav w, to emunaq grrigrrarla about" to thin all) or manoqmr vyr! 2i rroimubotq I: II” I arr-W )how ambient» (buzz. mtT ,MMN‘I ans dordw ablorlsauod bus and“ Marian! We gnibulom 2mm (Mltwitbew bus neoM um um an.“ that! nilrma't laubivillti activities, mainly in paid labor force, in order to achieve their goals (Garkovich and Bokemeier 1988). Thus, a multilevel framework that encompasses the interplay of macro-and micro-levels of analysis factors is needed to understand and illustrate the complexity of the construct of the household and family work strategies. The next section reviews factors associated with family and household work strategies. Across levels of analyses, a variety of factors influence family and household work strategies. The economic opportunities and their spatial distributions, as well as social and demographic composition of different labor market areas, constrain and enable family and household work strategies. At the same time, family and household social structures, resources (economic, social, and cultural) and other family and household characteristics such as geographic mobility and life course stages may influence family and household work strategies. Other family and household characteristics that may determine what work strategy will be adopted include age and educational levels of household members. Moreover, the prevalence and diversity of household work strategies will differ in metropolitan and non-metropolitan areas. Non-metropolitan families and households are more likely to be in economic distress and poverty than their metropolitan counterparts (Rural Sociological Society Task Force on Persistent Poverty 1993, Brown and Deavers 1987). Non-metropolitan labor market areas have a relatively limited employment and earnings opportunities and less diversified labor markets (Tickamyer and Duncan 1990, Killian and Hady 1988) compared to metropolitan labor market areas. Thus, non- l6 baa dynamism dang mm svarrlm ._: rat-rm m I; m): win! mm m {Inism 23m ioWiadtmzaqmoommm haw-mu! 1; '4! ’iurr' r. mi; 13801 niomsI-I' ammh baa baneuabrw or babe-m ,; (rum? err/ism.- 102$; ml-ouim him mm .W thaw films} but: blwiaamul am in ramming 911310 {film .mflm blorloauorl bars giants: rum batsman-an rumor awoivm Mina WWMMWU? anmamm W Wm mm: to (tens I a mun-mam 'iu Pia/3“ wow/8 “Miriam 1531i: boo mnmmoqqu 'v'rrrrmmfr id 1‘ 29313308113 m m WWW umoflrb’lo momma”) arrtqugomab has I“ hm .laiam .armomrm momma. maxim Ilia . ' “if” mm okrlqmgoog is done nnzriunmarlo blmlsauul , , mmmmww numb!) mr mrt'uuaiietmarb ”:5":""” "’ ' '7’ ' . ' Mmemblorteaumtiolzwllanonm 10M has mama ssh 13va may: scrim-non baa ashram baa commence at ad m mm em. as warm m laaigoloiaoa mm W nadeum-mw rm: an mm unimnqqo mil-n WWW com (380! W but-i“ e3! v.5“: metropolitan labor market areas offer fewer alternatives for potential family and household work strategies than metropolitan labor market areas. Family and household work strategies are likely to vary by gender, race, and social class. Feminists have criticized the limitations of research that aggregates gender interests and unequal power relations within the household (Wolf 1992; Hartrnann 1981). At the same time, they emphasized that families and households are fully integrated into wider systems of economic and political power (Ferree 1990). Empirical evidences indicate that, despite the increase in female labor force participation over the last decades, and the associated dual-earner marriages, women's economic opportunities remain conditioned and shaped by their disadvantage in the wage labor market, their oppression and domestic obligations, their high participation in informal and unpaid labor, and by state policies toward women, work and welfare (Tickamyer et al. 1993). Furthermore, families and households are changing at difi‘erent intensities, rates, places, and at different time as indicated by the increase in female-headed and non- family households (Skolnick 1995; Santi 1988). Research has tended to either focus on married-couple households or female-headed families. For example, recent research on poverty indicates that female-headed households, particularly those in rural communities, have been increasing and are more likely than married-couple households to be in poverty (McLaughlin and Sachs 1988; Tickamyer and Latimer 1993; Bane and Ellwood 1986; Hoppe 1993; Lichter and Eggebeen 1992; Wilson 1987; Duncan and Tickamyer 1988; Fitchen 1981, 1991). bna rum? hlmmuq a»: 29w 'rx. r'n' Full“ 11.1"" with“! mdal M 49.21:: 22- “us: win! mgrirqonam narii «again-tn in»: w . htupam Jabnog (d vuw m {la/i“ .. n"l::1 arr» J -( l hinnzwuri has '(llm‘l q. mm m acronym In / mutant"? 1 ' ’ ‘ «min and ”dimmed all.“ .( "Q‘ W .5991 WOW) Mariam»! M! nir‘w: menial"! mmrq lnupanu In.“ wafltfl mabmdazmm blur ‘illfirtil lui‘! !)-.r.'..u..'~'grrr, rad! ,amu mu“ WW3 10901 eons-l) ra'mm l-miiluq .‘mr urmvmooa'to mfi in film “MW 39rd! rods! flaw-.5? m 92mm” 3dr :uqzob Jflt m WWZ'W «teammate mnxmoieutx binlbrinmcl art: ba- ,8“ mm will manhunt!) 1m!) vii Lawn? Dfia benortrbtloo If“ i ’ l m rigid us!" ,enomagllrtu 'umrrrub has new Wham W brawn: mailer: am». '{d has an” “.mod bars winner mormadmfl muwm an aunt mound; urban M m .{W “a .1091 Moidmblodowod (lid MW 10 duodenum armchair“. W huh mm m ‘amubuanrmnnrrmdm www.mrmhmmmm WWMW»W!W WI J ,. , -f, mmumm a . _,y«‘ ,'.~,, , 1'! L! . However, little research on work strategies of male-headed families and that of the expanding "non family" households is done. A demographic profile of the diversity of family and household structure indicates that the proportion of male-headed families and that of the non-family household has increased Increases in single-parent households and non-family households are mainly due to the changing patterns of marriage, divorce, fertility, and increase in cohabitation of unmarried couples (Skolnick 1995, Santi 1988; Goldscheider and Waite 1991; Stacey 1991). Furthermore, analyses of family and household work strategies must acknowledge that families and households are hierarchically ranked by their social class. The economic situation of households limits choices and favors certain patterns (Crow 1989; Saraceno 1989; Moen and Wethington 1992). Households have different capacities to engage in various household strategies, and these differences reflect both household composition and economic resources. Households with more resources tend to have more choices and more information, and may be able to tolerate higher risks associated with certain strategic misions (Wolf 1992). Harvey (1989) indicates that those who live in poverty or near poverty may have neither materials nor political power to transform their niche in order to rise out of poverty or otherwise stabilize and neutralize its unpredictable nature. However, they may find ways of getting by in absence of social supports for family living. Thus, it is important to analyze how different household social structures as well as that of other environments surrounding the family/household, in particular recent widespread transformations of industrial organization and labor as 18 ‘90 mil bnr. asthma"! babaedolsm in vea'fitf'lw r2111" n1- :1 m. ~11 am? 73me " (mint) adt'to otitmq airtqmgumub in“ arm. .1 m: :4lu.-.'<-rl 'vltrTrM mm" umtmaqm‘ MMWW )0 noano‘vq 9H! trm’: warrgxf'uu amnrnn U1 11in Punfl baa Vilma“ Ward mam-ml 11'qu -r1:1 1» L004 limalwrvr' urir'tn ram h‘ ummw 0190b {imam '51:. :blhz‘u'l'u" .i'nngt non brie lbw mm Mic mirmrluic; w Jd... an: bar. ‘ ' i first ..':movib .335!” . . “091 Y‘Jflfile .101?! um H bnaaebrxioabrm) :88‘4 imaz .BQI mm.“ how blurbmwl baa Ilmm’l ’io eozvlrm: enrimnrlrlr‘l w M!!! bunt {lwoulzmttml m. ahloduaumi baa abilimu mm has noted: Mr'ml abimbeuorl‘tn mitwtie canon. mum river nolguuitew .‘x. a uvoM (139mm WWW has .zaigaim blorleeuort wo'nav ni a. J "MI: Marinara“ tamer Mimosa has «room I“! Unit at! vam has ,rwrtarmo'tn'r morn bars mode as. Wimu (:00: HOW) mosh 953315112 arr-ma rhi- mm “*3" watt lam (“5‘qu urea 10 Mi ,gfitwififlflm to wit. no van or who n'r :dotrt Tu Wth ball on trail! JemoH .a'unan airflow Wcmun am” gai‘nlvlrms'hrfi m Wm» m u rm as m that 81 well as social and demographic composition of difl‘erent labor market areas, have enabled and constrained families and households to adopt different work strategies. W In summary, this study attempts to link multiple perspectives to multilevel models of family and household work strategies. Family and household work strategy is a multidimensional concept that requires multiple approaches and multilevel model conceptualization. A general assumption is that economic, social as well as cultural resources controlled by members of families and households, as well as their social relationships, interactions, and activities are crucial factors for comprehending the diverse household social contexts that permit members of households to adopt any work strategy. At the same time, the way the household is situawd in the broader social, cultmal and economic environment (the economy, the state, and other institutions) affects the "strategies" its members can do and adopt Of particular importance is the spatial environment in which families and households can have access to employment and other opportunities. An integrated theoretical framework of family and household work strategies is presented in the next chapter. Chapter 2 presents a theoretical framework of family and household strategies. Two theoretical perspectives relevant to the study of family and household work strategies are proposed The first part of chapter 2 describes restructuring perspective. The second part of this chapter describes social embeddedness/capital perspectives. These two theoretical perspectives are integrated in an attempt to account for both macro l9 marl .amr. razlmm awful m'..r-xi....- ' ' mm" 1 1M '1. i-imqomsb burs him.“ 3913338113 how rtmal'ub richer ! (,9..1.1',/.uuri mu: asthma] bermmanuobnlw' .3.- r. 11.411.152.141 - -"*J,,‘L’.L1“.QULLW gammyfi . b leveliflam 01 maximal .alqmuzs: Jml n: eiqm'ma {tune am {13mm a1 ‘4 .. 3i W 1101' Malignant! brie '.'_llfl|li'i anaemia. A1111; Lind-19w 141 bus vlrml'ibm ”WNW udomnqqu u‘qnlmr nmur ' :v 7?. «mam lantvlarramlh'm Mallow as brow. .‘JlmmIO‘JJ 1M: 4 nm'm-tvw “$13138 /\ nmllfiim . mm“ lbw u .abio'tlewod bur; urimm'r !-1 nerin'nm (d balllmooo mm W film incur: an mums hm; .mu'mmv and?“ ‘ I. l H _ W10 endures" "(man 10111151191004 ‘maoa Narmada“ am a New orit yaw 3‘! .arm! ems: ah m .w ' ”It «mom 3‘!) mime grammes) hm m w *hd an autumn 1.2! Warp» '13?" ad! an. ' m=w ha autumn death» an .vrmnmnivn: in” w W rm .zartrmmo'nvgo tad» b .Wanmdl oi batman Ai Magnum m m 3mm S rewind? fit ‘- runk! W 1:90st ou’l' and micro factors of family and household work strategies. A series of propositions and hypotheses are summarized at the end of chapter 2. Chapter 3 describes the research design and analytical strategies of the study. This study uses existing data from the US. Census. The chapter first describes the data and variables used in data analysis. Finally, the chapter describes the analytical strategy of the study. A multi-level modeling approach is adopted in an attempt to account for difi‘erent level factors of family and household work strategies. Chapter 4 displays the results of the study. The first section of chapter 4 displays the descriptive statistics of family and household work strategies, the characteristics of families and households, and the characteristics of labor market areas in which those families and households reside and work. The second part of chapter 4 displays bivariate analysis of family and household characteristics with work strategies as well as that of labor market area characteristics with work strategies. Chapter 4 ends with two multilevel discrete models of self-employment and additional earners on family/household and labor market area characteristics. Finally, Chapter 5 discusses the results and their implications. This chapter summarizes the results in relation to corresponding hypotheses. Policy recommendations as well as the study limitations and future research are also discussed in this chapter. 20 bns ano'nizoqmq lo when A ”mm:- rho” blur‘usnfltl ‘ 1.. ,h: nil '10 atoms} , ~ C "turqurlo'tu 1m;- ‘3"? m iesrrsmmua at. w {blue edl'io asrgstam lawvvlunn brie navy!) 113mm .1 '1!“ rain nab £131an . stab at“ aadi‘neob mfi marl: arr! puma") 1": U ad) mm". rrnzb garterxo tea» may WWdfiwdmaab )qul') 9:11 ,vjlmn-l aw'llmrr. slab m beau ”M“ 101m 03 tqmsm an in Domains er nuarnqqr: gmlubom level mum A .M *1 .aeigsrena 110w bimIant bna virtue) '10 notorit level II“ WW’M 00m MIR 011T .{bme an: 111 "i’uflfl art: usher!) b 1:!qu \ humanity-gum m mm m {man in mums): ”mm-aha. .:- " ' mauiomarnmma ad) has mama muesli-i , ., '. " . “ ._l _ We!” Jhowbmabresublorimodhmm ' mm Wain block-anon bar: {lirruit‘ro an“ .m thaw thiw nim'nazornarto arm: trim“ {Ware-11:2 To ziabrxm stewed. 1min!” *. . ~ - Wmmmdalmm blorioeuortwlin‘ '9 ,Wm «a man a manic .yumn. ' ‘- .»mmmflaiarhnmamm mama with art: a than Chapter 2 Multiple theoretical perspectives relevant to the study of household work strategies have been developed and applied. These perspectives include micro-oriented theories such as status attainment and human capital theory, and macro-oriented theories such as structural theories that focus on social and economic organization and forces of production. Most research on household strategies has focused on structural factors that constrain and enable potential household strategies (Crow 1989) and fail to include individual factors. Other studies have relied on purely individualistic approaches (Becker 1975, 1981) and fail, not only to take into consideration household and family dynamics, but also the ways in which they are embedded in and connected with other social structures. What has often been missing is the connection between the macro and micro levels of analysis. An integrated theoretical and multilevel conceptual framework that bridges the gap between these theories and analytical strategies provides a better understanding of both macro-, mid-, and micro-level factors of household strategies (Tickamyer and Bokemeier 1993). In this chapter, two theoretical perspectives relevant to the study of family and household strategies are presented These perspectives include (1) restructuring perspective, and (2) the social embeddedness/social capital perspectives. First, I argue that family and household work strategies are embedded in the social structures of families and households. Family and household structures, available resources, including 21 2 ”mail , Wflih’tv 9.1.1:- airman 9W ‘ V“ m blorlawort'lo (but: an: 1? Mr Italy: 2'. . duo-1027 la‘ul‘florzvll :rlqitluM '- Wo-orarm ebulam mbesqzwq near! !' :mjnqr. but; mom-web mud and a“ WWW hm: ,‘noarlt limos; 'ilfifluli om. wvnmmu. «JUN? 28 than a?“ ‘io WWW almonoos um lam... nn 4; .5. ram armed) mum ufi‘ Wheat)!“ and urgently blrulwmrl t'u ’b‘lfrm‘fl 520M hum Wuhan: m» migelm blorlaauml arm-mm 3mm m nan-a : ”.7 . ' ._ ‘ View no boil-n OVIM whim 'arllu .momc‘t W mi enter or vln‘c um ,m’l baa t MM 370! M anathema 98 [all riotrlw m new ad) oala md mom an at War need who and MW «.amnmta la'uu I WHW Wt nA .aizvlms ’to zlaval m H hmmmqqmmbmfi hauntxmmodiomihum ,(mrwmmwun ”Waning-lo a'rrlml can Maui‘s-mum social, cultural, and economic capital, and other household characteristics, condition and constrain potential household work strategies. Second, the family and work strategies are conditioned by larger social and structural contexts in which families and households are located and interconnecmd At the same time, a spatial distribution of the social and opportunity structure which is the outcome of both historical, social, economic, and political change, enables and constrains the ability of families and households to adapt to their social environments. Each of these two theoretical perspectives provides a valued partial explanation of household work strategies as well as shared assumptions with other perspectives. Thus, propositions are derived from each of these perspectives and their integration, hypothesizing the unique, shared, and combined factors of household work strategies at both family/household and labor market area levels E . E . Restructuring perspective provides a lens to analyze how the ongoing transformations of the economy, industries and occupations, the associated spatial distribution of jobs and businesses, and the shifts in the social and demographic composition of places create social contexts that constrain and enable family and household work strategies. Restructuring refers to macro-economic changes in market demands for labor, capital, or technology, shifts in geographical distribution of people and enterprises, and the process of capital accumulation and competitive transformations in the industrial and occupational composition of labor markets such as the growth of the service sector and the decline in manufacturing industries (Redclifi and Whatrnore 1990; Bokemeier 1997; Tickamyer 1996). Restructuring not only has consequences for the 22 A bnanomnnm nif'cnw'srmlo iim'wmrl min but; imam: "memo: baa Jumbo wigalmrz )hOW 1mm liz‘zmm um em war. .uz-pnma Aw: blmtoauort lsbndqu' ' ablorbauorl bnnaellimril down 1;. (l/‘u'w, kahuna. ..i a mum «natal (d bamhM‘ baa labor: with nonudi'narb lanes". x .‘ mu um»? um .':' ammumtm brta man] has .aimonooo .hiooc ,lnommcrl (hurl i . ‘. muomu an; n 4.9:er utuuune Vim 0"me baa aatllmtil ‘ln (“Win an: (med-my; hm: zal-‘lcna agnarla It!“ Washimmmoqaaq marrow-nix ow: and: to rival! .amamnorrvm mm mm Wall bonds as New er; Pblg‘flfll}: :l in” blottauurl 'lu nnitamiqus I“ “W numbness mail bat/[tub an: enmtwmoiq .aurl'l cum WWW lam; lamb. ,wpmu ads 3fllYlfi‘Jll’flKl'-I_d W abut-m retinal 10an turn Monet-non ,‘Jlll'flb-i Mod to m WWW n W due! a abrvmq . Noumea annotations}! . f *mm Ml (moms: am to mom Watt-tau adr b. mm in moo mm W m luau m eeaelq to nomm “MWammua—a Wm thaw W “W to Jana-o w wit a...» - ”than air but uranium I. 1 spatial division of capital and labor, but is associated with changes in different patterns of work including part-time work, underemployment, contingent work, moonlighting, multiple earners, self-employment, industrial homework, bartering, and illegal jobs etc. To what extent does economic restructuring accormt for self-employment work and multiple earners within families and households? In the last three decades, the US. economy has experienced three interrelated major trends: (1) the increase in new technologies, especially in microelectronics and in other highotech industries; (2) globalization’ — given the increasing competition abroad, and associated corporate strategy of reducing labor costs by closing down firms, laying off some workers, relocating, merging businesses, and/or investing overseas (capital flight); and (3) the transition from extractive and manufacturing industries to service and information sectors‘ of the economy. These interrelated transformations of the economy can be explained in the context of contemporary world economy and its associated competition I argue that these economic transformations have created social contexts that make self—employment and employment of ‘non-core’ family/household members an adaptive strategy for both capitalists and for families and households. 5 Glohfinfionbmfifiofimflywedficpmcasofglobflwommicflfinmddmmagmmgmosewwd byapowerful global elites of financiers, international and nationalbureaucratsand corporate leaders (McMiclnel 1996228). 6Table 653 fi'omtheU. S. statistical abstracts producedbytheU. S. bureeuof labor statistics mdicnesthamepacamgeofworkasanployedmmacdwindusnieshawdedmedfiom11.18pereent in l9?0to9.83 percentin 1990. Thepercentageofworkersernployedinmamrfacnningindustrieshave declinedby 8.43 percent from 26.37 percent in 1970 to 17.93 percent in 1990. In contrast, however, the percentageofworkasanployedintheserviceindustryhasmereasedby7J$percentfrom25.9lpercentin l970to 33.09 percent in 1990. 23 atnanaq lflu‘nn'W' r- "a?" u . (..‘ .. a ' '2. 43110 which“. ,gnmigrlnomn JIM/I map-M's ' .m I. .11. u {..n emu-hm gallium “ .atoadoilagelh bns grin-4m” flu warms am an”; mum x..dr;nveo'tt92 21mm bnalmwflmrrfolqm'rlix wt ".2 W. h ,4 4x. -. -t r ,. .; am (1700 103mm“ . 'rblm‘flwnul m- ; .-:, m. M?! (”de rm M NM 55111] bummed/u mt (atom 1,, - , nit . - 4, ‘J‘JHU teal 5d: I'll - lll’bfll Wimrmm ni {Hanna‘- ,. "mount r... w n r A. nuoam um { l , Amen“ ”Wmlwom am new - - awluvh. iv"! ‘ "5 4‘2""‘viu‘? rim-Will “mm granola Id 81600 10d!“ :3: 2 1:11.17, 17 ‘11) I-cwmr/ ‘llanQ'lOO bulim b ”3w ”Mai to‘tbna ,aez? 9"va gmgnm .gfllllt'Xt‘V‘l mum” mfi “murmur mum“: baa ewmmt;q hunt m'rllL’fiL-ll w!) (E) has .‘(fl Wallflomm Mailman“ earl 5' (mar: 11,: a!!! to "mutual “031M Whvmm thfllm‘flo Mama) all: m ixmmqub 3d no MWWWI: 068100030 will mm 3mm I nohneqrme MW 'm'h mwlqrm baa mammlqm: 53b: 31am 1‘ mum to?! has mmma mod :01 mum». win-ha wwmbmmm manor-suns) . w. mwwwxhhfimoqa!‘ macaw Miami”! Uflmmm z Ushmiilbelda‘fo ' . ‘ .‘~; ~m alt .MWuI 0'!“ fitment!“ more: dmnjhttas-eimtuwm amine-mm“! Nd” .__, ..ar. WWW , ; .Mldmfifiia £9 m_iwmbmmMW" ' ‘ an mumaouMI' ‘ On one hand, restructuring has created new structures of work and employment In the context of increased competition, a changing international division of labor, and continued stagnation and economic crisis in the developed world, national and multinational firms have sought to decentralize production in order to lower labor costs, increase flexibility, and minimize investments risk — subcontracting, franchising, and self-employment are part of this broader trend (Mingione 1991, Dangler 1996). On the other hand, restructuring stresses the constrained choices available to workers in the labor market and at home and, ipso facto, the degradation of social and economic well- being, such as continuing race and gender inequality, increasing poverty, a more polarized class structure, and shrinking employment opportunities for those outside the technocratic elite (Tickamyer 1996214). Self-employment and number of earners within families and households can be viewed as alternative work strategies associated with both aspects of restructuring perspective. WWW One explanation of self-employment includes counter-cyclical response to recession closely related to unemployment (Steinmetz and Wright 1989; Myles and Turegun 1994; Portes and Sassen-Koob 1987; Pahl 1984). The argument is that in periods of economic crisis, with high levels of unemployment and economic hardship, self-employment activities and small business enterprises are likely to increase. Thereafter, a period of economic growth is expected to be accompanied by a slowdown or decline in self-employment Families and households experiencing unemployment and underemployment may get involved in self-employment in order to get by or 24 ,tnumvo'qma hm; now in r .w: .4 u . .;..u 1m grmutumreo'v. ,bnnrl ma bnanodnl'iouoir-mb :nnnumw w 1': .1‘m .. _n 411m my) oagnaruri‘io boa l'rznoitlm .hhm“ Lu: Muni- -,.;; . «x J ,m mm. 9 bar. mm .’ .aarn rodal rawni on "aim; in nomut. r. . .lmms, u. 0' irlgun- vnrl mnfl M)- bnagnieirbrm’t ,gmxammmd .w - '1 rum 2;” .. _. mum Lam Uilrdmflw ammo {$09! reigned ,IWI mogul-IE: him: 7'11: 70' «Mo “um 9131mm 3‘ -d|ni m or sidelines (.30me to: “manna 3d: 23am»: unnutwflm “*1. a" aflmmlflam '10 “U"Kbmg'lb wit moat Gerri ,bflfl omod u: ban aim“ "H‘- awning-am atria-mom ,0:le inbnug 1m. 39m germanium“; " . m ‘Iflaitimmorap Manuelqms gum-nth: bus mutant; mic M I_=' MMMmo-rbe (H «M mmnnr'lnulam ‘ I . I M Wis >3 bmw ad nu? eblorbeuod but so“ .wuoeqrmq anmmmtwr “to steep and ' Wan-Min uonmqua 31:0 or (Meier zlmlo m ' - “.mmerw hm m9 um MT improve a deteriorating financial situation Steinmetz and Wright found that while there is evidence to support the counter-cyclical explanation of self-employment as unemployment rises, this effect declines overtime (1989:997). They also indicate that self-employment was not just a consequence of an absolute lack of employment, but was also a structural response to declining opporttmities for good jobs (p .1008). The counter—argument is that self-employment may also responds positively to a decrease in unemployment Recessions may bring both uptums and dangers for self-employment and businesses. Economic growth may encourage some people to risk starting a small business, while others may take safer option of a better chance for paid-employment (Eardley and Corden 1996227). Also under the context of economic crisis, intense competition and uncertainty, the main explanation of self-employment appears to be the decentralization of economic activity, including decentralization of production, marketing, sales, and business services. Self-employment through franchising or subcontracting appears to be a corporate strategy to reduce costs, increase flexibility, and minimize investment risks in the changing economic environment, characterized by increasing competition and continued decline of the economy in advanced capitalist economies. The other explanation of self-employment associated with restructuring is the shift of manufacturing toward service industries. The number of jobs in manufacturing industries not only has declined in many places, especially in rural areas, but also the new created ones are in the small manufacturing firms which offer no benefits including health insurance, vacations, personal leaves or decent jobs (Lobao and Schulman 1991). 25 and: alrriw tart: boom Man x'l '." ‘ war» In" - -v..~ i1 'Jltflil‘l wumnmflbl asimmvolqnm-khm.» awn-diam . run-um -..'l.‘ nocmuaol \ tadiataoiimiozlsiarfl n‘t'z'w'l'v-..u~ '1'» i .de" -‘ trill .733" «w mu MM :dT .(‘mqudoltqu iii! d‘Jli'dmnz'glh' grandam u! swinger MMO‘. "I i mdenamvolqmoiozl bl mulch." it; :II 3': ‘cgum r a m. ”Wt“ [Mutter] gimme-,1 ”41:: um. wwm. mitt-"i 'l: 'uriJ at mamw 7 ,5‘ Wflfl‘lfl 81:13an bnu 'znrzriu'. n .111 gum) run encircwyfl 1W ' . IMIWfitdalqouq arm» agn'iuirmu mm strung nmmooil awmitudu V j, '. Mflmdo tuned a'to nonqu rain.» mint mm and"; alidw M i f H" (MN :13me has“ g mm mm .eizito aimomnu m iwmm m: robot: 031A H T Wflfim MB Memolqm-Tlcz m aortansqua Iii“ . ‘5‘ '_ mm .noitnubmq to nonexiiumm 'nb anioulani m ’ WEW dammit mavmolqms—'il.~9. ,zaaim ‘ V cm alumni .zlwo snubs: or 39315113 W p .. , _ .WMmrmeoimimM‘ 1". p W Wm a: mom with animal) W V, ".r At the same time, the number of jobs in the service sector has increased However, many of the jobs in the service sector were part-time and offered lower wages. Service industries offered the greatest possibilities of people who wanted to work on their own. Mingione (1991) argues that: The combination of the growing size of productive units leading to an expanding tertialization in employment structure and the productivity gap ..... between manufacturing and service jobs caused vertically integrated organizations to become less and less profitable, and at the same time increased the relative advantages of systems which had developed different forms 0 “organized vertical disintegration” (Sayer 1989), such as Japan or the Third Italy. In this sense, a current important factor in deindustrialization, and also one of the leading aspects contemporary trends towards vertical disintegration, is the contracting out of operations involving me intense use of service labor (201-202). This trend is not only reflected in decreasing manufacturing and increasing service employment but also and substantially in conditions of work and wage levels. Small subcontracting firms can easily hire casual, untenured and irregular workers characterized by low wages and lack of benefits (Gringeri 1994, Dangler 1996). The resulting effect of this transformation on communities not only has been the loss of “good” full-time jobs in some industries but also the diversification of working activities including self-employment, industrial homework, and casual and irregular forms of labor. 26 mam ,ramw..-'ii mum-w (u! "WI"- ;i um- Lin-xlmun wlr ,am‘n ‘.C' some? 31%ng 1'." 'mi' b- mu in. . - ‘ . L in"; 'L'i'h. my.“- ':‘)i 913?. all) Ill My", rruo rierlt no how in lJU'flBI- mi" ; gnu; gz, .muu ‘4. i.~-nn-,i;{ >1“ MOM mm ruigm ( NW)“ anibnqxa n3 0: gnitnal arm; a «incl/Mr: '1 iu,‘ , J“ "'.' "fl. 1!) Ii-bilfllfld‘m MT Oh WWW pivituubmn at” him -iVi'iu'a-W: r win-inhiw "r iroiiasillum 2".“ We human {llaaiiirw was; via. 3.: Nu? imu :gu'nmzac'im. - " m 50mm arrm am: an: in has .aidwr'imq an! hm: 22:! animal r WWW!) mm't ”numb heqolevab had (laid a: «mam; in mm v I“ m fill-MT all! to ml. in; rlouz .1980; r3 rad. ' iiom'igatniub :' Wk. Minimalist; m remit iratnuqttu mama ‘ de looms: cinema am" rimmmam \ ‘ "‘ ' mm “it! an: mam: am gnivlovni mohalsqo mm ni beta-flat vim ion «i him) airl'l' WWN We; in oetn we immoiqm acm- wMWMMvm-uh in! has sagaw 'eol vii beam Win-Momma imam- II Another structural explanation of self-employment is the increase and diversification of new technologies. New technologies and government incentives through regulations facilitated the decentralization process of economic activities. For example, emergent and successful new wchnologies encouraged self-employment activities involving computers and light technologies, particularly for those working at home. Christensen (1987) indicates that: The combination of automated offices, personal computers, and electronic communication services has led to increased decentralization of employment, including home-based work. As offices automate, valued employees increase their opportunities for flexible work alternatives, whereas full-time clericals face a shrinking job market (Leontief and Duchin 1986). Both trends promote home- based employment: valued professional employees may enjoy the luxury of working at home on their computer terminals; clerical employees may find fewer full-time salaried positions, but more opportunities as home-based independent contractors paid by piece rates(479-480). Also, technological development facilitated the logistics of overseas subcontracting and allowed de-skilling process. The development of compact, versatile components allowed formerly integrated production operations to be split and spread around the globe. Furthermore, decentralization is facilitated by government regulations and incentives. Government regulations permitted low effective tariffs on further processing abroad of semi-manufactures, and increased the attractiveness of offshore production for 27 L'su err-ram" --.i' 4 Hum m.’ yin not i.- tie-manta Immature m ”w: an rim-Jim titanium/mg hm; removing}; ”an animation) Merriam ' ‘ io‘l .dSllHl’I’JS aiinonoaa'iu «new unimmmiaawi; wit batmilioifi mornings!“ inemioiqme-Vlaa inguiiww. -- rim-"Hairy um iu’hz-iusiie him magnate“ ts audio” sand: m't msiuuiiizvi >'~.tm.‘~'.il' if"! 52w. .» iammmo gnivlrwni m rum (cit-Al. ‘l i-‘ivi , maintain?) .“ Wale but ,muiuqtnou lanoe‘iuq .2suiftu bait. »|U'UB 'tn nuiiamdmoo allT Munich consultant-Joel: W31)!!! "1 list an? . rt: 11.»: iii-'1mrnummoo WWW” .mrttotus zwitio o‘j‘l Amw aural-smut! gnihuluni MW W Warmth; how fildll'hrl ‘ili caiivnunoqqo M WM 1889i nn‘lootl b'i']¥l:05“1‘ J i t-‘hani out within!) a meW llnoraza'imq bwiav :tnain'mlqrna head MCI/mm W tannin) mm no email to swim WWW mm m .auottizoq beasts: emit-Hui imam saeiq yd bllq norm . Windham? Web tantalum! .octA .- WW-a mm bawolla has new humus-maintains...» ma . o.- ‘ fimmmamumbm i gigs TE many firms. The promotion of the ‘enterprise’ culture and associated financial incentives may engender self-employment activities. With the Schumacher (1973)’s idea that “Small is beautiful,” self-employment activities may be considered as an alternative way of producing and surviving against the suffocating and alienating control exercised by big corporations and the state. Here, the political character of some forms of informalization is underlined, but the possible relations of ‘exploitation’ and indirect control embodied in the expansion of “informal” sector are underestimated. According to Mingione (1991), another structural explanation is the neo-dualistic super-exploitative connections between informalization and new developments in the concentration of capitalism on a world scale. The later includes new forms of reproducing cheap labor, of direct exploitation either through self service or through various forms of subcontracting, or of indirect exploitation such as expansion of some privileged markets for do-it-yourself tools, electronic and information-processing equipments, financial control, and/or dismantling of state intervention Within core countries such as United States, the outcome of this decentralization process is the proliferation of individualized and small-scale economic activities in the form of self- employment, fi'anchising, and subcontracting (Dangler 1996; Gringeri 1994). Self-employment may also be considered a new option among the survival strategies available to families and households in industrialized countries. Inforrnalimtion is held to be one response to inflation, the job crisis, and the rigidity of formal work and consumption The argument here is that, under certain conditions, do- it-yourself or informal activities not only are advantageous from an economic point of 28 lei ‘n‘ti’iii but. sir-.5; ii: .‘L'Hlt . ,. u.“ ..mmm'nqarfl' nab: a'iETOI i 'I'thfin'tillih’. m :in .'i‘ mil-..' )1. ,r'1«m~qr-iq:':-..-'tlue ruboagno m SVhlmeIla as as bmabienm 30 IL"! «.22.- . l' iii-.v'iiuv'ifiu'tlw " trilituearizil beam lWflO-Q gfliZMfltlL hm. hfillll 1.1? :' 'in fun} 31:. km, 008 gram” 9*. "to eraro't amoa “to tsunami-1mm a; in: wane mi: brus 3000M i." Whoa 'qux3‘ '10 . iii..;i.i'ri ain‘w'rg wib :in biltllijjbnu a “mud?” . - 1‘ Mature am 10in "icimiitiii 30186181,“ am hi bwibodm m 2. muflfimflqxa lmwmnia reihmw ,o Iv i;‘.iitl‘;t':it'M «it gntbm ' ’9‘, I WW wan hm Wimwtru now/tad elimination; ”intro Mfg}? “madam an air 1. Dlrow i. no in, :limqua to new; WWW imitatiokzm mm: :0 mdul «madam all, ‘. ..‘“! We mibiii in '10 gm: ammudua 10 and! m_ ' o my. WI: .3100: humor-Mm mi minim W , -~ m Wk Wit: imbue .lonnou tuianani‘i .W. Idioms!” aroma heth as done at.“ Iii- . Wheatiesihnbivrbmtoaonm 'f‘ Wanamakumm g. "Wmmuuhmmotwtaa — . gag ».'~ _'. view but also they are the best way of obtaining certain goods or services quickly or the quality required E . I l l l . l E For a large number of households, restructuring has contributed to the increasing proportion of multiple-earner households (Gardner and Hertz 1992; Jensen and Tienda 1989). First, the lack of good jobs - jobs with good wages and benefits, good working conditions, chances for upward mobility, and security -- and the loss of one source of income due to unemployment of a family member may result in an increase in the number of earners within families and households. In married-couple families, for example, if the husband is displaced by current structural transformations, a previously non-working wife, an adult child, a relative, or a non-relative member of the household may be pulled into the labor force. Second, the creation of new jobs may contribute to the increase in the number of earners within families and households. The growth in service employment, for example, has been associated with, on one hand, a number of high-quality jobs offering high wages and benefits, security, and occupational mobility, and on the other hand, a number of low-paying jobs (Kassab 1992; Morris et al. 1994). Families and households with some members in secure jobs and other members also employed either full-time or part-time are able to achieve and maintain high household incomes and substantial aflluence, despite the individually weak labor market position of some of their members. At the other end of the spectrum, other households are increasingly becoming financially insecure, unstable, and working poor because they have little and decreasing access not only to job and self-provisioning opportunities and 29 9m 10 (Hoiup roar/we m alum; :v-._i-., 3m. .* in. ms; 'rd am am 43¢th ‘ m m , -; :J -I- an enigma M Wt ed! 01 bathtlirtnm m giiiiiiin: . mini-aw! ti. itirlmun natal efl‘ . r abriai'l' lineman-:1. :SWl £2121” hm: 151107151 ii -):i=iii-:euml renrns-olqitlum’tomm 8mm ,Ziflmd baa wagb’l.’ be)!» 4* i." “in. - . “hi ’JU‘a': '1 i 11ml Dill mfl M tomm‘tomtwb-bim mm m. .'i.lul'.‘1'1l2u'lst‘l Vii/wind: " ‘ . \ “Him! as or linear {um irriu‘uin v ."lfii ;. «a new. (r-iqrnuim at auburr ra't m airport-bottle!!! iii .wlortzmoii rims / mimisi ..vis. ‘I nemaa’io“ WW immme mum vrl imtiqaii, .‘i 1m mun all) r: ,9“ Wm whilst-nun a to ,w'nclan i. ,bliilu mil»: an aim and“ WW3. eon-m edt W emu] rods! ml) otm belluq at!" W Wit-W tum“? male redmmr ail) in ma mm m need all .slqmm roi .mam'golqrm m Wm .muuww «ism maria waiviW' mwwbflndetmwmomm a .bnerlrarlio school!- Wine-nono- m alrlorlewod rm zealirnsil ..—., mm“ SIM!) emit-lira unto Wade mmman Malawian-ash“ Mmmmwmwmw f( f .v if! to new advantages offered by investment in property and household technology, but also by complementary activities (Mingione 1991). 5 . ! EE . The above interpretation of restructining phenomena tends to underestimate the other aspects of social change such as (l) the consequences of the uneven spread of advanced technologies among classes and over different geographical areas; (2) the exploitative use of informalization and technological change through corporate restructuring; and (3) the critical feedback from this transformation in terms of under consumption resulting from increased unemployment, underemployment or job transfer and restructuring (Mingione 1991). The macro changes of the economy afi‘ect the livelihood and alternative choices of families, households, and communities in uneven ways. Jones and Kodras (1990) indicate that employment and income are unevenly distributed not just across social classes and strata, but also across geopolitical spaces. On one hand, it has been linked to the degradation of economic well-being, race and gender inequality, increasing poverty, especially working poor, a more polarized class structure, and declining employment opportunities (Brown and Hirschl 1995; Colclough and Tolbert 1993; Tickamyer 1996; Baca Zinn and Eitzen 1996). On the other hand, some families and households are increasingly becoming more fortunate (Pahl 1988: 251). The restructuring approach also focuses on the spatial location of economic opportunities, and its uneven impacts on individuals, families and households, and communities in difi'erent locations (Tickamyer et al. 1993; Lyson and Falk 1993; Lobao 30 "all: iud .‘cgolmmmi Uril'; im. . i“ : . "'4 at ”new: " {it er'l'io i '9“. gnu ’ 1" LT; mar/ms _ suitiini‘J-81mumAW 90mm: 01 abirai .ij'iiliilJHQ 55mm: .i .iui it, .imtmu'iqr-mi not!“ ' ”'0‘; O I me'mmt otli'io zaumunumw uzl: . i . «gs rich» «grinds lance lo M“ "O . - 14“ “3(8) :ue‘ia lacidqngoag luv-Hill m. i) u. avail) 3mm: zuigolomlausm ' j .l ”malignant! 35:16:13 in.) C-vi'HIll'.‘ .u bn’. iimlnilnm'uilnt'io 31w avw1.| .' b ‘ . “Emmi mimieiii: mm'i Airtime-I- .i. mm 'iriii'Cima :gn'nm'o “It . g“ ‘ ~00th Juanrwmmriu memu mml giiiilazar rim "3 ( ilWl mwiigniMigniwm .5. -' it «It ios't’ta (marrow 94110 zaarludv imam :dT «w W“ m m an aluminium. ban minimal pill-db Mmqum»quombw .~mharihqoqmoclamdmhle~ MHWW’MWIO ”Wm-mmdiwnhh—i ‘Vmwmimmmmm . , . , uni-madman rmimbumism WWW“ we. earmark-«aria WW 1990). Tickamyer and Duncan (1990) indicate that location in social space affects economic opportunities and life chances of persons in that locale, providing the parameters of aspirations and opportunities. The spatial distribution of economic opportunities is the outcome of different patterns of economic growth and development. Theories of uneven development focus on the specific social, political, economic, and historic context in which processes and outcome of development occur in a given location The diversity and structure of employment opportunities in an area, namely, the quantity and quality of jobs, determine options available to workers (Doeringer 1984; McLaughlin and Pennan 1991). A major consequence of the uneven development process is a spatial division of labor that produces economic inequalities between places (Colclough and Tolbert 1993). Many rural communities and regions lack stable employment, opportunities for upward mobility, investment in the community or regions, and diversity in the economy and other social institutions (Tickamyer and Duncan 1990). The poor distribution of jobs and wages result in low opportunity and high poverty rates for people and places (Tickamyer and Bokemeier 1993; Wilson 1987; Tickamyer and Latimer 1993 ). New industries are attracted by rural areas low labor costs, anti-union policies, and non- unionized workers, but also quickly relocate operations in search of cheaper labor (Bloomquist 1988; McGranahan 1988; Tickamyer and Duncan 1990). Theresultingneteffectsofrestructminginrural areashasbeenanincrease in poverty (Rural Sociological Society Task Force on Poverty 1993 ), unemployment and underemployment (Lichter and Constanzo 1987; Bokemeier and Kayitsinga 1997), and 31 «Delta 398'}. luiw r': um: "sw' H.111 'flmlbm 0’3"” ) Jean-JG but: ”Will“. 9!” gmbixmq .Jlfi’mt am; (in mix-1,; m 2 _ g.“ «'3 jh.‘ mu; ”3an oimonouo lo vioncmu..xl) 1mm“ .13! a? "nzmnqnn tum ant/:Imrqas'lo :4“ Wu!) brie drum: mum: m in arm r n trtnamb lu mmmo Mt! 31M,- ‘ 0. has m .taamloq 3500;, 41:51:11? Ma on man: wru'nqoi': ub nwanu'h Mirr'. Manimooourumol; wu'tr- woman 1,: r :»..-~: avnrr .bmu ui tram M ' I' J .‘mmmmmmummqu 'nunrmlqvm '0 ‘nm mm bus {Fiat-Nib MT M .mwm o! althl'wve encir'trro wimwmb 2:10; to zmlavp bun W" ".1 . W.“ 00810 aonwpmnm town; A ( l‘v’i’l moms‘l hm aim M . r, . , . . ’oirmmm mubmq wdnodal lanai: Hm meme-aim ..,’: . .1 {(091 :Vudlol mm- 1; WWW M? Malay! babaaumummoo [mm {HIM 4 V' ', .Wso W at: m trauma/m .vtilldom M '. “I can has WT) anomumm Imam 10th he. I, v”: ”Mum llama wot ni thaw m an 7' . 4‘29: new! ,{9- . mangled WM “When «vi-Mm am am ' " said Me: m , .- V ~31; r, 0- - MW“ m1 mow an m, declines of earnings (Shapiro 1989; Lichter et al. 1994; Whitener 1991). Non- metropolitan families and households are more likely to be in economic distress and poverty than their metropolitan counterparts (Rural Sociological Task Force on Persistent Poverty 1993, Brown and Deavers 1987). Non-metropolitan labor market areas have a relatively limited employment and earnings opportunities and less diversified labor markets (Tickamyer and Duncan 1990, Killian and Hady 1988) compared to metropolitan labor market areas. Economic restructuring has had uneven impacts on different places, and in particular, it has engendered an environment that threatens the economic survival of many families and households in rural areas. Furthermore, economic restructuring has exacerbated the already existing disadvantages of rural communities (Tickamyer and Duncan 1990; Lyson et al. 1993: 124)). Rural communities have lower population density and limited physical infrastructure (e.g., railroads, highways, postal services, water and sewer systems, educational facilities, and hospitals) than their metropolitan counterparts. Rural communities not only differ in size, physical infiasuucture, and economic base, but also in their social infrastructure ( Flora and Flora 1993). Thus, the effect of restructuring on rural communities is unevenly felt depending upon their community characteristics. Another important correlate of restructuring in rural communities is migration, especially the out migration of educated young people. Migration is a demographic process frequently associated with the concept of household strategy (I-lareven 1991; Tilly and Scott 1987). Economic explanation of migration has been concerned with differences in net economic advantages of places—with more net in migration where 32 «MM (2W: Luz-mu 3" LOW it: tea-2+); ! 7026’: mrqarlZMgnm baa Waib simmer: m at m u-Jr. ram (on. i «lzzumi ima wilt!“ .. ' trutarzro‘i no 307071 Jami lawyvinl )m’ luv?“ . anorexia-0'; “aluminium lied! w cwmmmmm 1mlnmtltqothfl'mM I'M“! wrest! but; mama .EQQI‘ . W'bofliersvib ml hm; zmnzxunmmo 4;..me bm: :mmmiqma betimil 01%?”th?! {am has nullixl ,WW mat-1:0 om vwmnbiT) . new mu bad and gnnutamtm )immoo?! arms 1:11am 1068i uni Muhamivno nu bembnagm (ml ii ,mlrumaq m ma .modq w . m bun ni ablodazuod W earlimu} mam )0 mm a .. ”M i m and aaimmmzor armmooe .awanedmll 3W”! m0 has minder! ) aenuwmmoo lawn 'lo - ,- . , ,. W rowel wart wuinummoo lmfl (0&1: I 1* ”MW .mm .35) M-» t M M (Iridium hm; ,aamliea'l “IBM '7‘ employment diversity and opportunities and wages are higher (Lyson et al. 1993). Historically, populations migrate from poor places to prosperous areas, particularly, the movement from rural areas to urban ones. The major assumption is that migration, particularly for the poor, is a response to economic opportunities (Wenk and Hardesty 1993, Nord et al. 1995). Geographic mobility may provide a better income or offer a new job, yet this depends on the opportimity structure of the place of the destination (Nord et al. 1995). Yet, poor families often move from poor nrral communities to poor or even poorer rural places where they are attracted to the meager but real opportunities that ofien exist in high poverty places such as entry-level and lower-skill jobs and inexpensive housing (Fitchen 1991; Fitchen 1995; Nord et al. 1995). At the same time, poor families may move into poor rural communities because they may experience less competition as young people with higher education and good job skills leave those commrmities to live in more prosperous areas (Cromartie 1993; Garkovich 1989; Lichter et a1. 1994; Fitchen 1981, 1991, 1994, 1995). In-migration may also increase poverty of nnal places if in migrants are older, poorer, less educated, and less connected to the labor force (Fitchen 1995; Lichter et al. 1994). Although labor market areas with relative advantages in employment and income opportunities may attract people from other areas, more recent studies on migration indicate tlmt non-economic factors, particularly family ties, play a key role in the decision to move from one community to another (Jobes et al. 1992; Wenk and Hardesty 1993). The decision to move, either for economic or non-economic reasons, is articulated at the household level and differ in the relative impact on labor or economic activity of 33 HP“ .115 13 “m, i' ‘ 4' exam: In W)" mmqqo lull! ilrztsvi yd) ,‘(l‘ialuwnnq ,aa‘rm 4.7mm; m1 1,, ram, 4 new? .‘lttl.:‘"l :r'vnteluqoq, .no'mnaim tad! 4 mm“; m; ' Hum 9r” rm did“! u: mam lmm M I ..'?” 13!:an hm. Jlnan aahmunmqrn ..‘!!Jn." . .x u .r mm a H tooq 3m 10'! vM‘z' ‘. Bimomsmoonirblbdaobwuq. am m: n '11)! 1'3 5: u .200! lsbmm' L, 1.." Marble wok; ad) to storm-x» «"llli memo an no alumni; mil to! m . "i MGWIBYU‘O 100a mtn'l mum as)!“ auditing .u "1,507 .t’Wl .hbm' ';‘*',. WMMW ad! or b name are rm: :1 rtw r. n :q lmur 151001;” ‘ “militia-19ml baa level-{um ea rl'me swig (ne'mq dgarl ni mat: mu ' .‘ MUM 890‘ hum 1901 Maori-l H'QI «Mair!» antauod who” . fimMMWWmm lmm mu; om: :wom '(sm W“ I. WWWWM'MM ulqrrxrgnuovumm‘ H m MDlmmqu mu m and or zoom ‘ . mu ,mr,19¢«1.t8wmfl:mlu nutmmobaummmamm Hem-c; we”. '5'?" - . (MN later-urinal :2Wlnm)“ Wham-mwwm . family members by gender (Tickamyer and Bokemeier 1993). Wenk and Hardesty (1993) indicate that although rural-to-urban migration is associated with an exit from poverty for women, a great proportion of women do not make such a move. When a family moves, the close ties they have developed and associated supports are undermined and ofien lost. Previous studies indicate that motivations for migration are associated with family ties and economic maintenance strategies (Jobes et al. 1992; Sell 1992; Stinner et a1. 1992; Voss et al. 1992). Such strategies involve a complex network of residences, employment, and income sources (F itchen 1991; Stack and Cromartie 1992). The restructuring of the economy has also been associated with structural changes of families and households. A demographic profile of changes in family and household structure indicate that the proportion of married-couple families has decreased since the 1960's, while the proportion of single families, especially female-headed households and that of non-family households has increased’. Goldscheider and Waite (1991) group changes in family and household living arrangements into two categories: “No families,” referring to unmarried adults living independently, a situation in which men and women avoid marriage and parenthood or living in families; and, “New families,“ referring to families in which men and women share family economic responsibilities as well as 7 From 1960 to 1994, the proportion of married-couple households hasdeereasedby33 percent. Incontrast,theproporfimoffennlehadedfufifieshshmuwdbywpumtwhilemnofnnlehaded familiesincreasedbynpereent. Also, theproportionofnon—familyhouseholdshastremendouslyincreased fi'oml960to 1994. Male-headedhmraeholdsinausedbyl27paeemwlfilefemde-headedhwaeholds increasedhy54pereent. 34 Umbra” a... Anvw .410; . ..nrn‘ ...m in" and"); 1‘) mung yd M mil furs m. dim Mimi-rwr: at Mr my": unit. ' ‘ u’.. w: [iguumib 18d! “ 1| “.1er ..wom a flow. 31am Mn viii swunm .. m-anoqmq meta amid M fl bemttnabmr era amqque bairzrjliaa; bun bonnie- r)”. 3" ..x (All ear; wub as: .m« . ‘ r We” nmungim ml moihmlma will 'r, J! «.1 a)" I" U: ( mower“. .m‘ 1 L I _‘.~ 3931392 .S‘Wl ls ta :eoduln aeigezau, :Jnrtllnlmum u'rw'uzu bus at! M 10% may I: wlovni 3913013112: drw ’ (CW7 la tn ”0'! 991 131. (MW name :19?! naduu’i) l‘nmur' ummni hm. tmmmlqm , , . WWW 1 ; .4. , f‘u' ' 1 . Jr".- . ,_ . I ' l | . ‘ l ', I .y. "r- mmahau'tommmoqauelm . " . fies-rear out $601132qu vim-soak“ ; ~ domestic tasks. Both these categories “no families” and “new families” are increasing, compared to the traditional nuclear families. Past research has tended to compare married-couple households to female-headed families. For example, recent research on poverty indicates that female-headed households, particularly those in rural communities, have been increasing and are more likely than married-couple households to be in poverty (McLaughlin and Sachs 1988; Tickamyer and Latimer 1993; Bane and Ellwood 1986; Hoppe 1993; Lichter and Eggebeen 1992; Wilson 1987; Duncan and Tickamyer 1988; Fitchen 1981, 1991; McLanahan and Sanderful 1994; McLanahan and Booth 1996). These families experience poverty and economic insecurity because of the difficulties in obtaining full- time employment reflecting limited labor market opportunities, especially for women, and because of insufficient child support from fathers, and low welfare payments. There is little research on male-headed families and that of the expanding "non-family" households. Increasingly, family scholars are emphasizing the pluralism of family and household arrangements (see Cheal 1991, 1993; Stacey 1990; Goldscheider and Waite 1991; Baca Zinn and Eitzen 1995). As non-marital fertility, divorce and remarriage rates increase, there is an increasing diversity of family forms. They include divorced- extended families living with their ex-spouses and their lovers, children, and fiiends, cohabiting households, single and unwed parents, divorced parents, matrilineal, extended and kin support networks, and dual-earner households. Also, families and households are changing in their structures due to transformations in the social and macro-economic 35 .amamwm3185.9“er New 1,“: ..'-11m“ ‘ I'f-Jg‘um until r1108 at. 51.: 1113‘;qu Zunmlilfl‘lltflla‘ ' I. o, mil-Dismal 01 ablorbrwml 'slqum im‘xmn 71: gm -, m mum? .Inrl domes! M I”: .1" v 1“ . ‘ I beheadurlamnt lad) 43mm)”. {TV/VJ" ..., 'I um ,n'api .qumn‘l w v 2' “s mmhnsso'taaemmmodwui . ,_ m. m. w, 1...)» u. mm (haluomuq diW“ :Im “Mammals: 19M) man; .9 gr! 1: («Curl—Mezi '-l:;u:r+t.amnmnflu -‘-‘ l: . bum :5Wl @1133“; lmmll'i hm; (”USU 1991 1'4“:th DRIM"" .‘ ..' ;mt,taet “Mimi? ,8891 1')’I_fil;.;i‘rl 1' hm: ummzru .1 WI Maw £991 on“ I; ’ mm (6991 rlJoofl bus Manuela”. .M’Ql lu'hzhnsa baa M i; ‘- . ... ' l y I ‘mdwbwlo earned virtue-um 'mnonmu i'mu ‘(nwoq m- - - ‘ WW minds; Instr-m gmzwrm "miniature“ 0‘! WW mam mm} W bl'uto lmior’ttutm to ma «WWIIIIM tub baa Mind Maxi-elem no claim elfli ablM' ,Wmum WM clam-ma! - -~ = mm“ ER! .1001 Wulnaenmmhlom ' Wm .(tQQl m4 bun mus ma ma W Whit-“W «- ei M M 7 mmm mumwu aetliau‘ibml if}- him ”newline-aw harsh-a WM ( _ 4 . 6 ‘ I 28 forces, and changes are most profound in poor and working-class families (Baca Zinn and Eitzen 1995; Stacey 1990). It is thus, important to analyze how economic restructuring, with its uneven impact across space and strata, as well as that of families and households have enabled or constrained families and households to adapt to various environments in which they are embedded In the next theoretical perspective, families and households are considered as the locus of social relations that help members to organize and participate in various activities in order not only to survive, but also to adapt to changing environments surrounding them. At the same time, families and households are embedded in larger social contexts, most notably, the labor market opportunities, social networks, and culture in their local communities. Families and households have different capacities to engage in work strategies. Work strategies cannot be understood unless we consider family and household different capacities, including their composition and structures, their economic resources, as well as their social and cultural capital. This not only would require to analyze family and household social structures and how these constrain or enable work strategies, but also to analyze families and households as social entities well connected and dependent on larger social contexts. In this section, I use two conceptual frameworks: social capital and social embeddedness perspectives to analyze family and household work strategies. I first argue that work strategies are embedded in family and household social structures. Secondly, I argue that families and households are embedded themselves in larger social contexts, most notably, labor market area social structures. 36 nmS rum}: ¢;-:lu.:i:'? ..'/5:2 ..--m:.luw 1.:u. 1m;- n. "mulmrv 'mm um . — 4 \ alumnae: ”ml .rl»;-:. « v « )1 ... . mt? ) 0"“: it'mlaim ..' eatlimfl'io mix “a: “an 2.. nilfllv i.....' ...g- . 1.13.. t..:..'mr: moan am I”!!! M, ’ o 0 molar: 0! M8 or ebloriccmrri bm ... Ir “ L‘JLHBHLIH“. :0 L-alrlrme 3v“ Bum W lannamarll 11.35. -H.‘ n’ in! ileum-r an. I_'.n'll :loirlw man-d], mmwm: anemia 11' ma in (cow! :1) r Unahraaoomw mm ,Mflue at Villa ton n. m m ..Isritwm. Illnllh‘l m :mqiomaqhfli . l 3' mm .smit m an: 1A merit anrhnuorrua. aruamnmwns saw“ . . 1 mm!!! .vkhtoa team .msmos lusw n. .91 m halibut-linen“, . M haul "am at amtluo baa uruman lama! MW; WWW 2:4. 12‘ wwd W With wart Medieval has aeriim‘l ..‘ n wk ;-1( ' Y“ “mowmbnomabmndmmawmw ‘4‘ ,. MW!“ mammium mm gm’bulaai w o’ $' WWWMUIH .latiqaolaummbnhiauam. Q MMWMMamtmul-ruw‘ wwmuuummwmmmm '53“- WWMmIMmMaI misuse“:'.f} Warn-mam ‘“ ‘ “ ‘ " WeMMWsa-muum . o£ 5 i I ; . l I! Social capital was initially identified by Jane Jacobs (1961), Pierre Bourdieu and Jean-Claude Passeron (l990[1970]), and Glenn Loury (1977), and developed more extensively by James Coleman (1988), Ronald Burt (1992), Robert Putnam (1995), Alejandro Portes (1998), and Portes and Sensenbrenner (1993). It has been extensively applied in many areas of social sciences and has attracted many interdisciplinary scholars as well as policy makers seeking non-economic solutions to social problems. However, there are difi‘erent conceptualizations of social capital. What they have in common is that they all seek to highlight properties of the social structure that facilitate or hinder social action (Wacquant 1998). Despite the different conceptualization of social capital, there is a consensus that social capital stands for the ability of actors to secure benefits by virtue of membership in social networks or other social structures (Portes 1998). What is social capital and what is not social capital? And most importantly, how does social capital affect work strategies? The concept of social capital is not new for sociologists. In their review, Portes and Sensenbrenner (1993) and most recently Portes (1998) contend that there are four different sources of social capital corresponding to each of the major theoretical traditions in sociology: value introjection, bounded solidarity, reciprocity exchanges, and enforceable trust They argue that from Marx and Engels, we derive the notion of “bounded solidarity,” —- the altruistic dispositions of actors are bounded together by the limits of their community. Other members in the same community can then appropriate such dispositions and the actors that follow as their source of social capital. From 37 -520me bnaueihruoa unar‘l HOW I edtnsl. an: ml lr.:srr..-;.i.; Alumni um lmrqn srorn boqolsveb bnn rrrui. mm : mam bnr: :ymtrmr . nmaeu‘l [MUM‘I hsdofl iivt'l r hu“. blrur :r'z’ 138‘“ r “ramble” carnal ltd MMMtl 45.951; unnsrdnaener’. but who? bns ,(8Wl)ufl“_, WWW!“ (Man lutarmtr. mrl Om. momma l..‘ )0. to costs (run-“J“ .‘lm .unoflmq loiooz at anortuloe armouow-mm gamma ‘03er {01M I“? I . ..‘ t ’ dim-Oi am {all MW .131qu larooe 'Io wortuilarnqaunw imam: i ' Warm M m loiooc all! lo aoihsqorq trlsnlrigrd or also: “I!“ “l Whmmwah watt r8901 tnwpaaN)M“rii; : 'l I Wmmhmafimmmmmmufi ’3 WMMWWWmmwmqubmmb“ A. mmmm twmwmmmmmn—I ~ v . Snmmmwo'fhl-p if WWflmnul-mhmflommm . -. Wmmmmrmnmu -; Simmel we learn of “reciprocal transactions” — the norms and obligations that emerge through personalized networks of exchange. The undersocialized view of human nature sees social capital as primary the accumulation of obligations from others according to the norm of reciprocity. In this case, donors provide privileged access to resources in the expectation that they will be fully repaid in the future. From Durkheim and Parsons, we learn of ‘Wue introjection” ——- the idea that values, moral imperatives, and commitments precede contractual relations and inform individual goals other than the strictly instrumental. From Weber, we learn of the idea of “enforceable trust” — formal institutions and particularistic group settings use difi‘erent mechanisms for ensuring compliance with agreed-upon rules of conduct - the former (e.g. bureaucracies) using legal/rational mechanisms, the latter (e.g. families) substantive/social ones. As in reciprocal exchanges, the motivation of donors of socially mediated gifis is instrumental. However, the expectation of repayment is not based on knowledge of recipient, but on the insertion of both actors in a common social structure. According to Portes (1998), the embedding of a transaction into such structure has two consequences: (1) the donor’s return may come not from the recipient but from the collectivity as a whole in the form of status, honor, or approval; and (2) the collectivity itself acts as guarantor that whatever debts are incurred will be repaid As a source of social capital, enforceable trust is appropriable for both donors and recipients. For recipients, it facilitates access to resources. For donors, it yields approval and expedites transactions because it ensures against malfaisance (Portes 1998). 38 ag'mma 11.11; rululfligddo hm. vnu ,..: r n-n am! hi, ”final" to“ v maximum ”tum an 1 gyfnjd‘p'g- man nsmud'fo «ICU 'v; napalm , manibmaan 2101110 m w 'rrltfl'nfiguo'n u.- .unml'unmw: .u'h nbmhqa at“ rti Wm 01 aaooon baguhmq am a. n7 ("mob an,» .41.!) M 30'0"“. :m M has mmhhufl mofl 'anii and m burg.“ ~(Tv ' ad "WI M! 'v‘ hmmasmnn {mum .doulsr/ um um» wm ' nun-19mm 4 “mm hmbinbni mmlm (nu. aummm muuznmm W WWW W" 10 83b! :11. "u nmuf aw junta}! mofl JIM “album :nm'nm aau agumw rzumg mmmlua'maq M :W .33) tame? am -- '1' WW indium} J; 5) mm: um ,zmzinflwm louhnoa 10 PJN' mun-hm m. , p ‘ 7 . I 4 .. _< , .‘JW "‘55- v"'1»‘.» ~’é‘~‘~"..!‘r ’5' ‘4‘- Among the contemporary scholars, Pierre Bourdieu (1986, 1990, 1993), James Coleman (1988, 1990), and Robert Putnam (1993, 1995), and more recently, Portes (1998), Michael Woolcock (1998), and Jan L. Flora ( 1998) offer distinct but complementary approaches of the usage of social capital. Bourdieu (1986) considers social capital as part of a generalized theory of capital that locates individuals’ positions with respect to their possession of available resources, including not only economic, but also social, and cultural capital. Bourdieu defines capital as “accumulated labor (in its materialized form or its “incorporated,” embodied form) which, when appropriated on a private, i.e., exclusive, basis by agents or groups of agents, enables them to appropriate social energy in the form of reified or living labor” (241). He indicates that it is impossible to account for the structure and ftmctioning of the social world rmless someone reirrtroduces capital in all its forms and not solely in the one form recognized by the economic theory (242). He distinguishes three general types of capital: economic capital, cultural capital, and social capital. Economic capital refers to monetary income as well as other financial resources and assets. Cultural capital exists in various forms: in the embodied state, i.e., in the form of long standing dispositions of the mind and body; in the objectified state, i.e., in the form of cultural goods (pictures, books, dictionaries, instruments, machines etc); and in the institutionalized state, i.e., the objectification of cultural capital in the form of academic qualifications. Social capital is the aggregate of the actual and potential resources that are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition (Bourdieu 19862248). Bourdieu indicates that "the profits which accrue from 39 asmal. .(EWI 0W: 03w I null-u u‘d 3711.9! armada). ~nmoqraatnooallm estro‘l .‘(lmsun ‘nnm Lin; t-‘i \5‘? '1‘” I (new)? hair” has .(099! .MI-I . Jud rannaib 12TH) (6W: 5 wall 1 ml, bna .1899! ) rioou’oow WM. Miamwamluommfl liZ’IQSJ hum»: '0 when grit?!) autumn. M2; mom 'vleub'rvibni cameo] mm lumpy, in (mm: haulmmag «'10 mo as him“ ’ i. Mam ‘(lno Don gnibuluni ..~:.~:rmar uldm’n. n. in nrmauuoq its!!! at M” 3 d'mwmmm“ as (snow cordial) maimmh’ lnnq»: [marina ill m Om mtiw .doidw (moi butlmdmu " l.‘.1n.,«":wr:m” Pti 10 mo? hm 3," WNW .umyflo aqumg m emgu m and .avrauloxa raj.“ fifivfl.mbni eH .( MS) "Iodal gniwl to beam 1!) (mot at!) ni W * ', 7mm him filth 310mm has autumn art] 103 "moons at M ' - W‘aifldmmhnaemw‘l an “I“. albumen-1mm ‘. ..‘ Wampum m Miugniraib m ‘95; need: ammo. d “m amid-pa mama mm brace m. .quaa mum mu WWWW ~~h mow) lcr'msnfi radio as M'fl -mmmwbm “$.91:er human WMWh-flw m .31 pm: beams-ad Ww‘mwuuxmmm .W _ sum-rampant main-mam.» um imam Ll; Whmhflnum Who: was 2:25 ‘3 mmhmbmwu "m -’ o membership in a group are the basis of solidarity which makes them possible.” He adds that social networks are not a natural given, or even a social given, —- they are the products of investment strategies, individual or collective, consciously or unconsciously aimed at establishing or reproducing social relationships that are directly usable in the short or long term (Bourdieu 1986:249). According to Bourdieu, the different types of capital differ in the degree of convertibility. Economic capital is the most convertible from the transformation into social and cultural capital. Through social capital, for example, actors can get access to economic resources (subsidized loans, investment tips, protected markets). By comparison, social and cultural capital possess their own dynamic and are characterized by less transparency and more uncertainty. While it is difficult to convert social into cultural capital, the transformation of cultural into social is easier. Strategies of reconversion are one of the means through which individuals, families, or groups attempt to maintain or improve their social position. The conversion rate between various species of capital, in turn, is one of the central stakes of struggles between groups, each seeking to impose the hierarchy of capitals most favorable to its own endowment or profile (Bourdieu 1986). Bourdieu argues that people struggle for economic, cultural, and social capital to maintain or improve their social positions within “fields.” These fields constitute networks of relationships among positions (Bourdieu and Wacquant 1992). The overall volume and structure of capital detained by an individual, group, or institution define their position in social space; changes in the volume and structure of one endowment chart one trajectory through it (Bourdieu 1986). Anheier et al. 1995 also 40 find that variations in capital are reflected in positional arrangements within fields. The presence or absence of these resources defines the social context in which individuals live. These objective conditions give rise to particular tastes, lifestyles, and ways of looking at the world as well as taking action Bourdieu also uses the concept of “habitus” -— a concept that refers to a system of lasting, transposable dispositions which, integrating past experiences, functions at every moment as a matrix of perceptions, and actions and makes possible the achievement of infinitely diversified tasks. Bourdieu argues that agents act within socially constructed ranges of possibilities, durably inscribed within them (even in their bodies) as well as within the social world in which they move. “The habitus is the source of these series of moves which are objectively organized as strategies without being the product of a genuine strategic intention —— which would presuppose at least that they are perceived as one strategy among other strategies” (Bourdieu 1977). An individual’s “habitus” is a direct product of the person’s structtual situation Different life conditions give rise to different forms of habitus and those exposed to the same conditions will develop the same habitus (Bourdieu 1990). The habitus, in turn, has a direct, constraining effect on the social action of individuals, which coming full circle, contributes to the reproduction of the social structure. Thus, it would be expected that work strategies are likely to be affected by the family/household structure. Coleman (1990) indicates that social capital constitutes a particular kind of resource available to an actor. He argues that social capital is defined by its function. It is not a single entity, but a variety of entities with two elements in common: they all 41 consist of some aspect of social structures, and they facilitate certain action of acto -- whether persons or corporate actors -— within the structure" (Coleman 1988:898, 19902302). For Coleman, social capital, like other forms of capital, is productive, making possible the achievement of certain ends that would not be attainable in its absence (Coleman 19902302). Unlike other forms of capital, social capital inheres in the structure of relations between and among actors (Coleman 1988:898). He identifies three forms of social capital including obligations and expectations, which depends on trustworthiness of the social environment, information-flow capability of the social structure, and norms accompanied by sanctions. He adds that some aspects of social structure facilitate social capital formation. They include close social networks and appropriable social organimtional arrangements. They do that through multiplex rather than simplex relations (Coleman 1988: 8108-8109). Coleman (1988) conceptualizes social capital at both family and community level. Social capital of the family is the relation between children and parents (and, when families include other members, relationships with them as well)(pS1 10). Social capital within the family depends both on the physical presence of adults —— structural deficiency -— and on the relationship between children and parents (or other adults in the household). He argues that single-parent families or those where both parents work outside the home lack the social capital that comes with the presence of parents during the day, or with grand parents or aunts and uncles in or near the household (Coleman 1988: S111). McLanahan & Sanderfur (1994) also indicate that social capital tends to be lower for children in single-parent families because they lack the benefit of a second at- 42 home parent and tend to change residence more often, leading to fewer ties to other adults in the community. However, social capital is more than the presence of adults in the family. Coleman (1998) indicates that even if adults are physically present, there is lack of social capital in the family if there are not strong relations between children and parents. Coleman (1988) also indicates that social capital exists outside the family -— “it can be found outside as well as in the community consisting of social relationships that exists among parents, in the closure exhibited by this structure of relations, and in parents’ relations with the institutions of the community (p81 13).” He argues that for families that have moved often, the social relations that constitute social capital are broken at each move. Leaving a community tends to destroy established bonds, thus depriving family and children of a major source of social capital (Hagan et al. 1996; McLanahan and Sanderfur 1994). Thus, social capital exists at both the family and community levels. At the community level, Putnam (1993) defines social capital as features of social organization, that include trust, norms, and networks. All of these aspects facilitate action and cooperation for mutual benefit Working together is easier in a community blessed with a substantial stock of social capital." (Putnam 1993: 35-36). He found that communities in certain regions of northern Italy were able to maintain and accumulate social capital which has been put to use in fostering economic development. By contrast, social capital in poor regions of southern Italy remains bounded in patron-client relations which hamper economic and political development He suggests that this difference can 43 be understood in light of norms. Of special interest is “generalized reciprocity,” that is, the assurance community members have that their altruistic actions will be rewarded at some point ensures their willingness to contribute to others’ welfare. People are thus less likely to opt out of civic responsibilities and social attachments, thereby creating certainty and stability, as well as becoming models for future cooperation. Historically, families have relied kin and relative ties to access or maintain their economic resources. The family plays a considerable role not only in the transmission but also in the management of the economic heritage, especially through business alliances which are often family alliances (Bourdieu 1996, 1998). Social capital originates in kinship ties but extends beyond kin and relative ties. Granovetter (1974) used the term "strength of weak ties" to refer to the power of indirect influences outside the immediate circle of family and close fiiends to serve as an employment referral system. The common sense would be that, dense networks such as available through family circles, would be most effective in finding jobs. Burt (1992) built on Granoveter and developed a concept of "structural holes." According to Burt (1992), social capital is based on the relative paucity of network ties rather than on their density. He defines social capital as friends, colleagues, and more generally contacts through whom you receive opportunities to use [other forms of] capital (Burt l992:9). According to Burt, it is the relative absence of ties, labeled "structural holes,” that facilitates individual mobility. This is so because dense networks tend to convey redundant information, while weaker ties can be sources of new knowledge. An extensive literature on social capital outside the immediate family comes from ethnic entrepreneurship and enclave studies. Networks and the social capital that flow through immigrant and ethnic groups are identified as a key resource for the creation of small businesses. Light (1984) and Light and Bonacich (1988) emphasized the importance of rotating credit associations for the capitalization of Asians immigrant firms in the United States. Social capital comes from the trust that every member has in the continuing contribution of others even after they have received the pooled funds. Enclave studies consistently highlight the role of ethnic networks as a source of vital resources, including start-up capital, tips about business opportunities, access to markets, cheap and disciplined and mostly co-ethnic labor force. For example, entry level openings are frequently filled by contacting kin and friends in remote foreign locations rather than using available local workers (Sassen 1995). Also, mobility opportunities through niches are entirely networks driven. Examples of such ethnic enclaves include New York Chinatown (Zhou 1992), Miami little Havana (Portes 1987, Portes & Stepick 1993, Perez 1992), and Los Angeles Korean town (Light & Bonacich 1988, Nee et al. 1994). The literature on studies that emphasize the “social isolation” concept highlight the use of social capital in poor communities. Survival in poor urban communities frequently depends on the close interaction with kin and fi'iends in similar situations (Carol Stack 1974). However, the problem with such ties is that they seldom reach beyond the inner city, thus depriving their inhabitants of sources of information about employment opportunities elsewhere and ways to attain them (Portes 1998). Wilson 45 (1987, 1996) emphasize how the departure of both industrial employment and middle- class families fi'om Black inner city areas have left the remaining population bereft of social capital, a situation leading to extremely high levels of unemployment and welfare dependency. Fernandez-Kelly (1995) also indicates that the dense but truncated networks of inner-city Black families not only cut off members from information about the outside world, but alternatively support alternative cultural styles that make access to mainstream employment even more difficult The concept of social capital is without limitations. Woolcock (1998) argues that social capital is regarded as other capital that can be maximized -— “if a little trust, group participation, and cooperation is a good thing, should not more of it be better?” Social capital has both positive and negative effects (Portes and Sensenbrenner 1993; Portes 1998). There may be different types of social capital and that collectively they are resources to be optimized not maximized (Woolcock 1998). There are different types, levels, or dimensions of social capital, different performance outcomes associated with different combinations of these dimensions, and different sets of conditions that support or weaken favorable combinations (Woolcock 1998). SEW The concept of embeddedness was originally introduced by Karl Polanyi et al. (1957), then used by Granovetter (1985) and Mingione (1991), and more recently by Portes and Sensenbrenner (1993). Granovetter (1985) argues that: The embeddedness approach to the problem of trust and order in economic life, then, threads its way between the oversocialized approach of generalized morality 46 and the undersocialized one of the impersonal, institutional arrangements by following and analyzing concrete patterns of social relations. Unlike either position, or the Hobbesian position, it makes no sweeping (and thus unlikely) predictions of universal order or disorder but rather assumes that the details of social structure will determine which is found (p493). Following Mingione (1991) and Granovetter (1985), Bloomquist et al. ( 1993) argue that work activities are purposive actions of individuals or groups, but these actions are embedded in particular social contexts. The concept of embeddedness emphasizes social relations and network ties within and between social groups, organizations, and institutions. Granovetter (1985) indicates that social structures can advance and constrain individual goal seeking and they can even redefine the content of such goals. He argues that Actors do not behave or decide as atoms outside a social context, nor do they adhere slavishly to a script written for them by the particular intersection of social categories that they happen to occupy. Their attempts at purposive action are instead embedded in concrete, ongoing systems of social relations (1985: 487). Granovetter (1985) indicates that social networks can be viewed as social capital that actors can use to pursue their own goals or interests. He indicates that action is embedded in social relations. These social relations in turn can limit and influence actor’s choice of strategy as well as provide actor with opportunities to further his or her interests and influence others. 47 Social embeddedness refers to more than the network of social relations -— it highlights the processes and social-structural context that maintain the interdependence of those structural ties. Bloomquist et al. (1993) indicate that: Work activities are embedded in household structures (including the internal distribution of power). These are in turn embedded in other reciprocal networks and in associative structures of redistribution (p96). From this perspective, family and household work strategies depend on not only the family/household social structure in which these activities are embedded, but also on labor market area social structure in which families and households reside. The analysis of family and household work strategies requires a multi-level approach as Bloomquist et al. (1993) put it: The conceptualization of the embeddedness of work activities requires one to consider not only how work opportunities are organized in a particular social context, but also how household members organize their work activities in the context of the overall reproductive strategy (p96). Mingione ‘s (1991, 1994) focuses on the social embeddedness of economic behavior. He indicates that industrial societies are based on systems of social integration founded on complementarity between forms of institutional regulations and associative organizations and the role of adapted reciprocity networks. In his conceptualization of work activities as a component of household reproductive strategies, he highlights the considerable importance of the regulatory process based on reciprocity in addition to what he calls the "associative regulatory process.” 48 I am convinced that nowhere can industrial development be fully understood without devoting great attention to the adaptation and change of reciprocity networks and to family and kinship strategies, which have maintained a crucial role as the fundamental social and organizational background to the biological reproduction of human kind As such, complex mixes of reciprocity loyalties greatly condition individual economic behavior (1994:24). Mingione’s approach highlights the fact that work structures are embedded in social contexts as complex mixes of associative and reciprocal kinds of relationships organized to regulate the tensions produced by competitive market behaviors. He indicates that the basic unit of social reproduction is, for the most part, located in the household. The most important social network in which the household is strictly embedded is the family (19912133). Recent studies also indicate that families and households are critical and strategic social organizations, through which individuals shape and adapt to social transformations (Bokemeier 1997). The social embeddedness perspective provides a lens to analyzing family and household work strategies in both relations of production and social reproduction. Family and household work strategies are conditioned and constrained by available resources including social relations within households, considered as units of social reproduction, as well as its social relations with surrounding environments, particularly, labor market conditions and processes. Social reproduction involves the process of reconstituting the social relations of society necessary for human social and economic activities. It covers all activities necessary to sustain the household and the economy, 49 including childbearing, child rearing, housework, household consumption, emotional labor, and other non-labor market activities (Tickamyer 1992; Collins 1990). It encompasses diverse conditions and organizational relations that make possible household members survival in the social and economic environment (Mingione 1991: 124) as well as their adaption to changes in such environments. The internal activities and resources controlled by household members are crucial factors for comprehending the diverse household conditions that permit members of households to adopt an adaptive strategy. At the same time, the location of the household in the broader social, cultural and economic environment affects the "strategies" its members can do and adopt. Mingione (1991) argues that: The combination of resources utilized in reproduction is connected with, among other factors, labor market conditions and processes . . . within the household as a unit of social reproduction, embedded in different supportive reciprocal networks, decisions are taken according to the varying and changing internal distribution of power on which is the best possible allocation of available resources to meet subsistence needs, given the existing relations between work opportunities and income and the parallel possibilities of saving subsistence expenses through labor for direct self-provisioning, and/or given the existence of other reciprocal or redistributive resources. In this sense, the complex structure of reproduction expresses, among other processes, the formation of the labor supply at given condition of labor demand and of overall work opportunities (Mingione 19912141). 50 Given variable social context in which households are embedded, household members organize their work activities in the context of the overall reproductive strategy (Mingione 1991). The restructuring of the economy has eliminated or redefined many jobs in the paid labor force, allowing workers and their families to support themselves in different ways (Garrett et al. 1993). That situation has encouraged some household members to enter the paid labor force, engage in self-employment work, and seek different sources of income, given the conditions, organization, and availability of social, cultural, and economic resources of labor market area structures in which they are embedded Social embeddedness perspective is useful for understanding how social structure affects economic activities. However, it does not concretely explain how social ties affect economic outcomes (Uzzi 1996). He argues that The core statement —- that economic action is embedded in social relations which sometimes facilitate and other times derail exchange -- is conceptually vague (p674). In a review of sociology of markets, Lie (1997) indicates other criticisms of the embeddedness approach. It neglects nonsocial or nonstructural factors such as culture, technology, and even macroeconomic forces. It eschews analyzing historical and cultural variations in markets. Power, especially in non-economic realms, remains elided, and the role of the state is not accounted for. The embeddedness approach must itself be embedded in larger, historically transient social structures -- not only state institutions 51 and superstate organizations, but also historically shifting transnational relations and structures (Lie 1997: 351). Woolcock (1998) argues that all forms of exchange are inherently embedded in social relationships. He cited Braudel in indicating that “it is easy to call one form of exchange economic and the other social. In real life, all types are both economic and social.” He cited Zukin and DiMaggio” indicating that embeddedness could take several distinct forms: social ties, cultural practices, and political contexts, all had a powerful effect on shaping the types of opportunities and constraints individuals faced as they sought economic advancement He also indicates that many benefits gained by embeddedness in a given network were not indeed without corresponding costs. . The high degree of density and closure characterizing the social relations undergirding the relatively simple, small-scale, informal exchange in village markets, for example, could in facts impose considerable constraints on successful members of these communities as they attempted to make the transition to membership in larger, more extensive, and sophisticated exchange network coordinated by formal institutions and the rule of law (p13). Woolcock proposes a framework that integrates “embeddedness” and “autonomy” as distinct forms of social capital at both micro and macro-level of analysis (p15). Embeddedness at micro refers to intra-community ties, while at the macro level it 8 Zukin and DiMaggio (1990) classify embeddedness into fourforrns: l) structural -- material quality and structure of ties among actors; 2) cognitive -- structured mental processes that direct economic logic; 3) cultural—shared beliefsandvaluesthatshapeeconomicaims;and4)political-—-institutionallimitson economic power and incentives. 52 refers to state-society relations. Autonomy at the micro level refers to extra-community networks, while at the macro level it refers to institutional capacity and credibility. The micro-level, what he calls “bottom-up dilemmas of development,” encompasses individuals, households, small groups, and communities. Bottom-up development typically functions in and through social relations among people with common neighborhood, ethnic, religious, or familial ties (i.e., those with high endowments of social integration). In such cases, integration constitutes an important source of social capital, enabling participants to provide one another with a range of services and resources — the more intensive the social ties and generalized trust within a given community, the higher its “endowment” of this form of social capital (p21). However, he indicates that more is not generally better — where generalized trust extends only to immediate family members and blood relatives, a stark of non- developmental reality is likely to be present. He refers to “amoral familism”-- presence of social integration but the absence of linkage (extra-community networks). “Amoral individualism” on the other hand exists where there is neither familial nor generalized trust, where narrow self-interest literally permeates all social and economic activity, and where members are isolated — either by circumstance or discrimination -- from all forms of cohesive social networks. It is thus characterized by absence of both integration and linkage. He also describes anomie (as in Durkheim) referring to the situation where individuals have newly found freedom and opportunity to participate in a wider range of activities but lack the stable community base to provide guidance support and identity, i.e., they have linkage but no integration. For example, without a strong community 53 group to provide initial financial resources, small businesses fail to get started or go bankrupt in the early stages. Strong intra-community ties, can be highly beneficial to the extent they are complemented by some measure of linkage. He argues that, people who are able to forge new social ties into the wider business community, even in less dramatic circumstance, are the ones who enjoy greater economic success. In summary, this study integrates (1) the restructuring perspective, and (2) the social embeddedness/social capital perspectives to explain family and household work strategies. First, family and household work strategies are conditioned by the social structure of environments in which household and families are embedded and interconnected. At the same time, a spatial distribution of opportunity structure which is the outcome of both historical, social, economic, and political changes enable or constrain family and household work strategies. Secondly, family and household work strategies are conditioned and constrained by available resources including economic, but also, social and cultural capital of families and households. The restructuring and social embeddedness perspectives, taken together, suggested the following propositions and hypotheses: P11.: Family and household structure determines work strategies. Families and households have different capacities to engage in work strategies. The diversity of family and household social arrangements may facilitate or inhibit the adaptation to social environments in which families and households are embedded Family and household composition, as indicated by the presence of marital partners and 54 Vt: other related adults, such as adult children, parents, and in-laws help its members to survive and adapt to different environment (Coleman 1990‘; Femandez-Kelly 1994). At the same time, social relations, in particular, those of gender and power, degree of mutual obligation, respect, trust, collaboration, and solidarity enhance their chances of adapting to changes in the social context in which they are embedded Social capital resides in those relations between household members and may result in family/household members’ contribution in terms of labor and financial inputs (Sanders and Nee 1996:233). According to Femandez-Kelly (1994), social relations vary across social groups, social and physical locations. Thus, it is expected that the family/household social structure will have a differentiated impact on the likelihood of families and households to have other members (other than the head, the spouse, or partner) employed and/or a self-employed member. The hypotheses to be tested in the analyses are summarized as follows: H1 1 la. Married families are more likely than “non-married” families and households to have a self—employed member. H112a. Cohabiting families and households are more likely than single-headed families and households to have a self-employed member. H113a Female-headed families and households are less likely than male-headed families and households to have a self-employed member. H11 1b. Married families are less likely than “non-married” families and households to have a self-employed member. H112b. Cohabiting families and households are less likely than single-headed families and households to have a self-employed member. H113b. Female-headed families and households are less likely than male-headed families and households to have a self-employed member. 55 In addition to family and household types, other household factors influence family and household work strategies. Family and household types are likely to differ on a variety of factors that can influence the likelihood of self-employment and having additional earners, such as the presence and age of children and the presence of adult relatives. I expect family and household work strategies to vary by, in addition to family/household structure, these other composition and life cycle factors because they differentiate the social environment in which family and household work strategies are embedded The presence and age of children affect not only family and household economic needs but also the ability and availability of parents to respond to such needs. I expect that families with young children, especially preschool children, because of the amount of work and time involved in their care, are less likely to involve their members, particularly mothers, in the paid labor force. However, because of flexibility of small family businesses, some of them done at home or nearby, I expect families with children, especially preschool children, to have a self-employed member. This suggests the following hypotheses: H114a. Families and households with children under 18 years of age are more likely than those without children to have a self-employed member. H1 15a. Families and households with preschool children are more likely than those with school-age and/or adolescent children to have a self-employed member. H116a Families and households with school-age children are less likely than those with adolescent children to have a self-employed member. In addition, family and household work strategies vary by family and household structure and the presence and age of children because of these factors, but also because 56 the presence of other adults, especially adult relatives help either in family/household responsibilities including the care of young children or participate in family/household work strategies such as involving in family business or accessing the paid labor force. Thus, the presence of adult relatives in the household is used as a control variable. Families and households have different capacities to engage in work strategies, including not only the family-based social capital, but also the external social capital, including the social networks with relatives and friends in their community. Therefore, I propose that: P12: Family and household social ties to the labor market area affect work strategies Family and households move within and across labor market areas. People move from different reasons, mainly economic but also social reasons. As indicated earlier, economic resources in a community, enable and constrain families and households to adopt any strategy. Even in poor communities, especially rural communities, families and households that move in are willing to take any kind of jobs, including low-wage part-time jobs and self-employment activities because they satisfy greater proportion of their needs (F itchen 1991;1995). Although, families and households’ residential move may be associated with a better job for the head of the household, an affordable house, or a better school for children, it is likely to be associated with a decrease in family members in the labor force, particularly wives and women partners (Tickamyer et a1. 1993). At the same time, family and household geographic mobility is also associated with less social capital in the community (loss of old fiiends and adjustment to new 57 fiiends and neighbors, and co-workers), therefore greater chances of few members in the paid labor force and in self-employment. In contrast, family/household residential stability is likely to be associated with greater social capital (e. g., participation in exchange networks with kin, friends and neighbors, and sharing information on jobs and businesses’ opportunities. Thus, I hypothesize that: H12 1a. Family/household residential mobility decreases the likelihood of self- employment. H121b. Family/household residential mobility decreases the likelihood of additional earners. Family and household work strategies are affected by family and household characteristics but also they are conditioned by larger social and structural contexts including the characteristics of labor market area in which they are located, most notably the labor market area opportunity structure and its associated spatial location The following propositions include the impact of labor market area characteristics on family and household work strategies. P21: Non-metropolitan-metropolitan residential location determines work strategies. The restructuring of the economy has placed much greater burdens on non- metropolitan families and households. Employment in agriculture and in manufacturing industries that has for a long time sustained the well-being of a large number of non- metropolitan families and households underwent sharp contractions while the service sector has increased However, the rise in service sector employment in rural areas has 58 been limited to low-wage jobs while the urban areas have experienced a rise in jobs at both ends of the wage spectrum (Gorharn 1992). The resulting effects of the economic restructuring in rural areas have been greater rates of poverty, unemployment and underemployment and lower levels of earnings for non-metropolitan families and households when compared to their metropolitan counterparts (Lichter and Constanzo 1987; Lichter et al. 1994; Fitchen 1991; Whitener 1991; Rural Sociological Society Task Force on Poverty 1993). Thus, H211a. Non-metropolitan families and households are more likely than those in metropolitan areas to have a self-employed member. H211b.Non-metropolitan families and households are less likely than those in metr0politan areas to have additional earners. The between labor market area differences in family and work strategies, especially the metmpolitan and non-metropolitan differences could be due to the labor market area differences in opportunity structure. I, thus, propose that: P22. Industrial structure of a labor market area affects work strategies. The restructuring of the economy has been associated with a shift in the quantity and quality of jobs. The decline in extractive and manufacturing employment, and the rise in service jobs has negative implications for the economic security of a large number of families and households (Brown and Hirschl 1995). This shift in employment structure has resulted on one hand, in a number of hi gh-quality jobs offering high wages and benefits, security, and occupational mobility, and on the other hand, a number of low-quality jobs (Kassab 1992; Morris et al. 1994). At the same time, the restructuring of the economy has changed the patterns of work The displacement of a large number of 59 workers in manufacturing industries and the low-paying jobs in the service sector has resulted in an increase in self-employment, unpaid work, industrial home working, and a variety of informal work On one hand, families and households have increased the number of earners in order to cope with the loss of employment and underemployment of one of their members. On the other hand, families and households started or increased their involvement in self-employment activities. For families and households experiencing economic hardships, self-employment is a source for extra income. For other families and households, self-employment is new window for opportunity because the new technology facilities do not require a lot of capital, time flexibility, and the need for work autonomy. Thus, it is hypothesized that: H221a. The greater the prevalence of core industries in a labor market area the lower the likelihood of self-employment. However, H222a Differences in self-employment between non-metropolitan and metropolitan areas are primarily due to their differences in the proportion of core industries. H221b. The greater the prevalence of core industries in a labor market area the greater the likelihood of additional earners. Labor market areas differ not only in industrial structure but also in the levels social capital. I propose that: P23. Residential stability of a labor market area affects work strategies. At the labor market area (or community level), the longer the length of residence (an indication of social ties to the community) the greater the strength of enforceable trust and the higher levels of social capital stemming from it (McLanahan and Sanderfur 1994). Residential stability is a reasonably good proxy of social capital because it 60 measures the long-standing connections between families in the community. Social ties and trust are an important source of social capital within a community. They not only enable families and households to access to a range of resources and services, including financial resources that may help start a small business, but also they protect businesses to go bankrupt in their early stages. They may also attract new investments (businesses) from outside the community. Thus, H23 la. Residential stability of a labor market area is positively related to self- employment. The net impacts of restructuring have been an increase in poverty, unemployment, underemployment, decline of earnings, and public assistance receipts. Labor market areas tint experience economic hardships, either because of the recent restructuring or historical deterioration the economic structure, not only do not have enough jobs for its residents, but also the likelihood of small businesses is minimum. Thus, I propose that: P24. Economic disadvantage/inequality of a labor market area affects work strategies. This leads to the following hypotheses: H24la. Economic disadvantage/inequality of a labor market area is negatively related to self-employment. H24lb. Economic disadvantage/inequality of a labor market area is negatively related to additional earners. A summary of propositions and hypotheses is presented in the following table. 61 Table 1. Summary of Propositions and Hypotheses. 62 Propositions Hypotheses Self-employment Additional Earners Level-l: Family/household P11. Family and H1 1 la. Married families H1 1 lb. Married families household structure are more likely than “non- are less likely than “non- determines work married” families and married” families and strategies. households to have a self- households to have a self- employed member. employed member. H112a. Cohabiting families H112b. Cohabiting families and households are more and households are less likely than single-headed likely than single-headed families and households to families and households to have a self-employed have a self-employed member. member. H1 13a. Female-headed H1 13b. Female-headed families and households are families and households are less likely than male- less likely than male- headed families and headed families and households to have a self- households to have a self- employed member. employed member. H1 14a. Families and households with children under 18 years of age are more likely than those without children to have a self-employed member. H115a. Families and households with preschool children are more likely than those with school-age and/or adolescent children to have a self-employed member. Propositions Hypotheses Self-employment Additional Earners H116a. Families and households with school-age children are less likely than those with adolescent children to have a self- employed member. P12. Family and H121a. Family/household H121b. Family/household household social ties to residential mobility residential mobility the labor market area decreases the likelihood of decreases the likelihood of affect work strategies. self-employment. additional earners. Level 2: Labor market area P21. Non-metropolitan- H211a Non-metropolitan H21 lb. Non-metropolitan metropolitan residential families and households are families and households are location determines work more likely than those in less likely than those in strategies. metropolitan areas to have a metropolitan areas to have self-employed member. additional earners. P22. Industrial structure H221a The greater the H221b. The greater the of a labor market area prevalence of core prevalence of core affects work strategies. industries in a labor market industries in a labor market area the lower the area the greater the likelihood of self- likelihood of additional employment. earners. H222a. Differences in self- employment between non- metropolitan and metropolitan areas are primarily due to their difl’erences in the proportion of core industries. P23. Residential stability H231a. Residential stability of a labor market area of a labor market area is affects work strategies. positively related to self- employment. 63 mp... if f 4 i W V ' ” Hypoth 9‘ V * Self-employment Additional Earners P24. Economic I-I241a Economic H241b. Economic disadvantage linequality disadvantage/inequality of a disadvantage/inequality of a of a labor market area labor market area is labor market area is affects work strategies. negatively related to self- negatively related to employment. additional earners. Note: Propositions and hypotheses are numbered using letters and mrrnbers. The letter P refers to a proposition while the letter H stands for a hypothesis. The first digit following the letter, indicates the level of analysis, 1 for a proposition and hypotheses to be tested at level-1 and 2 if a proposition and hypotheses are to be tested at level-2 of the analysis. The second digit refers to a proposition’s number. The third digit indicatesahypothesis’ number. Thehstletterindicatesthedependernvafiableforwhichthehypothesesare beingtested-«aforself-employmentandbforadditionaleamers. Chapter 3 THE RESEARCH DESIGN AND ANALYTICAL STRATEGIES Data This study uses two data sets: The 1990 Public Use Micro Data Samples (PUMS- L) Labor Market Areas and the 1990 Summary Tape Files (STF3). The PUMS-L is produced by US. Bureau of the Census and funded by the Economic Research Service, US. Department of Agriculture and Agricultural Experiment Stations and land-grant institutions affiliated with U. S. D. A. Regional Project 8-259. Labor market areas (LMAs) are groups of counties that encompass the county of residence and the county of work The labor market area PUMS-L data provides an excellent sample for linking individuals, households, and labor markets areas, thus permitting the incorporation of multilevel factors in models of household and family work strategies. The other advantage of using the PUMS data for labor market areas is their coverage of both rural and urbm local labor markets. For this study a sub-sample of 112 North Central labor market areas is first selected and groups counties mainly in the states of Michigan, Ohio, Indiana, Illinois, Wisconsin, Minnesota, Iowa, and Missouri. The boundaries of labor market areas are not limited to geographic lines of these states. Some counties are from the neighboring states including Kentucky, West Virginia, Arkansas, North and South Dakota, and Nebraska. See Map of Labor Market Areas. 65 (hOvza Shaw 82¢, .21.: .33 s as: ._ 23... 66 The major limitation of these labor market areas is the Census Bureau confidentiality requirement for any geographic identifier on the PUMS to contain a minimum of 100,000 people (See Tolbert and Killian, 1987, for greater details). A sample of civilian working-age individuals (age between 16 and 64) is selected from the PUMS-L data Four types of households are selected. They include married-couple families, male and female-headed households, and cohabiting households (unmarried persons living with their partners). Households of one member are excluded in this study. This corresponds to 75,280 households with about 36 percent of them (27,013 households) living in non-metropolitan areas. The summary tape files data from the 1990 census (STF3A) is used to construct characteristics of labor market areas. Data from the summary tape files is at the county level and is aggregated at flre corresponding labor market areas. Characteristics of labor market areas include the opportunity structure, i.e., the availability and types of jobs, residential stability and labor market area inequality. Was The variable of interest, family and household work strategies, is derived from a combination of the employment and self-employment activities of family/household members. Using the employment status of every member in the household (employed, unemployed, or not working) and the class of worker (employed in private or in government sectors, self-employed including unpaid family members), a typology of family and household work strategies is developed and include six excluded categories including: 1) families and households in which none of the members is working 67 (employed, self-employed or unpaid family member), 2) Only one spouse/partner is employed, 3) Both spouses/partners are the only ones employed, 4) At least one or more additional earners is employed, 5) At least one spouse/partner is self-employed, and 6) a combination of additional earners and self-employment For simplicity and clarity, this study focuses on two separate dependent variables: (1) Employment of non-core members within families and households”, and (2) self-employment"). Employment of other members refers to employment of any person 16 years of age and older within the family/household other than the head of the household and the spouse or partner in case they are present. Previous studies have focused on the conventional one-earner employment strategy - the employment of the head of the household. Starting with the 705, with the increasing participation of women in the labor force, recent studies have shifted the focus to dual-eamer employment strategy, that is, the employment of the head of the household and that of spouses. This study analyzes ’ Nonmmanbmhchdeofiumanbasoffiehwsehoflofiafianfiehmmeholda,fiespoummflor partner (if present). Using the label “non-core” is not intended to diminish the importance or attach a tertiary status on the employment of those members, rather it is intended to highlight and emphasize an emerging employment strategy of families and households other than the “traditional” one-earna household and the increasing dual-earner households. ‘° A distinction is being made between three major categories of self-employment: (l) Self-employment in own unincorporated business (Census definition), (2) self-employment in own incorporated business (considered as paid employees for tax purposes), and (3) unpaid family workers (work on a fimily firm or family business for at least 15 hours per week). There are dificulties in distinguishing self-employed to employees since some corporations may register employees as self-employed to avoid taxes and social secmity payments (Linder and Houghton 1990; Myles and Turegun 1994). The self-employed indicator is derived from the class of worker variable. Self-employed is defined as CLASS=6, self-employed in own not incorporated business, professional practice or farm, CLASS=7, self-employed in own incorporated business, professional practice or farm, and CLASS=8, working without pay in family business or firm. Not self—employed includes individuals employed in the private, Federal, or State sectors --CLASS=1,2,3,4,and 5. Individuals less than 16 years old, unemployed who never worked, and those not in labor force (last worked prior to 1985), and the armed forces. Individuals who are classified as not self-employed but have received a self-employment income (INCOME-12 non-farm self-employment income or INCOME3 farm self- employment income) are included in the category of self-employed. 68 the employment of other members of families and households that are employed as a family and household employment strategy. They include adult children living in the household, brothers and sisters, parents, uncles and aunts, other relatives as well as non- relative members who live in the household Conceptualized at the household level, employment data for all employed members of the household is used to create a variable that identifies various household employment strategies. This variable identifies three groups of families and households based on their employment status in 1990: (1) Families and households with none of the members is employed, (2) families and households with core members employed and no other members employed, and families and households in which non-core members are employed. The employment of non-core members within families and households is transformed into a dummy variable indicating (1) families and households with other members employed, and (0) families and households with no other members employed Later in the analysis, I refer to the employment of non-core members as additional earners. Two main questions are of interest in regards to self-employment: (1) what is the prevalence of self-employment and (2) what are the factors contributing to household involvement in self-employment? Previous studies have focus on the analysis of self- employment at the individual and national level, especially in metropolitan areas and among immigrants (Portes and Zhou 1996; Sanders and Nee 1996), but few have analyzed self-employment at the family/household level and in rural settings. This study examines self-employment at the family/household level in both metropolitan and non- metropolitan areas. Self-employment measure is conceptualized as a dichotomous 69 variable, indicating (1) Whether at least one or more members of the household are involved in self-employment activities? (0) Otherwise. The key independent variables are measured at both family/household and labor market area levels of analysis. I argue that both family and household characteristics and labor market area characteristics affect the likelihood of families and households to have additional members in the labor force or to get involved in self-employment activities. At the family/household level, eight indicators are constructed to reflect differences on family and household structure and composition, life-course stages, mobility status, social class, race and ethnicity, and immigrant status. These variables are assumed to determine the diverse family/household social contexts that shape family/household work strategies. Along this study, the family/household has been used interchangeably because they are overlapping, reciprocal, and interdependent (Bokemeier 1997 ). Family/household structure is defined by family household type", as defined by the US. Census”. Past research has tended to focus on married-couple households or on “ Family/household type is defined as a transformed and combined variable ofRHHFAMTP and RELATl. RHHFAMTP includes married-couple finrily household, male and female-headed fimilies, and non-family households -- male householder living alone, male householder not living alone, female householder living alone, and female householder not living alone. RELATl indicates the relationship to the householder and includes the householder, spouse, related (son/daughter, brother/sister etc. .) and unrelated individuals (roommate, partner etc. .). Married-couple family is defined as RHHFAMTP=L Male-headed households aredefinedasRHHFAMTP=20rRI-IHFAMTP=12andthereisnounmaniedpartnerlivinginthe household, RELAT1=10. Female-headed households are defined as RHHFAMTP=3 or WWZZ and there is no unmarried partner living in the household, RELATl=lO. Cohabiting households are defined as being headed by an unmarried male or female , RHHFAMI'P=2,3,12,22who coreside with unmarried partner, RELAT1=10. Excluded in the analysis are households of living alone individuals, RHHFAMTP=11,21, as well as group quarter/vacant households. 12 TheU.S.BureauoftheCensusdisfingrfishesbetweenhouseholdsandfimifies. Ahouseholdisdefinedas a statistical or physical unit of coresidentiality, composed with one or more individuals who occupy the same livingquarters—ahouse—anapartmentoraroom. Mernhersofhouseholdsmayberelatedinmostofthecases, unrelatedorlivingalone. Afimflyisdefinedastwoormorepersonsrdatedbybloodmarriage,oradoption 70 female-headed households but few have analyzed household work strategies in male- headed families and in cohabiting households. Family/household type is coded into four categories: (i) married-couple families, (ii) male-headed families, (iii) female-headed families, and (iv) cohabiting households. Single households are excluded in the analysis. The presence and age of the youngest child under 18 present in the household are use to construct five life course stages: (i) no children — there are no children under 18 living at home, (ii) preschool -- the youngest child is under six years of age, (iii) school- age -- the youngest child is between 6 and 12 years of age, and (iv) adolescent --the youngest child is between 13 and 17. The family/household type indicates whether or not there is a spouse or a partner in the household The life course stage adds whether or not there are children under 18 years of age in the household. The next variable measures the number of adult relatives in the household Adult relatives include not only adult children but also other relatives including brothers, sisters, uncles, and aunts and other relatives. Residential mobility indicates whether or not at least one or all members of families and households have moved in the last five years (period of 1985 to 1990). Residential mobility is used as a proxy for families and households’ integration into wholivetogetheratleastpartofthetime, whoshareeconomicresources, andwhofunctionasacooperative social unity in several ways. Non-family households include persons who live alone or with unrelated individuals. Non-family households are a diverse group. They may consist of elderly individuals who live alone, college-age youths who share an apartment, cohabiting couples, individuals who can delay or forego marriage, or those who are between marriages. Over the past three decades, the number of non-family households has increased dramatically because of the aging of the population, high divorce rates, and the tendency of young adults to live apart fi'om their parents before marriage. Family ties extending outside the irrrmediate household are not considered in the above definition of the family. In this definition, cohabiting households are defined as non-fimily households. In this study, cohabiting households include non—family as well as single families living with their partners. 71 commrmities. It is assumed that families and households that have experienced geographic mobility in the last five years are relatively less connected to their community, and thus, less social capital stemming from it, when compared to those that have not moved in the last five years. Residential mobility is measured as (1) if at least one or all members of the family/household have moved in the last five years, and (0) otherwise. Indicators of the economic capital include household income and poverty status. It is assumed that family and household work strategies vary greatly depending upon the family/household economic capital. Household income refers to all sources of income of all members in the household. Family/household poverty is measured using the 1989 poverty threshold”. Poverty is a dummy variable scored 1 for families and household in poverty, that is, whose household income is less than 1.25 the poverty threshold and 0 for those who are not poor. The other household indicator tlmt reflects both the social and economic status is home ownership. Home ownership is a dummy variable scored 1 for respondents who own their own home and O for those who do not. A The other explanatory variables measure education, race and ethnicity, age, and immigrant status of core members (householder, spouse/partner (if present». Education differentiates families and households into four categories according to the level of education of core members: (i) At least one or both (if spouse/partner present) core ‘3 Poverty thresholds are defined and revised each year by the omce of Management and Budget based on familysize, thenumberofrelatedchildrenunder l8yearsandincorne. The l989povertythresholdsareused in the computation of household economic well-being. For a detailed discussion on poverty measure, see U. 8. Census, Current Population Reports, Series P-60, No 171, Poverty in the United States: 1988 and 1989. 72 members has less than high school (excluded category), (ii) At least one or both has high school education; (iii) At least one or both have some college, and (iv) Both core members have college (bachelor) or higher. Race includes six categories. (1) Non- Hispanic Whites; (2) Non-Hispanic Blacks, (3) American Indians, (4) Hispanics or Latinos, (5) Asian Americans, and (6) Mixed-race households. Later in regression analysis, race is coded into two groups: (1) Minority, and 0 Non-Hispanic White (excluded category). Immigrant status is measured as (1) if at least one member or all is an immigrant member and (0) Otherwise. Age is measured in years and refers to the mean age of householder and spouse/partner or the age of the householder if no spouse/partner is present At the labor market area level, indicators were constructed from STF3A data to indicate the spatial location of families and households as well as labor market area differences in opportunity structure, i.e., the availability and types of jobs, residential stability, and levels of inequality/economic disadvantage. Labor market areas are groups of counties that encompass both the location of residence and employment. Labor market areas are not only the geographical space where people reside and can find jobs, but also they include social, economic, and cultural opportunity structures that influence potential household work strategies. A labor market area’s conceptualization allows the analyses of how the opportunity structures of an area, in combination with other aspects of social life, affect the likelihood of families and households to adopt different adaptive strategies. The underlying assumption in using a labor market area is that location in a social space 73 affects economic opportunity and life chances of its residents and provides the parameters for family/household strategies. The spatial distribution of economic activities and other opportunities conditions the repertoire of potential household work strategies. Recently, sociologists of labor market analysis have begun to specify how employment relationships within the labor market interface with other aspects of the social organization of society and how these social relationships affect labor market behavior (Snipp & Bloomquist 1989). Past studies have tended to focus on labor markets by analyzing characteristics of workers, occupations, organization of industries, firms, occupations, and class (Kalleberg and Berg 1987). Past studies on spatial variations in social and opportunity structure have considered space as a contextual given, emphasizing its physical view. Most studies that attempt to analyze the spatial location effect on individuals and families use a metropolitan/non-metropolitan dummy variable. The metropolitan-non—metropolitan dichotomy collapses the rich and dynamic aspects of the non-metropolitan and metropolitan differences in social, historical and economic opportunities. Soja (1989:79- 80) indicates that space may be primordial given, but the organization and meaning of space are a product of social translation, transformation, and experience. Industrial and occupational structure, along with social and demographic characteristics such as population size, education, housing facilities, race and gender relations, unions of workers, social services, churches and other social organizations interact differently within each labor market area to create an environment of alternative household work strategies. 74 Non-metropolitan and Metropolitan residence is defined using non- metropolitan/metropolitan definitions“ and the size of the largest city, town, or place. It is coded as (1) Non-metropolitan small -- population of largest place is between 5,000 and less than 20,000 in 1990, (2) Non-metropolitan large—population of largest place in 1990 was at least 20,000; (3).metropolitan small -— population of the largest MSA was less than 250,000 in 1990, (4)metropolitan medium—population of the largest MSA was at least 250,000 but less than 1,000,000, and (5)metropolitan large—population of the largest MSA in 1990 was 1,000,000 or greater. In a multivariate analysis, non- metropolitan/metr0politan residence is collapsed into two effect categories: (1) Non- metropolitan, and (-1) Metropolitan. Industrial structure is defined by the presence and type of industries present in a labor market area Industries are grouped into five major categories based on expected earnings and using the three-digit standard industrial classification code: 1) Agriculture, Forestry, and fishing; 2) traditional high-wage industries (mining, government, and high- wage manufacturing); 3) construction and low-wage manufacturing; 4) high-wage services; 5) and consumer services". In order to assess the relative impact of the 14 Non-metropolitan labor market areas (LMAs) are those containing no MSAs -Metropolitan Statistical Areas and metropolitan LMAs are those containing one or more MSAs or PMSAs ~Primary Metropolitan Statistical Areas (Office of Management and Budget Bulletin 93-17, ERS Stafi‘ paper, Rural Economic inision, 9614. Low-wage mamdircturingincludes tobacco, textilemill products, apparel, lumberproducts and furniture, rubber products, leather products, and miscellaneous manufacturing industries. All the other manufacturing industries not classified as low-wage industries are grouped into high-wage manufacturing industries. Consumer services include retail trade, entertainment and recreation services, automotive services and repair, and personal services. High-wage services include services not included in consumer service sector and those include business, professional services, finance, insurance, and real estate services, and transportation, communication, and utilities (Kassab 1992 and Kassab et al.1995). 75 presence of high-quality jobs versus low-quality jobs in an area, one aspect of the restructuring, a ratio of core industries (traditional high-wage industries and high-wage services) to peripheral industries (agriculture, forestry, and fisheries, construction and low-wage manufacturing, and consumer services) is computed A ratio of 1 indicates an equal balance of core and periphery industries. A ratio greater than 1 indicates a greater prevalence of core industries over periphery industries while a ratio less than 1 would indicate a greater prevalence of periphery industries over core industries in a labor market area. Industry structure reflects one important element of the opportunity structure, that is, the quality of jobs available in a labor market area It is assumed that the opportrmity structure of a labor market area enables and constrains family and household work strategies. In addition, the effect of the opportunity structure on families and household work strategies in non-metropolitan areas is hypothesized to differ from that of metropolitan areas. Therefore, an interaction variable is included to test whether the opportunity structure has different effects in non-metropolitan and metropolitan areas. The other characteristics at the labor market area level included in the study are the percentage of residents that have not moved in the last five years; the percentage of housing owners; social and economic well-being measures including a gini" measure of income inequality, the percentage of households in poverty, the percentage of households 011 public assistance, the percentage of people unemployed, and the per-capita income; ..‘ l6 TheGinicoeficienttakesvaluesbetweenOand1,with0representingcompleteequalityofincomesand 1 intlicatingcompleteinequality. 76 educational attainment (percentage of residents over 25 years of age with a college, some college, high school, and less than high school education); race and ethnicity; immigration concentration; and age structure (percentage of residents less than 18 years and percentage of residents greater than 65 years of age). Table 2 displays the descriptive statistics of variables used in the analysis. Tablez. Descriptive Statistics of Selected Characteristics" Variable Codes and range Mean Std. Dev. Family and Household Characteristies (N='75280) F (wily/household work strategies Self-employment l = Self-employed . 1 6 7 . 3 7 2 0 = Otherwise Non core members’ 1 = Non-core members employed . 1 4 4 . 3 5 2 employment 0 = Otherwise F amin/household structure (contrast coding) Married-couple Families .750 = Married-couple families - 8 l 0 . 3 93 -.250 = Non-married households Cohabiting households .667 = Cohabiting households - 0 3 9 . l 9 5 -.333 = Married households .000 = Single households Female-headed .500 = Female-headed households .122 .327 households -.500 = Male-headed households .000 = Married/Cohabiting households l7 Fordaaipfiwwmomfimd‘ormdwfibluwnhwmmthmmcomedudumyvmables (I, 0) for computation ofmeans, sothat theycan reflect proportions. 77 Variable Codes and range Mean Std. Dev. Life course stages (contrast coding) Presence of children .250 = Children < 18 years - 5 5 0 . 4 98 under 18 years of age -.750 = No children < 18 years Preschool children .667 = Preschool children . 2 4 5 . 4 3 O -.333 = School-age and/or adolescent children .000 = No children < 18 years School-age children .500 = School-age children - l 9 0 . 3 92 -.500 = Adolescent children .000 = No children under 18 years of age/Preschool Number of adult relatives 1 = One or more adult relatives - l 8 7 . 3 9 0 in the household 0 = No adult relative Household Mobility l= Moved in the last 5 years . 4 6 6 . 4 9 9 0= Not Moved (same house) Household Income (0, $455,359) 3 8 , 1 92 2 8 , 3 9 1 Home Ownership l=Ownahome .788 . 422 0==No home (rent) Family/household Education (excluded category is less than high school) High School (0,1) .376 .484 Some College (0, l) . 2 91 . 4 5 4 College education (0,1) . l 9 4 . 3 9 5 Race and Ethnicity (excluded category is non-Hispanic White) AfiicanAmericans (0,1) .039 .195 Native Americans (0,1) . 0 O 4 . 0 65 Hispanics (0,1) .002 .043 AsianAmericans (0,1) .013 .113 Mixed Households (0,1) . 0 O 6 . 0 7 5 ImmigrantStatus 1=Atleastoneorallmembers .018 .132 are rmmr' ' grants 0 = No member is immigrant Age (16,64) 40.93 11.71 Age square (00's) (.26, 4.10) 1. 812 1 . 000 78 ll Variable Codes and range Mean Std. Labor Market Area Characteristies (N=112) Dev. Non-metropolitan/Metropolitan Residence (excluded category is metro large) Non-metro small urban (0,1) . 2 4 1 . 4 3 O Non-metro-large urban (0,1) . 1 9 6 . 3 9 9 Metro small metro (0,1) . 2 9 5 . 4 7 6 Metro medium metro (0,1) . 1 7 O . 3 7 7 Industrial Structure (excluded category is consumer service industries) % Agriculture, Forestry, (.9, 23.5) 6 . 2 96 4 . 604 and Fishing Industries % High-wage industries (10.8, 32.9) 2 O . 7 3 O 4 . 555 % Construction/ (5.6, 19.3) 1 O . 8 1 4 2 . 63 8 low-wage industries % High-wage services (28.2, 49.9) 3 8 . 05 9 4 . 4 1 7 Ratio ofcore to periphery (.41,1.11) .744 .162 Industries Socioeconomic Indicators Income inequality (Gini) (.369, .462) . 4 O 6 . O 1 8 % ofhouseholds in (6.15, 25.65) 12 . 0 1 7 3 . 4 5 5 poverty % ofhouseholds on public (3.73, 17.38) 7 . 2 9 1 2 . 2 64 assistance % of people unemployed (2.85, 11.52) 6 . 4 2 5 1 . 8 91 Per capita income (8,476, 17,194) 1 1 , 3 7 O 1 , 4 4 2 Demographic Characteristics %ofpeople withahigh (26.59, 46.19) 37.39 4.115 school (over 25 years) 79 Variable Codes and range Mean Std. % of people with some (15.58, 32.00) 22 . 7 91 31.);‘66 college (over 25 years) % of people with college (7.59, 27.80) 1 4 . 8 1 3 4 . 2 2 3 (over 25 years) %Black (.04,21.44) 4.116 4.862 %Hispanics (.21,11.15) 1.251 1.417 %Asian Americans (.16, 3.34) . 705 . 545 % Native Americans (.05, 8.60) . 67 4 1 . 1 61 %immigrant (.13, 6.73) .782 .769 % Female-headed (4.35, 14.65) 8 . 93 1 . 993 households %ofpeoplelessthan l8 (2469,3138) 27.635 1.433 Three types of analyses are planned A descriptive analysis on the prevalence of families and households with other members other than the householder and spouse/partner (if present) employed and self-employment will be first presented. Second, an exploratory factor analysis of labor market area characteristics will follow. Finally, a multi-level model for discrete data to account for the combined effects of labor market area and family/household factors on these dependent variables is envisaged In this study, I first examine the relationship between household social structures and the likelihood of families to have other members employed I expect to find that the odds of other members to be employed (additional earners) vary depending upon family 80 household social structures. Are the odds of having other members employed different for married couples, male-headed households, female-headed households, male-headed cohabiting, and female-headed cohabiting households? If so, what other family and household characteristics that may directly or indirectly influence the likelihood of having other members employed versus that of not having other members employed? Second, I examine the relationship of family/household social structures and the odds of families and households’ involvement in self-employment. Do the odds of self- employment vary depending upon family/household social structures and other family and household characteristics? I argue that family and household social, economic, and cultural capital influence the odds of having other members employed or that of self- employment. Assuming a constant variance at the labor market area in the odds of self- employment and in the odds of family and households to have other members employed, a logistic regression model would fit well the data However, the odds of self- employment or of having additional earners are assumed to vary depending upon the social structures of places in which families and households are located and embedded. Thus, I expect to find that the odds of self-employment and the odds of having additional eamersvary from labormarketareato labormarketarea. Ialso expectthatthe oddsof self-employment for families and households living in non—metropolitan areas and those of having other members employed are different from those of families and households in metropolitan areas because of their historical and continuous differences in opportunity structures. Thus, multilevel models that take into consideration different 81 levels of analysis, specifically different sources of variations in the odds of self- employment or having additional earners are required The next section describes multilevel models and their advantages over standard regressions and conceptually explains how these models would fit this study. Midfikysl Medals Multilevel models refer to a family of models that help analyze hierarchically structured data. In sociology, multilevel models are used to explore the link between the macro and micro levels of social phenomena (see Diprete and Forristal 1994 for a review). There are a variety of multilevel statistical models that are designed to integrate and test multilevel theories using multilevel data Previous attempts to model and analyze hierarchically structured data have been constrained by conventional statistical techniques that do not allow for analysis of multilevel models. Recent research on multi- level models by Bryk and Raudenbush (1992), Goldstein (1995), and Longford (1993) contributes to a large number of research in social sciences and most importantly find the missing puzzle of studies that attempt to integrate multilevel theories to multilevel analytical strategies. This study applies multilevel models for discrete response data to analyze the effects of both family and household factors as well as the effects of labor market areas’ factors on the odds of families and households to have other members employed or the odds of getting involved in self-employment activities. As a general illustration, the next section describes a basic 2-level linear multilevel model. Two levels of analyses are considered The household is the level 1 unit, and the labor market area in which the family/household locates and its members find work is 82 considered the level-2 unit of analysis. Following Bryk and Raudenbush (1992), let Y,- denotes the response variable for household i nested within a labor market area j. Yi=50r +911“ Xi+eij Where j refers to labor market area imit and i refers to household level; [50,. is the intercept and [3,,- is the regression coefl'icient indicating how the outcome Yij varies in a labor market j as a function of the household characteristic Xi; and e,- the residual or error term. It is assumed that e, ~N(0,o’). E(e,j)=0, and Var*(eijFo2 The above equation is the linear regression equation for each labor market area j. For j level-2 units, there would be j number of equations. In case there are few level 2 units and the goal of the study is to focus on specific level 2 units, the equation above can be analyzed. However, in case there are many j level-2 units, analyzing the equation above would require to estimate a large number of parameters and may result in imprecision estimates given the potential variation in level-2 units (Goldstein 1995; Bryk and Raudenbush (1992). In order to have a multilevel model, 00,- , 8.,- are assumed to vary randomly: BOJ‘ :50 + "Oi ; Bu =51 + 1115 ; where no, and un- are random variables and E(uo,-) = Em.) =0 Var (P05) =03 . Var (mt) =0..’ . C°V(Po,' . al.-Fowl The combined model, then becomes Yijzpm‘ +ij Xij+(llq+llrjxij+eij) 83 u Lilla 7.1155 r 13.. rials The model is composed by two major components, a fixed part, [30,- + [5,,- X,-, and a random part (qu + 11,in + e,- ). It is this random part that distinguishes the 2-level model to the standard linear regression For estimation of fixed and random parameters and the introduction of more fixed and random parameters at level 1 as well as parameters at level 2, see Bryk and Raudenbush (1992), Goldstein (1995), and Longford (1993). For a review of software for multilevel models see a review by Krefi et al. (1994) and Goldstein (1995, chapter 11. Given the dichotomous nature of the outcome variables in this study, a multilevel model for discrete response data is described next. In the following description of multilevel model, two levels of analysis are considered. Level-1 units are families and households nested within level 2 units, labor market areas. Multileysl Medal fut cm W Data In this study, having additional earners and self-employment are defined at the family/household level as binary response variables. Having additional variable takes a value of 1 if at least one or more non-core members are employed, and 0 otherwise. Self-employment variable also takes a value of 1 if at least one or more members within a family/household are self-employed, and 0 if there is no self-employed member in the family/household. Next, I will describe a multilevel for discrete response data using self- employment dependent variable. The model for additional earners is similar to that of self-employment because they both have two levels of analysis with family/household and labor market area characteristics as predictors. For each family/household i and labor market j, and a set of explanatory variables Xi, the logit link fimction for the probability that family/household i within labor market j have a self-employment member is: n.=f{X.B+ u.}={1+exp(-D<.a+ am" where it,- is the probability of being involved in self-employment for family/household i within labor market j, and f, a non-linear ftmction of the linear predictor X, B which has fixed coefficients, and ll,- is the random component for labor market j. It is assumed that u,- ~ N(0, 0,2). The response variable Y,i is assumed to have a binomial distribution: Y,- ~ Bin (1:,- , nil.) Where n,- is the number of “trials” and it,- is the probability of “success” in probability terms. In the case of this study, n,,- =1 which refers to the Bemouilli distribution. Var(Yij Ire,- )= 1:,- (14:9) The model thus becomes, Y,,-=1rij + e,- Z,- Where Z,- =\/rr,,- (Mtg); E(e,,- =0; and Var (e,- )=1. For full description of multilevel with binary responses see Goldstein ( 1991,1995) and Rodriguez and Goldman (1995). Figure 2 shows a multilevel model of self-employment while figure 3 shows a multilevel discrete model of additional earners. Each model combines the level-1 and level-2 predictors. In both models, level-1 predictors are centered to their grand mean, except household income which is centered on the median income. This allows to model the intercept, 0,3, the labor market area’s mean log-odds of self-employment (or 85 additional earners), adjusting for the covariates at level-l (family/household). Then, the adjusted mean for each labor market area j is modeled as a function of level-2 variables plus a random error. Figure 2: Multilevel Discrete Model for Self-employment selferrrpy.~ Binomial(dcnorrr, 4}) } selfenpij = 74,-” ewbcons. 108%) firms + flimettj + 7830M}; + Mm”); + flsChil‘l; + fitmschy + firmer; + flamma- + figmobilityy. + [310qu + 511mg- + fiuhighsclr}. + i9 13somecolli1. + [3 1490mm; +fi15min0fityij + [916385;- + [31738934] + flame“; + flutes“) + 13min“; + 525%“; +flnm°findj 511:5: “‘1; [“11] "N“ 9" i Q": [03.] boons’=bconsrt,(1-t,)/denont,r“ [904'] ~ (0» Q.) i 9.11] 86 Figure 3: Multilevel Discrete Model for Additional Earners naddearr). ~ Binomial(denorrr.1, 7%.) } naddeanfi = ,5]. + co, cons. logit(7;.j) = fl 1Icons + fizmriedij + 53cohab’}. + fifenflr}. + fiSChlldy. + fiépresclr}. + fi7schageq + figmobilityl}. + figuring). + filohighsclri + flusomecollij + flucollegeij + filminorityij + [314388; + 161538381} + 516mm; + fl17°°°dis% + fl 18”]er l9 1; = 51 + "1; [at] “N”: ‘4 81°11] bcons.=bcons[;;.)(l -,;}.)idenom,}]°5 [it] "' ‘°’ 9') ‘ ‘2': [I] 87 Fem! nit-ill , Enos Chapter 4 RESULTS E . .v S i . Figure 4 shows how many individuals within families and households are working. About 31 percent of families and households have one earner and 47 percent of all families and households are dual-earners. About 9 percent of families and households have three earners. About 3 percent of families and households have four or more earners. The remaining 10 percent consists of families and households in which no member is employed. In addition to the number of earners within the household, this study is interested in knowing which family/household member is working and in which activity. F amily/household members include the householder, the spouse or partner (if present), adult relatives (child, brother and sisters, parents, uncles and aunts etc), and non relatives. The study distinguishes whether a family/household member is employed, self- employed, or not working. The combination of family/household composition and work status yields what is referred as family and household work strategies. The distribution of family and household work strategies is presented in Figure 5. About 61 percent of families and households have only a spouse or partner (core members) employed In this category, about half of these families and households have one spouse or partner employed The other half includes families and households in 88 which both spouses/partners are employed In about 15 percent of families and households, there is at least one additional earner other than the householder, spouse, or partner. This is the case when adult children, relatives, and non-relatives living in the household are employed. About 12 percent of families and households are classified as self-employed, i.e., at least one member is self-employed This category of households includes cases in which one spouse/partner is employed and the other is self-employed, both spouses are self-employed, one spouse/partner is self—employed and the other is not employed, or additional members are self-employed. In about 2 percent of cases, Figure 4. Number of Earners Within Families and Households 3% 9% 10% B No Earner IOne Earner I Two Eamers a Three Famers El Four Earners and mre 47% families and households combine employment of non-core members and self- employment activities. About 11 percent of families and households have neither member employed nor self-employed. Figure 5 shows that about 17 percent of families 89 and households have additional earners and about 14 percent of families and households are self-employed Table 3 displays the characteristics of families and households by non- metropolitan and metropolitan residence. F arnilies and households in non-metropolitan areas are more likely than their metropolitan counterparts to have a self-employed member. About 20 percent of families and households in non-metropolitan areas are involved in self-employment work while about 15 percent of those living in metropolitan areas are. Families and households living in metropolitan areas are more likely than those in non-metropolitan areas to have additional earners. About 16 percent of Figure 5. Percentage Distribution of Family and Household Work Strategies 2 % INone ofthe Members is l l % Working IOnly One Spouse/Partner is employed IBoth Spouses/Partners are the Only Ones Employed 3 0 % EIAt Least one or more Additional Earners B At Ieast one spouse/partner is self- employed ElA Combination of Additional Eamers and Self-employment 90 metrOpolitan families and households have additional earners while only 12 percent of non-metropolitan families and households do. In the sample, about 36 percent of households live in non-metropolitan labor market areas and about 64 percent live in metropolitan areas. About 81 percent of all households are married couples and about 12 percent are female-headed households. Approximately 3 percent and 4 percent of the sample are respectively male-headed and cohabiting households. Married households are more likely to be found in non- metropolitan than in metro labor market areas. The rest of the households, especially female-headed households, are more likely to live in metropolitan areas. About 55 percent of the households have children under 18 years of age living at home. Table 3. Family and Household Characteristics by Non-metropolitan and Metropolitan Residence Family and Household Total Metro Non metro Characteristies Percentages Self-employment." NotWorking 11.4 10.4 13.3 Not Self-employed 72 . O 7 5 . 1 66 . 3 Self-employed 16 . 6 14 . 5 20 . 4 Additional Earners'" NotWorking 11.4 10.4 13.3 NoAdditionalEarners 74.1 73.7 75.0 AdditionalEamers 14.4 16.0 11.7 91 Family and Household Total Metro Non metro Characteristics Family and household structure'" Marriedcouples 80.9 79.8 82.9 Male-headed households 2 . 9 3 . 1 2 . 6 Female-headed households 1 2 . 2 1 3 . O 1 0 . 7 Cohabiting households 3 . 9 4 . O 3 . 8 Life course stages' Nochildrenunderl8years 45.0 45.2 44.7 Preschoolchildren 24.5 24.5 24.5 School-age children 19. O 18 . 7 19. 5 Adolescent Children 1 1 . 5 1 1 . 7 11 . 3 Presence of Adult Relatives'" Noadultrelative 81.3 80.0 83.6 Oneormoreadultrelatives 18.7 20.0 16.4 Residential mobility". Movers 46.6 47.4 45.1 Non-movers 53.4 52.6 54.9 Immigrant status'“ At least one member or all 1 . 8 2 . 2 . 9 is immigrant Nomemberisimmigrant 98.2 97.8 99.1 Own a home.” Yes 7 6 . 8 76. 3 77.8 No 23 .2 23. 7 22 .2 Household Income'" Lessthan$10,000 8.9 7.7 11.0 Family and Household Total Metro Non metro Characteristics $10,000-24,999 24 . 3 21 . 1 30 . 1 $25,000-34,999 1 9 . 7 1 8 . 8 2 1 . 5 $35,000-49,999 23.3 24.2 21.6 350,000-74,999 16.4 19.1 11.6 $75,000-99,999 4 . 3 5 . 3 2 . 6 $100,0000rgreater 3.0 3.8 1.7 Household Poverty.” Poor 13.6 11.7 16.9 NotPoor 86.4 88.3 83.1 Core education". Lessthanhighschool 13.9 13.2 15.1 Highschool 37.6 35.7 41.0 Somecollege 29.1 29.4 28.6 Collegeormore 19.4 21.7 15.3 Corerace'” Non-Hispanic White 93.6 91.7 96.9 Blacks 4.0 5.5 1.2 AsianAmericans 1.3 1.7 .6 Native Americans . 4 . 3 . 6 Hispanics/Latinos .2 .2 .1 Other Races, including Mixed . l . 6 . 6 Core average age'” 1468311131130 19.2 19.2 19.3 30-39 31.0 31.4 30.3 40-49 23.9 24.5 22.8 93 Family and Household Total Metro Non metro Characteristics 50-59 17.7 17.2 18.5 600rmore 8.2 7.7 9.1 Number ofcaseg 75280 48267 27013 "‘p<.001; "p<.01; ‘p<.05; mot significant About a quarter of households report living with preschool children and about 19 percent of all households have an adult relative living in the household Overall, about 47 percent of all households have moved in the last five years. Movers include household members whose residence in 1985 was in a different house in 1990. This includes households that have moved within a labor market area and those whose residence in 1985 was in a different labor market area (either within the same or outside the state or abroad). Households in metropolitan labor market areas have experienced more mobility than their non-metropolitan counterparts. Respondents in more than three fourths of households report to own a home. Home ownership is slightly more prevalent in nonometmpolitan than in metropolitan areas. Respondents in about 78 percent of households in non-metropolitan areas compared to 76 percent in metropolitan areas own a home. The distribution of households in the sample according to the levels of income shows that about 9 percent of households have income less than 8 10,000, about 44 percent of households have income between $10,000 and $35,000, about 23 percent have 94 f 801 9U income between $35,000 and $50,000, about 16 percent of households have income between $50,000 and $75,000, and about 7 percent of households have income greater than $75,000. Note that household incomes are lower in non-metropolitan than metropolitan areas. Households with incomes less than $35,000 are more likely to be in non-metropolitan than in metropolitan areas. In contrast, households with incomes greater than $35,000 are likely to live in metropolitan than in non-metropolitan areas. Overall, about 14 percent of households in the sample are poor. As expected, a greater proportion of poor households live in non-metropolitan than in metropolitan areas. About 17 percent of all non-metropolitan households compared to only 12 percent of metropolitan households are poor. The distribution of households according to the level of education shows that households with a high school or less level of education are more represented in non- metropolitan than in metropolitan areas. In contrast, households with college levels of education live in metropolitan than in non-metropolitan areas. About 6 percent of households are non-White with about 4 percent Blacks. Non- White households, except Native Americans, live in metropolitan than in non- metropolitan areas. Considering the age structure, a greater proportion of households whose core members on average are over 50 years of age live in non-metropolitan than in metropolitan areas. In contrast, those in the age groups between 30 and 50 years are more likely to live in metropolitan than in non-metropolitan areas. 95 mmmcmama Table 4 displays differences between non-metropolitan and metropolitan labor market area characteristics. The selected households are located in 112 labor market areas, 49 of which are classified as non—metropolitan labor market areas. Table 4 classifies the 112 labor market areas according to industrial structure. The spatial location of industries is as expected, with a greater proportion of agriculture, forestry, and fishing as well as those in construction and low-wage industries in non-metropolitan labor market areas while producer service and mining, government, and high-wage manufacturing are more concentrated in metropolitan than in non-metropolitan areas. Compared to metropolitan labor market areas, non-metropolitan residents are more likely to be non movers. Also, compared to metropolitan, non-metropolitan labor market areas have on average greater percentage of owned occupied housing. Table 4 also classifies labor market areas according to different levels of socioeconomic well-being indicators. Compared to metropolitan labor market areas, the average percentage of households living in poverty, the average percentage of households receiving public assistance, and the average percentage of unemployed residents are higher in non-metropolitan labor market areas. Moreover, the average per capita income is lower in non-metropolitan labor market areas than in metropolitan ones. The average per capita income in non- metropolitan labor market areas is estimated at $10,501 while in metropolitan labor market areas, it is estimated at $12,046. 96 Table 4. Labor Market Area Characteristics by Non-metropolitan and Metropolitan Residence (N=112). Labor Market Area Characteristies Mean S .D. Mean S.D. (N=49) (N=63) Industrial Structure Agricultme, Forestry, and Fishing'" 7 . 92 5 . 74 5 . 03 3 .24 Mining/Government/ High-Wage 19.25 4 .29 21 . 88 4 . 45 Manufacturing” Construction/Low-Wage 11.71 3.15 10.11 1.91 Manufacturing” ProducerServices" 36.83 3.91 39.01 4.58 ConsumerServices‘“ 24.28 2.95 23.96 1.90 Residential Stability PercentNonmovers' 55.20 4.21 52.98 4.80 PercentOwners'” 73.28 4.13 69.83 4.07 Socioeconomic Well-being Income Inequality(Gini) .410 .020 .403 .016 %ofHouseholdsinPoverty'" 13.40 4.09 10.94 2.40 %ofHouseholds onPublic 7-89 2-76 6°82 -235 Assistance' PchapitaIncome°" $10,501 $944 $12,046 $1,404 % ofPeople Unemployed" 6 . 99 2 .25 5 . 99 1 . 42 Educational Attainment LessThanHighSchool'” 27.26 5.80 23.25 3.74 HighSchool" 38.04 3.54 36.89 4.47 Some College' 22.09 3.36 23.38 3.09 College”. 12.67 2.86 16.48 4.37 97 Non-r0 '5 if f if " Metro " Labor Market Area Characteristics Mean SJ). Mean SJ), (N=49) (N=63) Race/Ethnicity Concentration Non-Hispanic Whites'" 95.51 3.92 91.43 6.34 Blacks'” 2.30 3.64 5.52 5.24 Native Americans” .89 1.52 .50 .75 Asian Americans.” .45 .29 .90 .61 Hispanic (orI.atinos)” .81 .61 1.59 1.74 OtherRaces’” .03 .02 .05 .03 Immigration Immigrant” .46 .31 1.03 .92 Family Structure “‘ p<.001 ” p<.01 ‘p<.05 ns=not significant Note that there is greater variation in the per capita incomes across metropolitan labor market areas than in non-metropolitan ones. Thus, consistent with previous studies, non- metropolitan labor market areas are more economically deprived than metropolitan labor market areas. Table 4 also shows significant differences between non-metropolitan and metropolitan labor market areas in levels of education, race and ethnicity concentration, and family structure. Residents in non-metropolitan labor market areas have on average lower levels of education than their metropolitan counterparts. Non-metropolitan labor market areas have on average a greater percentage of residents with high school or less 98 education than metropolitan areas. In contrast, metropolitan labor market areas have on average greater percentage of residents with a college level of education. Blacks, Asian Americans and Hispanics are concentrated in metropolitan labor market areas than in non-metropolitan areas. On average, a greater percentage of female-headed households live in metropolitan than in non-metropolitan labor market areas. Next, a bivariate analysis indicating the relationship between family and household characteristics and their work strategies, is presented E39152: Analysis at 1m: Made! me smears: To assess whether a smaller number of linear combinations of characteristics describe the social context of labor market areas, an exploratory factor analysis was used. Three factors are obtained. The first factor describes the labor market area inequality or economic disadvantage. Variables that load high on the economic disadvantage factor include the percent of households in poverty, the percent of households receiving public assistance income, and the percent of people unemployed These three variables have loadings over .84. Based on these loadings, the labor market area inequality consists of areas with high concentration of poverty, high percent of households dependent on public income, and high unemployment. The second factor is termed residential stability. Variables that load high on the residential stability factor include the percent of housing owners (loading=.85) and the percent of non movers (residents that had stayed in their homes in the last five years) (loadings =.69). The two variables are positively associated with the factor. Based on the sign of these loadings, labor market area residential stability consists of areas with 99 high proportion of housing owners and high proportion of residents that have not moved in the last five years. The third factor is immigration concentration Variables that load high on the immigration factor include the percentage of Hispanic residents (loading=.92) and the percentage of immigrants within a labor market area (loading=.67). Factor loadings are presented in table 5. These three factors, economic disadvantage, residential stability, and immigration concentration are consistent with those in Sampson, Raudenbush, and Earls’ study (1997). Table 5. Oblimin with Kaiser Normalization Rotated Factor Pattern (factor loadings 2.60) in 112 Labor Market Areas. iVarrabl—e ' ‘ FactorLoadrng '3 Residential Stability Percent Owner Occupied Housing . 8 5 3 Percent Nonmovers . 6 8 8 Economic Disadvantage Percent of Households on Public Assistance . 9 4 8 Percent Unemployed . 8 8 4 Percent of households in Poverty . 8 4 3 Immigrant Concentration Percent of Hispanics . 9 1 8 Percent of Immigrants . 6 7 4 Extraction: Alpha-Factoring Method 100 Table 6 displays the distribution of families and households with self- employment by family and household characteristics. The data in table 6 show that married families are more likely than non-married families and households to have self- employment. F emale-headed households are the least self-employed than other forms of families and households. Families with preschool children are less likely than other families and households to have a self-employed member. Families living with adult relatives are more likely than families without adult relatives to have a self-employed member. Families and households that have moved in the last five years are less likely than those that have stayed to have a self-employed member. Also, immigrant families are less likely to have a self-employed member. Families and households living in an owned home are more likely to have a self-employed member. Table 6 also shows significant differences in self-employment across levels of household income. Families and households with incomes less than $10,000 are less likely to have a self-employed member than other families and households. In contrast, families and households at the top of the class ladder (starting at $50,000 up to $100,000 and higher) are more likely to have a self-employed member. Families and households with incomes between $10,000 and $25,000 are more likely than those with incomes less than $10,000 to have a self- employed member. However, families and households with incomes between $25,000 and $34,000 are less likely than those with incomes between $10,000 and $25,000 to have a self-employed member. 101 Table 6. Percentage of Families and Households with Self-Employment by Family and Household Characteristics (N=75280). — Family and Household Not Working Not Self- Self-employed Characteristics employed Percentages Family and household structure‘ * ’ Marriedcouples 9.1 72.0 18.9 Male-headedhouseholds 13.7 73.9 12.4 Female-headed households 2 6 . 5 6 9 . l 4 . 5 Cohabitinghouseholds 10.5 79.5 10.0 Life course stages’ ’ " NochildrenrmderlSyears 14.4 68.7 16.8 Preschoolchildren 10.6 73.6 15.8 School-agechildren 7.5 75.5 17.0 Adolescent Children 7 . 8 7 5 . 2 17 . 0 Presence of Adult Relatives’ ’ ’ Noadultrelative 12.4 71.3 16.2 One or more adult relatives 7 . 0 7 4 . 8 1 8 . 2 Residential mobility‘ ’ ’ Movers 11.9 75.6 12.5 Non-movers 11.0 68.8 20.2 Immigrant status‘ ‘ ‘ Atleastonememberorall 12.1 76.1 11.8 is immigrant Nomemberisimmigrant 11.4 71.9 16.7 Own a home’” Yes 9 . 2 7 2 . 0 1 8 . 8 102 — Family and Household Not Working Not Self- Self-employed Characteristics employed No 1 8 . 8 72. 0 9. 3 Household Income ’ ’ ’ Lessthan$10,000 49.3 37.6 13.1 $10,000-24,999 17.8 65.5 16.8 $25,000-34,999 6 . 8 77 . 2 16 . 0 $35,000-49,999 3 . 5 81 . 9 14 . 6 $50,000-74,999 2 . 4 82 . O 15 . 6 $75,000-99,999 1 . 7 74 . 2 24 .1 $100,0000rgreater 2.2 57.4 40.4 Household Poverty’ ’ * Poor 39.2 45.7 15.1 NotPoor 7.1 76.1 16.9 C are education‘ “ ‘ Lessthanhighschool 30.0 57.8 12.2 Highschool 11.8 71.8 16.5 Somecollege 7.3 75.7 17.0 Collegeormore 3.7 76.9 19.4 Corerace‘” Non-HispanicWhite 10.6 72.0 17.3 Blacks 26.5 69.5 4.0 AsianAmericans 14.8 74.8 10.3 NativeAmericans 29.0 65.4 5.6 Hispanics/Latinos 20.9 68.3 10.8 OtherRaccs,includingMixed 10.9 76.5 12.6 C are average age’ ‘ * 103 Family and Household Not Working Not Self- Self-employed Characteristics employed Lessthan30 11.8 78.2 9.9 30-39 6.3 76.9 16.8 40-49 5.5 75.4 19.2 50-59 14.4 65.1 20.4 600rmore 40.5 43.5 16.0 Number of cases 8592 54176 75280 '/. Self-employed 11 . 4 72 . 0 16 . 6 ”" p<.001; ”p<.01; ‘p<.05; ns=not significant Note also that the rate of self-employment decreases further for families with incomes between $35,000 and $50,000 compared to those with incomes between $25,000 and $35,000, and those whose incomes are between $10,000 and $25,000 income brackets. Note also that the relationship between household income and self- employment varies by non-metropolitan and metropolitan residence — families and households with incomes greater than $50,000 in non-metropolitan areas are greatly more likely than their metropolitan counterparts to have a self-employed member. The relationship between household income and self-employment is presented in figure 6. Table 6 also shows that poor families and households are less likely than non- poor families and households to have a self-employed member. Also, families and households with lower levels of education are less likely than those with higher 104 Figure 6: Percentage of Families and Households with a Self- employed Member by Income Categories (000's) so 50 4 i. .. « 2 at g so 13 <10 9,. a 10.25 g 20 1'25-35 I35-50 10 . BSD-75 I75-100 o .a 100+ All Households Metro Househokls Non-metro Househokls education to have a self-employed member. Minority families, especially Blacks and Native Americans, are less likely to have a self-employed member. The rate of self- employment increases as age increases and then decreases for age 60 or greater. Table 7 shows the fiequency distribution of families and households with additional earners across the categories of a series of family and household characteristics. The results indicate that single-headed families and households, especially male-headed households, are more likely than married couples and cohabiting households to have additional earners. Cohabiting households are the least likely of all household types to have non-core members employed. 105 Al Pres N0 .9877 Table 7. Percentage of Family and Households with Additional Earners by Family and Household Characteristics (N=75280). Family and Household Not Working No additional Additional Characteristics Earners Earners Percentages Family and household structure." Marriedcouples 9.1 79.1 11.8 Male-headedhouseholds 13.7 40.1 46.2 Female-headed households 2 6 . 5 4 6 . 4 2 7 . 1 Cohabiting households 1 0 . 5 8 3 . 9 5 . 6 Life course stages'” NochildrenunderISyears 14.4 64.3 21.3 Preschoolchildren 10.6 86.3 3 1 School-agechildren 7.5 84.0 8.5 Adolescent Children 7 . 8 7 0 . 6 2 1 . 6 Presence of Adult Relatives” Noadultrelative 12.4 83.7 3.9 One or more adult relatives 7 . O 32 . 8 6 O . 2 Residential mobility'“ Movers 11.9 75.2 12.8 Non-movers 11.0 73.2 15.8 Immigrant status”. Atleastonememberorall 12.1 69.1 18.9 rs rmmrgrant Nomemberisimmigrant 11.4 74.2 14.4 Own a home.” Yes 9 . 2 7 5 . 9 1 4 . 9 106 ’ - fan PM No: l‘ 1.1er Family and Household Not Working No additional Additional Characteristics Earners Earners Percentages No 18.8 68.3 13.0 Household Income ... Lessthan310,000 49.3 44.8 5.9 $10,000—24,999 17 . 8 72 . 5 9. 8 325,000-34399 6. 8 81 . O 12 . 1 53100049399 3 . 5 80. 6 15 . 9 $50,000-74,999 2 . 4 75 . 0 22 . 5 375,000-99399 l . 7 71 . l 27 . 2 $100,0000rgreater 2.2 78.8 19.0 Household Poverty". Poor 39.2 53.7 7.1 NotPoor 7.1 77.3 15.6 Core education'" Lessthanhighschool 30.0 52.6 17.4 Highschool 11.8 72.6 15.7 Somccollege 7.3 79.0 13.7 Collegeormore 3.7 85.3 11.0 Coreracem Non-HispanicWhjte 10.6 75.0 14.3 Blacks 26.5 56.3 17.2 AsianAmericans 14.8 69.3 15.8 NativeAmericans 29.0 56.8 14.2 Hispanics/Latinos 20.9 61.2 18.0 OtherRaces,includingMixed 10.9 80.3 8.8 107 Family and Household Not Working No additional Additional Characteristies Earners Earners Percentages Core average age." Lessthan30 11.8 79.9 8.2 30-39 6.3 87.4 6.3 40-49 5.5 69.7 24.8 50-59 14.4 64.4 21.2 600rmore 40.5 44.4 15.1 Number of cases 8592 55814 10874 °/. Self-employed 11 . s 74 . 1 14 . s "*p<.001; "p<.01; ’p<.05; ns=not significant About 46 percent of male-headed households and about 27 percent of female-headed households have additional earners. In contrast, about 12 percent of married couples and 6 percent of cohabiting households have additional earners. Families with no children living a home and those with adolescent children are more likely than those with preschool or school-age children to have additional earners. This is expected because families with young children have greater household responsibilities than those with no children or those with adolescent children. Non-core members of the household including adult children and other relatives may help in child responsibilities instead of seeking employment. Families and households that have moved in the last five years are less likely than those that have not moved to have additional earners. Home owners are also more likely 108 than those without a home to have additional earners. Families and households with higher income have as expected higher likelihood for additional earners. Poor families and households are less likely than non-poor households to have additional earners. In terms of education though, families and households with low education (high school or less) are more likely than those high education (some college or higher) to involve additional members in the paid employment. Hispanic, Black, and Asian American families and households are more likely than non-Hispanic white and Native Americans to have additional earners. Families and households in which core members are between 40 and 49 years of age, followed by those between 50 and 59 years of age, are more likely to have additional earners than the others, especially the younger generations (less than 40 years of age). 99mm of Labor Masks: Asa W Table 8 displays the correlation matrix of labor market area characteristics including means, standard deviations, and bivariate correlations. At the labor market area level, the average percentage of families and households involved in self- employment is about 17 percent with a standard deviation of approximately 7 percent. This indicates greater variations in self-employment across labor market areas. There is a strong and positive correlation between the percentage of families and households involved in self-employment and the percentage of agriculture, forestry, and fisheries within a labor market area Which indicates that self-employment is concentrated in labor market areas that are agriculturally dependent. There is also a significant positive correlation between the percentage of self-employment and the percentage of residents 109 that have not moved in the last five years. This suggests that self-employment is positively associated with residential stability within a labor market area. However, high concentration of traditional high-wage industries (mining, government, and high-wage manufacturing) and consumer service industries are negatively associated with self- employment. Economic disadvantaged areas, that is, with high unemployment rates and high dependency on public assistance, are strongly and negatively associated with lower rates of self-employment. 110 6s 66... 66 88> M: 66.5 .x. 636- 2&6- ~36 626 636- 65.6- H26 216 H~m6 mm~6 2:6 366 6:6 626: v-6.. 636 3~6 6666- 626 6666- 6:6- 636 8: mu_o:om=o:vovuom-o_ufiom$ ~36| 6:6- 636- 866.. $66- ~36 6-6- 686- 6-6u :36. ~36 8: SquameEHfi. 686 3H6 2.66 866 ~36 H36- m36 686- «$6 t.~6 ~36- Ab: m§otoEuaZ$ $~6- 2&6- 6666- 266- 626. ~36 $~6u 6:6- 3H6- ~36. :36 3: §0t05<§_m<.x. 6666- 686. 236- 636- 866- 636 Z~6u 656 1.86- 63.6- 636 An: 858.6566 63.6 686 £66- ~36- 626 636 ~86- 3H6 ~36- 636- 636 Av—v Mow—mfi. ~26- «$61 636; H36- ~86- 2.66 mom6- «R6- 686- 686 626 $3 Bosomnmfcafibfiocwoe :66 ~36 686 636 636 636- MR6 866 :36- ~86- 2&6- AN: 3§mmmmomEmEogom=om$ 6664 366 $66 656- $26 636 ~36 >56 656- A3 w:_m=omuoasooo._o=30$ 86; ~86- 286. 866 626 -m6 S~6 686- A8 Eo>oEéoZ§ 6664 -66 6666 £66- 686.. 366.. 2:6. A5 moomtomcofiameoo$ 6664 6666- H36- 636- H36- $~6 GV moomtomuoosvoig 6664 $66 :36. H26- 636- A9 mufignfi owagioqfi. 6664 ~26- 636- 626 Avv moEQvSomaBAmifi. 6664 686 626- A0 mGEwE$bmeom6h£=otm<$ 83 8...- E 692686.266: oooé A: when—Ham Ecfl§w:ovebee 6664 mm~6 S~6- 626 636 $66 366 62.6 3: mEonomzcmvowaoE-ofissmfi. 666; m-6- 6666 62.6 626 $66 626- 8: $§E_E§o\e 86; ~26- ~36- ~m~6- ~86 686 p5 manor—08.6.9923 6664 :66 ~36 866 ~v~6- G: gotogcumzce 6664 686 $86 366- 6: «omega—mix. 6664 8~6 $~6 Av: ace—max. 83 E..- a: 6666685558620: 83 a: 8:568... 235.6 n36 mg; 62.6 63:“ 9.66 :64 «3.. n36 .36 >HGA—hm 36.3 866 «2.6 2.66 366 #34 6:6 mooén an... Zea—2 SS 33 and ~53 GS 95 $5 Ana Ana .3388 82:86:20 8:. .952 .33 6 552 5222.5 .6 as: 112 Greater concentration of minorities (Blacks and Hispanics), immigrants, and female- headed households are associated with lower rates of self-employment. However, labor market areas with greater concentration of Native Americans are associated with higher rates of self-employment. The average percentage of families and households with additional earners is estimated at about 14 percent with a standard deviation of 3 percent. This indicates relatively not much variation across labor market areas compared to self-employment. Labor market areas with a greater concentration of agricultural industries, construction and low-wage industries, residential stability, (non-mover residents, housing owners), economically disadvantaged areas, and those with a high concentration of Native Americans have lower rates of families and households with additional earners. In contrast, labor market areas with high-wage and producer service industries, as well as labor market areas with a high concentration of Blacks, Hispanics, and Asian Americans and a high concentration of residents with greater than high school level of education, high concentration of immigrants and female-headed households are associated with greater rates of families with additional earners. The preceding analysis reports the bivariate relationship of household and labor market area characteristics with family and household work strategies. However, it does not tell us about inter—correlations among these factors, nor it does not indicate the relative importance of independent variables in influencing the likelihood of families and households to have additional earners or having self-employed members. In the following analysis, multilevel discrete models are used to determine and assess the 113 relative importance of multilevel factors influencing the odds of additional earners and self-employment. WWMLEWQQWMM Table 9 displays the results from a multilevel discrete model of self-employment. Model 1 reports estimates of the unconditional model of self-employment. The fixed part of the model describes the average log-odds of self-employment. The average log- odds of self-employment are estimated at -l.753 (se=.027). This corresponds to an estimated average odds of self-employment of approximately .173”. In terms of probabilities, it is equivalent to an average probability of self-employment of .148. Table 9. Multilevel Model Estimates of Self-employment -- The Unconditional Model _ Parameter Model 1 Estimate S. E. Fixed Effects: Intercept -l.753 .027 Random Effects: Level 2: Variance, 0’”, - 068 - 01 1 18 Themodelestimates fl areexpressedintermsoflogitsorlogodds, i.e., log(1ti/l-1tij). Thiscanbewn'tten inter-ms ofodds by computingthe exponential ofB. Hence, exp (-1 .753)=. 173 isthe estimate ofthe average oddsofself—employment. Thelogoddscanalsobeexpressedintermsofprobabilitiesbythefollowmgfonmlm n35: exp(B)/(l+exp(fl)). Hence, the average probability of self-employment is estimated at .173/(l+.173)'=.148. 114 The random part of the model indicates a between labor market area variance of 0.068. This gives a 95% confidence interval of the average odds of self-employment that ranges from .104 to .289 '9. Thus, there is evidence of significant differences amongst labor market areas in the likelihood of families and households to have a self-employed member. Next, family and household characteristics as well as labor market area characteristics are included in the model as predictors of self—employment. E81118! and Hans-shalt! 3mm and W Do families and households with different structures differ in their likelihood of self-employment? The results in model 2 indicate a significant and positive difference between the log-odds of self-employment for married families and the log-odds of self- employment for non-married families and households (i.e., male-headed, female-headed, and cohabiting families and households combined), fl=.75 8 [se([3)=.040]. Taking antilog, this indicates that the odds of self-employment for married families are estimated to be 2.134 [i.e., exp (.758)] times as much as non-married family/households, other things being equal. Thus, when the presence of children and adult relatives in the household are controlled, married families are about 113% more likely than non-married families and households to have a self-employed member. l9 Ammo-minim ofthe log-odds ofself-employrnmtiscomrtcdusingthefonmrla ,6 :t 1.96./se(fl) and then converted into odds. 115 Table 9. Multilevel Model Estimates Predicting Self-employment —— Explanatory Models. — Parameter Model Estimates (S.E.) Model 2 SE Model 3 SE Model 4 SE Fixedeffect: Intercept —1.806 0.027 -1.830 0.028 —1.857 0 030 MarriedFamilies 0.758 0.040 0.516 0.041 0.442 0 042 Cohabitinghouseholds 0.290 0.081 0.332 0.081 0.297 0 082 Female-headedhouseholds -0.681 0.093 -0.731 0.093 -0.671 0 093 PresenceofChildren -0.069 0.023 -0.045 0.024 -0.045 0.024 PreschoolChildren —0.109 0.032 0.006 0.032 0.027 0.033 SChOOI-ageChildren 0.006 0.042 0.029 0.042 0.036 0.042 PresenceofAdultRelatives 0.084 0.030 0.029 0.030 0.005 0 031 Residential Mobility -0.207 0.025 -0.225 0.025 OwnaHome 0.587 0.036 0.502 0 037 Household Income (‘000) 0.005 0.0004 High School 0.027 0.042 SomeCollege 0.144 0.043 College 0.198 0.046 RandomEffects: Level 2; Variance, 02,-- 0.066 0.011 0.067 0.011 0.079 0.013 Level 1: Variance, e20; 1 0 1 0 1 0 Non-married families and households also differ in their likelihood of self- employment The difference in the log-odds of self-employment between cohabiting families and single-headed families, is positive and statistically significant, 8 = .290 [se([3)==.081)]. Translated into odds, this indicates that the odds of self-employment for cohabiting households are about 1.336 [i.e., exp(.290)] times the odds of single-headed 116 families and households (male and female-headed combined). This indicates that when other variables in the model are held constant, the odds of self-employment are about 34 percent greater for cohabiting families and households than single-headed families. As expected, the relationship of female-headed families and self-employment, in contrast with male-headed families, is negative and statistically significant, 13= -.681 [se( B)=093]. In other words, the odds of self-employment for female-headed families and households are estimated to be .506 times lower than male-headed families, i.e., about 49% lower. Thus, female-headed families and households are less likely than male- headed families and households to have a self-employed member. Thus, the results in model 2 show that families and households with different household structures have different likelihood of self-employment. Next, model 2 controls for household composition variables including the presence or not of children under the age of 18 at home, the age of children, and the number of adult relatives in the household. flexes-s 2f cam and W Do the presence and age of children explain the differences amongst families in the likelihood of self-employment? It is expected that families with children under 18 years of age, especially younger children, are more likely than those who do not have children in their homes to lmve a self-employed member. The presence of children under 18 years of age in the household yields a significant negative relationship with self- employment, [3= -.069 [se( 13):.023]. This indicates the difference between the average log-odds of self-employment for families with children and those without children Expressed in terms of odds, this means that, the odds of self-employment for families 117 with children under 18 years of age at home are estimated to be 0.933 times the odds of sel-employment for families without children at home, i.e., about 7 percent lower. Thus, families with children under 18 years of age are less likely than those with no children under 18 years of age at home to have a self-employed member. Furthermore, comparing families with children under 18 years of age, I find that families with preschool children are less likely than those with school-age and/or adolescent children to have a self-employed member. The odds of self-employment for families with preschool children are estimated to be 0.897 [i.e., exp(-0.109)] times the odds of self-employment for families with children between six and seventeen years of age, i.e., about 10% lower. Thus, families with preschool children are less likely than those with school-age and adolescent children to have a self-employed member. However, the difi‘erence between the log-odds of self-employment for families with school-age children and those with adolescent children is not statistical significant. The results in model 2 indicate that, regardless of the family/household structure and the presence of adult relatives, families with children under 18 years of age at home are less likely than families without children at home to have a self-employed member. Also, families with younger children at home are less likely than those with older children to have a self-employed member. Families with preschool children are less likely than those with school-age and/or adolescent children to have a self-employed member. Next, the presence of adult relatives living in the household is included in the model as a control. 118 Model 2 controls for the presence of adult relatives in the household The associated assumption is that the difference in the likelihood of self-employment between families and households with different structures may be due to the number of adult relatives in the family (adult children, brothers, and sisters, uncles and aunts, parents, and other adult relatives). Holding other variables in the model constant, that is, the family/household structure, the presence and age of children under 18 in the household, the results in model 2 indicate that the relationship between self-employment and the presence of adult relatives (other than the householder, the spouse or partner) living at home, is significant positive, B= .084 [se(li)=.030]. This indicates that each additional adult relative in the family increases the odds of self-employment by about 9 percent, i.e., 100*[exp(.084)-1]. In summary, controlling for the presence of adult relatives and young children in the family does not change the relationship between family structure and self- employment. Married-couple families remain more likely than non-married families and households to have a self-employed member, cohabiting families and households are more likely than single-headed family/households to have a self-employed member, and female-headed families and households are less likely than male-headed family/households to have a self-employed member work. Model 3 adds family residential mobility as a proxy for family social capital. Does residential mobility of a family/household explain the differences in the odds of self-employment? Results in model 3 indicate a significant and negative relationship 119 between family residential mobility and self-employment, B= -.207 [scaly—1025]. In terms of odds, this indicates that the odds of self-employment for families and households that have moved in the last five years are about 19 percent [i.e., 100‘(exp(- .207)-l)] lower than the odds of self-employment for those that have not moved in the last five years. Thus, family residential mobility reduces the likelihood of self- employment Model 3 also controls for home ownership. Home ownership affects residential mobility. People who own homes are less likely than those who do not to move. Home ownership may also affect self-employment in two ways. It can be used as an asset to borrow a start-up capital or as a convenient place to do self-employment work, including homework activities. It is expected that families and households with home ownership to have greater chances of being involved in self-employment than families and households who do not own a home. The results in model 3 indicate a significant and positive relationship between ownership of a home and the log-odds of self-employment, B=0.587 (se= .036). In terms of odds, this indicates that, after controlling for the other variables in the model, families and households with a home are about 80 percent [i.e., 100*(exp(.587)-1)] more likely than those who do not own a home to have a self- employed member. Thus, home ownership increases the likelihood of self-employment. Hansehald Income and Want Model 4 controls for household income as a proxy of household economic capital. As expected, the relationship between household income and self-employment is significant positive, fi=.005 (se=-.0004). This indicates that, for each additional $10,000, 120 the odds of self-employment increase by 5 percent. This implies that the greater the income the greater the chances for self-employment”. This is probably so because those with greater income have investments to start a small business and/or can have access to loans fi'om banks, because they are economically reliable. Emil! 5411128890 and Wm Model 4 also controls for family education. The effect of college education on the odds of self-employment is significant and positive, net of all other effects. Families with a college or more education are more likely than those with some college or less education to have a self-employed member, B= 0.198 (se=0.046). In terms odds, this indicates that the odds of self-employment for families and households in which at least one or both core members have a bachelor or greater level of education are about 22% greater than the odds of self-employment for families with less than college education Also, families with some college education are more likely than those with high school or less education to have a self-employed member, B= 0.144 (se=0.043). In terms of odds, this indicates that families and households with some college education have about 15 percent greater chances for self-employment than those with high school or less education. Families and households with a high school level of education are not significantly different from those with less than high school education in their likelihood of having a self-employed member, B=.027, [se(B).=.042]. 20 Because household income was measured in 1989 and employment, including self-employment’s date of referenceistheCensusreferenceweekin 1990, householdincomeisassumedtocausetheinvolvementinself- employment in 1990 (one direction causality). However, if it was the same reference time, it would have been possible for self-employment to increase household income, which would suggest double-causality, i.e., household income causes self-employment and self-employment causes household income. 121 W and mum Model 5 includes race/ethnicity as a control variable. The relationship of race/ethnicity and self-employment is significant negative, 13= -0.472, [se(B)=.064]. This indicates that minority families have significantly lower (about 38 percent) likelihood of self—employment than non-Hispanic White families. Model 5 also controls or age of core members in the household It is expected that the odds of self-employment would increase with age. Age square is added in the equation to control for curvilinear relationship between age and self-employment. The estimated coefficient of age is significant and positive, [35031, [se(B)=.002]. This indicates that each additional year of age increases the odds of self-employment by about 3 percent, i.e., 100*[exp(.03l)-l]. As expected also, the estimated coefficient for age square is significant and negative, [i=- .074 [se (B)=.Ol 1]. This means that the odds of self-employment increase at a diminishing rate with age. The introduction of these control variables in model 5 significantly improves the explanatory power of the model. In comparison to model 4, the estimated coefficients in model 5 for many variables have changed considerably. Married families remain more likely than non-married families and households to have a self-employed member. However, the average difference between the odds of self-employment for married and non-married families decreased Cohabiting households also remain more likely than single-family households to lmve a self-employed member, but the difference between the two in their effect on self-employment increases. Cohabiting households tend to be younger. 122 Table 9. Multilevel Model Estimates Predicting Self-employment, Continued... — Parameter Model Estimates (SE) Models SE Model6 SE Fixed effect: Intercept —1.791 0.031 -1.748 0.029 MarriedFamflies 0.316 0.042 0.315 0.042 Cohabiting 0.387 0.082 0.385 0.082 Femalfiheaded -0.850 0.094 —0.849 0.095 PresenceofChildren 0.147 0.029 0.147 0.029 PreschoolChildren 0.257 0.036 0.258 0.036 School-ageChildren 0.154 0.043 0.154 0.043 PresenceofAdult -0.069 0.031 -0.069 0.031 ResidentialMobility —0.080 0.026 -0.080 0.026 OwnaHome 0.344 0.038 0.343 0.038 Householdlncome 0.004 0.0004 0.004 0.0004 HighSchool 0.102 0.043 0.102 0.043 SomeCollege 0.259 0.044 0.261 0.044 College 0.268 0.047 0.272 0.047 Minority -0.472 0.064 -0.467 0.064 Age 0.031 0.002 0.031 0.002 Agesquare(‘00) —0.074 0.011 -0.074 0.011 Non-Metro 0.160 0.025 RandomEfl’ects: Level 2; Variance, 02’” 0.074 0.012 0.050 0.009 Levell=anc€~ ,_ ,_ e _ 9 Once age is controlled, they are way more likely than single-family households to have a self-employed member, suggesting that the presence of a partner makes a difference. Also, female-headed households remain less likely than male-headed households to have a self-employed member. However, the difference between female- and male-headed 123 households in the odds of self-employment widens up. The most significant changes occur in the effect of having children under the age 18 and the age of children on the odds of self-employment Once the age of the householder and/or the spouse/partner is controlled, having children under the age of 18 in the household significantly increases the likelihood of self-employment Families with preschool children also become significantly more likely than families with school-age and/or adolescent children to have a self—employed member. Families with school-age children are also more likely than those with adolescent children to have a self-employed member. This suggests that once age is controlled, having children at home, especially younger children, is associated with greater likelihood of self-employment The effect of residential mobility also changed after age was introduced in the model. This is expected because younger families are more likely to move than older families. Despite the change in the absolute value of the effect of residential mobility, residential mobility remains significantly and negatively associated with the odds of self- employment. The effect of family education also is triggered by the introduction of age in the model. Families with a high school education are more likely than those with less than high school education to have a self-employed member. Before age was introduced (model 4), there was no significant difference between families with a high school education compared with those with less than high school education. Also, families with some college education are more likely than those with high school or less education to have a self-employed member and those with a college education are more likely than 124 other families (i.e., with less than a college level of education) to have a self-employed member. The random part of the model indicates a slightly increase in the between labor market area variance from 0.068 in the unconditional model to 0.074 in model 5 —— approximately 9 percent increase. Even after controlling for family and household characteristics, there is evidence of significant differences between labor market areas in the likelihood of self-employment. Next, the labor market area (level-2) variables21 are included in the model to determine if they account for the between labor market area differences in the likelihood of self-employment Wit-an 1 MW 85514299.: and Wm Does metropolitan and non-metropolitan residence explain any difl‘erences between labor market areas in the likelihood of self-employment? The theory described led me to expect that families and households living in non-metropolitan labor market areas would be more likely to have a self-employed member than those living in metropolitan labor market areas. The results are displayed in Table 9 (model 6). I find that families living in non-metropolitan areas are more likely than those in metropolitan areas to have a self-employed member, B=.160 [se(0)=.025], controlling for family and household characteristics. This indicates that the odds of self-employment for families living in non-metropolitan labor market areas are about 17 percent greater than those of families and households living in metropolitan areas. Notice also that by introducing 2! Thelabormarketvariablesinthemodelhaveajsubscriptwhilefann’lyandhouseholdvariableshaveij subscripts. 125 non-metropolitan variable in the model, the random part of the model decreases from 0.074 (model 5) to 0.050 (model 6). Residential stability and industrial structure measures are next entered in the model. Industrial Emma: and 55mm Does the industrial structure of a labor market area affect any differences amongst labor market areas or the differences between metropolitan and non- metropolitan in the likelihood of self-employment? A measure of the labor market area industrial structure -- the ratio of core industries (hi gh-wage manufacturing industries and producer services) to peripheral industries (agriculture, fishery, and forestry, low- wage manufacturing, and consumer services) is included in model 7. I find that labor market area with greater proportion of core industries to be associated with lower likelihood of self-employment, all other effects held constant, B= -.795 [se(p)=.093]. This indicates that families living in labor market areas with greater proportion of core industries are less likely than those in areas with peripheral industries to have a self- employed member. Thus, families living in labor market areas with greater availability of better jobs are about 55 percent less likely than those living in labor market areas with poor jobs to have a self-employed member. In other words, the greater proportion of good jobs (proportion of core industries) the lower the likelihood of self-employment. DIM Mikel .A_r§a 8551512111181 3151211113: and W In addition to industrial structure and non-metropolitan residence, model 7 includes the labor market area residential stability indicator. 126 Table 9. Multilevel Model Estimates Predicting Self-employment, Continued... Parameter Model Estimates (SE) Model7 SE Model8 SE Fixed effect: Intercept 0.388 0.408 0.812 0.387 Married Families 0.313 0.042 0.314 0.042 Cohabiting households 0.380 0.082 0.381 0.082 Female-headed households -0.850 0.095 —0.847 0.094 Presence ofChildren 0.146 0.029 0.145 0.029 Preschool Children 0.260 0.036 0.258 0.036 School-age Children 0.155 0.043 0.154 0.043 PresenceofAdultRelatives ’0-068 0-031 -0-067 0.031 Residential Mobility —0.079 0.026 -0.080 0.026 OwnaHome 0.343 0.038 0.343 0.038 Household Income (000) 0.005 0.0004 0.005 0.0004 High School 0.103 0.043 0.100 0.043 Some College 0.264 0.044 0.261 0.044 College 0.275 0.047 0.271 0.047 Minority -0.462 0.064 -0.459 0.064 Age 0.031 0.002 0.031 0.002 Agesquare(‘00) -0.074 0.011 —0.074 0.011 Non.Meu-o 0.062 0.025 0.066 0.028 Residential Stability -0.016 0.005 -0.016 0.005 Industrial Structure -0.795 0.093 -0.863 0.092 Economic Disadvantage -0.037 0.008 Non-Metro’Industrial Structure '0 - 1 66 0 . 090 Random Effects: Level 2; vafiauoo, 02..- 0.025 0.005 0.018 0.004 Level 1: Variance, e2 .. 1 0 1 0 b To what extent families and households living in labor market areas with greater residential stability involve in self-employment work? It is expected that the labor market area residential stability would be positively related to self-employment, net of 127 family and household characteristics, non-metropolitan residential status, and the labor market industrial structure. I find that labor market area residential stability is significantly and negatively related to the adjusted log-odds of self-employment, B= -.016 [se(B)=.005]. However, in another analysis (data not shown), when non-metropolitan residence and industrial structure variables are not included in the model, the effect of residential mobility on the log-odds of self-employment is significant positive. Therefore, the negative relationship between residential stability and self-employment at level 2 is probably due to the suppressing effect of industrial structure and non- metropolitan residence indicators. The final model (model 8) includes the labor market area economic disadvantage/ inequality indicator and the interaction of non-metropolitan residence and industrial structure. I expected to find that the greater the labor market area disadvantage/inequality the lower the likelihood of self-employment The labor market area inequality yields a significant negative relationship with the log-odds of self- employment, B= -.037 [se(BFDOS]. In terms of odds, this indicates that the labor market economic disadvantage decreases the odds of self-employment by about 4%, i.e., 100*[exp(-.037)-1]. Thus, families living in labor market areas with high economic disadvantage/inequality, i.e., with high poverty and unemployment rates as well high dependency on public monies, are less likely than those living in other labor market areas to have a self-employed member, all other variables in the model held constant. 128 55135101012150: It is expected that the effect of the labor market area industrial structure on self- employment varies depending on the metropolitan and non-metropolitan residence, given the metropolitan / non-metropolitan differences in industrial structure. An interaction of the metropolitan / non-metropolitan residence and industrial structure is included in model 8. The interaction effect between the two labor market area variables yields a significant and negative effect on the mean adjusted log-odds of self-employment, B= - .166 [se(B)=.089]. This suggests that families and households living in non-metropolitan areas with greater and better job opportunities are less likely than those in metropolitan areas with good job opportunities to have a self-employed member. The overall effect of industrial structure on the mean adjusted log-odds of self-employment is estimated at - 1.029 [i.e., -.863+(1)(-. 166)] for families living in non-metropolitan areas and -.697 [i.e., -.863+(-1)(-. 166)] for families living in metropolitan areas. Thus, families living in non- metropolitan areas are far less likely than their metropolitan counterparts to have a self- employed member if they have access to good jobs. The growth of new employment opportunities in non-metropolitan areas has been limited to low-paying jobs while the jobs in metropolitan areas are at both ends of the wage spectrum (Gorham 1992). Thus, families living in non-metropolitan areas with greater job opportunities may get involved in self-employment in order to complement their incomes, especially if quality of such jobs does not allow them to satisfy their needs. However, if the available jobs offer good 129 wages, then families, especially those in non-metropolitan areas, are less likely to have a self-employed member. 5mm The likelihood of families and households to have a self-employed member depends upon family and household social structure as well as the labor market area social context in which families and households are located The results in model 8 show that married families are more likely than non-married families and households to have a self-employed member. Among the ‘non-married’ families and households, cohabiting households are more likely than single-headed family/households to have a self-employed member. Also, female-headed households are less likely than male- headed households to do self-employment work. Furthermore, the results indicate that families living with young children under 18 years of age are more likely than those with no children at home to have a self-employed member. Among families with children under 18 years of age living in their homes, those with preschool children are more likely than those with school-age and/or adolescent children to do self-employment work while families with school-age children are more likely than families with adolescent children to have a self-employed member. The results in model 9 also reveal that family residential mobility is significantly and negatively associated with self-employment Although residential mobility is not a very good proxy for social capital, the results imply that family stability increases the likelihood of self-employment. As also expected, home ownership increases the chances for self-employment 130 The results show that the greater the household income the greater the changes of self-employment. Family education is positively associated with self-employment. Families with higher educational levels are more likely to have a self-employed member. Minority families are less likely than non-Hispanic White families to have a self- employed member. Age is found to be positively associated with the chances for self- employment —— with the older generation more likely than the younger generation to have a self-employed member. At the same time, the relationship between age and self- employment is found curvilinear — self-employment increases at a diminishing rate with age. The likelihood for self-employment is not only influenced by family and household characteristics, but also on characteristics of places in which they live in, most notably whether they reside in metropolitan or non-metropolitan, the industrial structure, the residential stability, and the labor market area economic disadvantage/inequality. At the labor market area level, when all family and household characteristics are controlled, living in non-metropolitan labor market areas is significantly and positively associated with self-employment. The results also show that families living in labor market areas with greater proportion of core industries, i.e., with good jobs, are associated with lower chances for self-employment. The interaction between non-metropolitan residence and industrial structure is significant negative. This suggests that families living in non- metropolitan areas with greater proportion of good jobs are less likely to have a self- employed member. The labor market areas with greater residential stability are 131 associated with lower chances for self-employment. Finally, the greater the economic disadvantage of the area, the lower the chances for self-employment WMMfiWmmWM MMMW For this analysis, a sample of families and households with at least one adult member other than core members is drawn. The results of multilevel models for the number of additional earners other than core members are presented in Table 10. The fixed part of the unconditional model (model 1) indicates a non significant average log- odds of additional earners, B= 0.004 [se=.022]. A non significant coefficient means that it is possible that [3 is zero in the population or it could be simply due to sampling error. The average odds of additional earners are estimated at 1.003. Converted into probabilities, this corresponds to a probability to have additional earners of about .501. Table 10. Multilevel Model Estimates of Additional Earners - Unconditional Model Parameter Model 1 Estimate 8. E. Fixed Effects: Intercept 0 . 004 . 022 Random Effects: Level 2: Variance, a”,l - 029 . 007 Level 1: Variance, e2 .. 1 o * The random part of the unconditional model indicates a between labor market area variance of 0.029. This gives a 95% confidence interval of the average odds of 132 having additional earners of [.670, 1.338]. This implies that the likelihood of having additional earners within families and households varies between labor market areas. Email! 8511811115 and the mains: of Additional Earners It is expected that the presence of additional earners within families and households varies by family and household structure. The results in Model 2 show that married families are less likely than “non-married” families and households to have additional earners, 0= -.401 [se(8)=.044]. In terms of odds, this indicates that married families are about 33% times as less likely as “non-married” families and households to have additional earners other than core members. The results in Model 2 also indicate that cohabiting families and households are less likely than single-headed families and households to have additional earners, I3= -.987 [se(]i)=.108]. Converted into odds, this indicates that the odds of additional earners within cohabiting families and households are about 63%, i.e., [lOO‘exp(-.987)-l] lower than the odds of additional earners within single-headed families and households. F emale-headed families and households are less likely than male-headed families and households to have additional earners, B= -.373 [se(fi)=.066]. In terms of odds, this indicates that the odds of additional earners for female-headed families and households are estimated to be 31% as lower as those of male-headed families and households. To sum up, “non-married” families and households are more likely than married families to have additional earners. Also, single-headed families and households are more likely than cohabiting households to have additional earners. Finally, female-headed families and households are less likely than male-headed households to have additional earners. 133 Table 10. Multilevel Model Estimates of Additional Earners — Parameter Model Estimates (S.E.) Model 2 SE Model 3 SE Model 4 SE Fixed effect: Intercept 0.017 0.021 0.030 0.022 -0.278 0 043 MarriedFamilies -0.401 0.044 -0.173 0.047 —0.143 0.048 Cohabiting households -0.987 0.108 -0.540 0.115 -0.551 0.115 Female-headed households -0.373 0.066 -0.115 0.069 -0.094 0.069 Presence ofChildren -1.374 0.033 -1.385 0.033 Preschool Children -0.347 0.061 0.326 0.061 School-age Children -0.005 0.046 -—0.0002 0.046 Residential Mobility 0 . 094 0 . 033 Random Effects: Level 2; Variance, 02,.- 0.026 0.007 0.025 0.007 0.036 0.007 Level 1: Variance, 620-,- 1 0 1 0 1 0 Model 3 includes the life course stage variables. I find that when controlling for family/household structure, families with children under 18 years of age at home are less likely than those without children to have additional earners, B= -1.374 [se(B)=.033]. This means that the odds of having additional earners for families with children under 18 years of age are about 75% lower than the odds of additional earners for families without children under 18 years of age in their homes, other things hold constant. When comparing families with young children living in their homes, those with preschool children are more likely than those with school-age and/or adolescent children to have 134 additional earners, B= 0.347 [se(B)=.06l]. This implies that the odds of additional earners for families with preschool children increase by about 42% when compared to families with school-age and/or adolescent children in their homes. However, families with school-age children are not significantly different fiom those with adolescent children in their likelihood of having additional earners, B= 0.005 se(B)=.046]. In brief, the results in model 3 show that, regardless of family/household structure, having children under 18 years of age in the household decreases the odds of additional earners. However, having preschool children increases the odds of additional earners. Note that the estimated log-odds of additional earners for married, cohabiting, and female-headed families decrease substantially when life course variables are included in model. This indicates that when the presence and age of children are controlled, the gap between married and non-married families, between cohabiting families and single-headed families, and between female-headed and male-headed households in terms of having additional earners narrows. To what extent residential mobility influences the likelihood of additional earners? The next model includes residential mobility as a predictor of having additional earners. Results in model 4 indicate that residential mobility is positively associated with the log odds of additional earners, [3: 0.094 [se(B)=.033]. This indicates that the odds of having additional earners for families that have moved in the last five years are about 10 percent as higher as those of families that have stayed in their homes during the last five years. In brief, the results in model 4 show that families and households that 135 have moved in the last five years, i.e., with less residential stability, are more likely to have additional earners. It was expected that residential mobility would be negatively associated with additional earners. However, it is possible that people who have moved in the last five years are in migrants, i.e., people who are coming back to their communities. Therefore, they could be attached and well connected to their communities, and therefore, may have greater likelihood for additional earners. Model 5 adds household income and education as controls of household economic and human capitals. The relationship between household income and having additional earners is significant positive, 8: .009 [se(BF.0006]. In terms odds, this implies that the odds of additional earners increase by 9 percent for each additional $10,000, controlling for other variables in the model. Model 5 displays also the results of the relationship between family education and additional earners. I find that families with a college or higher level of education are less likely than families with less than college education to have additional earners. The odds of having additional earners for families with a college or greater education are estimated to be 41 percent lower than those of families with less than college education. Also, families with some college education are less likely than those with less than some college education to have additional earners. The odds of additional earners for families with some college education are about 16 percent lower than those of families with lower education Families with a high school education are not significantly different from those with less than high school education to have additional earners. 136 Table 10. Multilevel Model Estimates of Additional Earners, continued... Parameter Mael Estrmlates 75E; Models SE Model6 SE Fixed effect: Intercept —0.441 0.058 -0.580 0.062 MarriedFamilies -O.263 0.048 -0.269 0.052 Cohabiting households —0.682 0.115 -0.641 0.115 Female-headed households -0.064 0.070 —0.054 0.073 Presence ofChildren -1.339 0.033 —1.146 0.038 Preschool Children 0.336 0.062 0.299 0.064 School-age Children 0.020 0.047 0.066 0.047 Residential Mobility 0.120 0.033 0.126 0.034 Household Income (000's) 0.009 0.001 0.009 0.001 High School -0.066 0.046 -0.003 0.047 Some College -0.177 0.049 -0.094 0.051 College -0.529 0.057 -0.458 0.058 Minority Status ~0.270 0.059 Age 0.015 0.002 Age Square 0.160 0.015 Random Effects: Level 2; Variance, 02,-- 0.016 0.006 0.017 0.006 Level 1: Variance, 32015 1 0 1 0 — Model 6 controls for race and ethnicity. Results in model 6 indicate that minority status is significantly and negatively associated with the log odds of additional earners, 1% -0.270 [se(BF059]. This indicates that minority families and households are less likely than non-Hispanic Whites to have additional earners. The odds of additional earners for minority familiesand households are about 24 percent lower than the odds of 137 additional earners for non-Hispanic White families and households. In this model, imrrrigrant status was also included as a control. Because immigrant status was found not statistically significant with the odds of additional earners, it was removed in the analysis. As: and Additional carriers Model 6 also includes age and age square variables. Results in model 6 show that age is significant positive, B= 0.015 [se(B)=.002]. Age square is significant and positively associated with having additional earners, B= 0.160 [se(BFDlS]. This indicates a curvilinear relationship between age and the odds of additional earners, signaling an increase in the odds of additional earners as age increases. The introduction of control variables in model 5 and model 6 significantly enhances the explanatory power of the model. Note that the effects of mobility and family structure variables, but female headship, compared to their effects in model 4, have increased when income, education, race, and age variables are controlled The difference between female-headed fanrilies and households and male-headed families and households in terms of having additional earners is no longer significant once these controls are introduced in the models. Also, the intercept, i.e., the adjusted mean log- odds of additional earners has decreased considerably from model 4 to model 6. The next model includes a measure of industrial structure -— the ratio of core industries to peripheral industries in the labor market area. The results in model 7 indicate a significant and positive effect of industrial structure on the log-odds of 138 additional earners, B=0.233 (se=0.074). As expected, this indicates that the greater the prevalence of core industries in a labor market area the greater the likelihood of additional earners. Model 8 accounts for the economic disadvantage/inequality of a labor market area. It is expected that economic disadvantage/inequality of a labor market area would be negatively associated with the likelihood to have additional earners. I find that families and households living in economic disadvantaged labor market areas are less likely than those in prosperous areas to have additional earners, fi= -0.038 (se=0.008) . In terms of odds, this means that the odds of additional earners for families living in disadvantaged labor market areas are about 3 % lower than the odds of additional earners for families living in more prosperous labor market areas. Note also that the effect of industrial structure on the log odds of additional earners decreases, but remains positive. Thus, the labor market area economic conditions influence the likelihood of additional earners. W l W Essidsnss and Additional Earners The final model includes non-metropolitan/metropolitan variable. Does non- metropolitan residence explain any difference between labor market areas in the likelihood of families and households to have additional earners? As expected, the results in model 9 indicate that families and households in non-metropolitan areas are less likely than their metropolitan counterparts to have additional earners, B: -.049 [se(B)=.024]. The odds of additional earners for non-metropolitan families and . 139 households are about 5% as lower as metropolitan families and households. Note that the effect of industrial structure on the log odds of additional earners becomes not significant once non-metropolitan/metropolitan variable is included in the model. Table 10. Multilevel Model Estimates of Additional Earners, continued... — Parameter Model Estimates (S.E.) Model 7 SE Model 8 SE Model 9 SE Fixed effect: Intercept —0.901 0.122 -0.495 0.144 -0.414 0.147 MarriedFarnilies -0.264 0.052 -O.268 0.052 -0.267 0.052 Cohabitinghouseholds —0.637 0.115 -0.632 0.115 —0.632 0.115 Female-headed households -0.058 0.073 -0.055 0.073 -0.056 0.073 presence ofChildl-en -1.142 0.038 -1.145 0.038 —1.144 0.038 Preschool Children 0.297 0.064 0.298 0.064 0.297 0.064 School-age Children 0.069 0.047 0.067 0.047 0.068 0.047 Residential Mobility 0.125 0.034 0.124 0.034 0.125 0.034 Household Income (000's) 0.008 0.001 0.008 0.001 0.008 0.001 High School -0.002 0.047 -0.007 0.047 -0.007 0.047 Some College ~0.095 0.051 -0.097 0.051 -0.096 0.051 College —0.459 0.058 -0.467 0.058 -0.467 0.058 Minority Status -0.287 0.059 —0.275 0.059 -o.276 0.059 Age 0.015 0.002 0.015 0.002 0.015 0.002 Age Square 0.160 0.015 0.159 0.015 0.159 0.015 Industrial Structure 0.223 0.074 0.166 0.068 0.074 0.081 Economic disadvantage -0.038 0.008 —0.033 0.008 Non-Metro -0 . 049 o. 024 Random Effects: Level 2; Variance, 02,-- 0.013 0.005 0.007 0.004 0.006 0.004 Level 1: Variance, ezdj 1 0 1 0 1 0 140 Summary The likelihood of families and households to have additional earners depends on family and household characteristics on one hand, and on the other hand, it depends on non-metropolitan Imetropolitan residence, industrial structure, and economic conditions of the labor market areas in which families and households reside. Using the results in model 9, which includes all variables, I find that married families are less likely than “non-married” families and households to have additional earners. Cohabiting families and households are also less likely than single-headed families and households to have additional earners. Among single-headed families and households, female-headed households were found not significantly different from male-headed households to have additional earners. Having children under 18 years of age at home is negatively associated with the likelihood of having additional earners. Among families with children at home, I find that the younger the children are the greater the likelihood for additional earners. Fanrilics with preschool children are more likely than those with school-age and adolescent children to have additional earners. The higher the household income the greater the likelihood of additional earners. However, the higher the level of education, the lower the probability of additional earners. Minority families and households are less likely than non-minority families and households to have additional earners. Also, the chances of additional earners increase at a diminishing rate of age. Living in non-metropolitan labor market areas is negatively associated with the likelihood of additional earners. I also find that the greater the quality of jobs in a labor market area, as indicated by the greater prevalence of core industries, the greater the 141 chances for families and households to have additional earners. Finally, I find that families and households living in economically disadvantaged labor market areas have lower likelihood of additional earners. 142 Chapter 5 SUMMARY AND CONCLUSIONS This study has examined family and household work strategies focusing on self- employment and employment of additional earners in non-metropolitan and metropolitan labor market areas. The important contribution of this study is that it accounts for family and household factors as well as labor market area factors of family and household work strategies. The main finding of this study is that family and household work strategies vary within and across different labor market areas, between non-metropolitan and metropolitan labor market areas, and across different households given different family and household as well as labor market area social structures. The propositions and hypotheses tested were derived from the restructuring and the social capital / embeddedness perspectives. The results indicate that family and household work strategies, in this case self—employment and having additional earners, are influenced by both family/household characteristics and labor market area characteristics. This study contributes to the understanding of family and household strategies from a multi-level theoretical and analytical fi'amework This study, thus, contributes to research that attempts to link micro- to macro-phenomena of family outcomes. In this chapter, the results are first summarized in relation to corresponding hypotheses. Secondly, policy implications and recommendations are proposed Finally, limitations and future research are also discussed. 143 mammmmnmmcmmmm 122151295111: In relation to the social embeddedness/capital perspective, it was expected to find that family and household structure determines work strategies (Pl 1). A series of hypotheses was developed and tested It was hypothesized that married families are more likely than “non-married” families and households to have a self-employed member. This hypothesis is supported and consistent with previous studies on self- employment (Sanders and Nee 1997; Boyd 1991; Carr 1996), Married families are more likely than “non-married” families and households to have a self-employed member. It was also hypothesized that cohabiting families and households would be more likely than single-headed families and households to have a self-employed member. This hypothesis is also supported Thus, Cohabiting families and households are more likely than single-headed families and households to have a self-employed member. Among single-headed families and households, it was hypothesized that female- headed households would be less likely than male-headed households to have a self- employed member. The following result is found: Female-headed households are significantly less likely than male-headed households to have a self-employed member. The above results indicate that the likelihood of self—employment differs depending upon the family/household structure. The above relationship between family structure and self-employment remains the same after other family/household characteristics, such as family income, and labor market characteristics, such as the non- 144 metropolitan and metropolitan residence and the labor market area opportunity structure, has been controlled Why self-employment is more prevalent among married families, and to a lesser extent among cohabiting families and households, than among single- headed families and households? Why female-headed households are less likely than male-headed households to have a self-employed member? One can only speculate on the reasons for the differential effects of family structure on the likelihood of self- employment. One possibility is that married couples are more likely to have a self- employed member than other households because marriage offers an excellent partnership that is preferred to the use of other employees in a business. It helps not only to maximize the family income (rational interpretation) and solve business problems that employees may shirk (Borjas 1986). A related interpretation is that capital to start a business can be obtained fi'om one spouse. When one spouse is working, that may insure against the risk of a fluctuating self-employment income. The spouse can work in the family business on a part-time business or on a full-time basis when the business is successful. The other alternative explanation to why married couples have greater chances for self-employment is that a spouse is a trustworthy person and a valuable source of information to business ownership. The data does not reveal any explanation to why these differential effects of family structure on self-employment, however, the findings suggest the need for further understanding how family organize its resources to adapt any work strategy. In relation to the social embeddedness/capital perspective, the family is a valuable source of social capital for its members that may help or eventually constrain certain activities including work related ones. Family and household work 145 strategies are embedded in family/household social relations. Most studies on social capital at the family level have focused on the effect of parental and kin support on children (Coleman 1988; McLanahan & Sanderfur 1994; Parcel and Menagham 1994; Hao 1994). More research is still needed to understand fully the mechanisms through which the family-base social capital influence different family work strategies. Cohabiting households are like married couples on one account. They both have a very closely related adult (spouse and partner). This may explain why they are more likely than single-headed households to have a self-employed member. However, it does not explain why married households are more likely than cohabiting households to have a self-employed member. One explanation of the difference between married and cohabiting households in their likelihood of self-employment may reside in their differences. That is, the length of the relationship between married partners may have strengthened their social ties, thus their bounded solidarity. The other explanation may be cultural and legal ~— the legal ramifications of marriage make it strong and trustworthy. Therefore, married couples may have greater social capital based on bounded solidarity and enforceable trust than cohabiting households, and thus the greater the likelihood of self-employment for married couples than for cohabiting households. Comparing female to male-headed households, this study finds that female- headed households are less likely than male-headed households to have a self-employed member. Gender is an important factor of self-employment and its effect on the likelihood of self-employment cannot only be explained by the headship or the composition of the household Self-employment, on one hand, is very important for 146 families because it is an alternative and complementary productive activity to wage- salary employment Thus, to understand it, one needs to locate it in the larger structure of the economy. On the hand, it is equally important to analyze the social aspects of self- employment. The social embeddedness perspective offers a lens to analyze how social relations mediated by gender and class shape particular work experiences. Mingione (1991) is critical to the tendency to interpret informalization fiorn points of view almost exclusively concerned with its macro-economic origins and impact. Instead, he stresses the importance of social organizations, life-styles, and survival strategies as factors explaining the heterogeneity of working activities rather than as mere consequences of it (76). Self-employment can be seen as a result of the interaction between both economic and non-economic factors at the family and household level. Family/household structure also affects the likelihood of having additional earners in a different way than it does for self-employment. It was expected that: (I) married couples would be less likely than “non-married” households to have additional earners; (2) cohabiting households would be less likely than single—headed households to have additional earners; and that (3) female-headed households would be less likely than male-headed households to have additiOnal earners. The above hypotheses were supported and the following significant results were found: Married families are less likely than “non-married” families and households to have additional earners; Cohabiting families and households are less likely than single-headed families and households to lurve additional earners. 147 Female-headed families and households are less likely than male-headed families and households to have additional earners. These findings reveal that the family and household structure affects differently family and household work strategies. Single-headed families and households, particularly, male-headed households, are more likely than married or cohabiting households to have additional earners. One plausible explanation is in the male-headed households’ composition. They are more likely than other households to host non- relative members. This finding insights more research on cohabitation, including not just partners, but also non-related individuals who share the same roof. To what extent the social relations within this later type of cohabiting households affect their members’ work strategy? The presence and age of children were included in the model as controls. Families living with young children under 18 years of age are more likely than those with no children at home to have a self-employed member and the younger the children at home the greater the likelihood of self-employment. This may be explained by the fact that families with young children, especially preschool children, have greater time demanding for child care and rearing responsibilities. If there are few or no child care facilities, either because they are not available in the vicinity or because they are expensive, self-employment, particularly home-working self-employment, may be an attractive work strategy for families with young children, especially for women. Families with young children are found to have relatively fewer additional earners than those without children. This may indicate that family members other than core 148 members help in child rearing responsibilities, therefore, are less likely to join the paid labor force. Another alternative is that there may be no room in home for extra adults. However, when families with children are compared, the younger the age of children, i.e., the greater the responsibilities at home, the greater the chances for additional earners. For families with younger children, having additional earners may help out by increasing the family income and/or sharing the family living expenses. The above results help reveal the importance of family/household-based social capital in shaping work strategies of family/household members. In this study, I find that the involvement in self-employment activities and the employment of non-core members are influenced by the social relations within families and households, as indicated by the structure, composition, and stages in the life course. This is consistent with the embeddedness argument that -- economic actions are embedded in social relations (Granovetter 1985). Family/household work strategies are embedded in family/household social structures. An important finding here is that the effect of family/household social structure on family and household work strategy is not unique -- it varies by each household work strategy. A key implication of this finding is that family-based social capital can, on one lmnd enhance certain work strategies, while on the other hand, it can inhibit other types of work strategies. Families and households are interconnected and embedded with other social institutions in their community. The child care facilities, the schools, businesses, the extended families, a network of fiiends and co-workers and other institutions affect somehow the survival and adaptability of families and households to social and 149 economic changes. I proposed that family and household social ties to the labor market area affect work strategies. (P12). I hypothesized that family/household residential mobility decreases the likelihood of self-employment and having additional earners. The results indicate that: Family/household residential mobility decreases the likelihood of self- employment and increases the likelihood of additional earners when all other variables are controlled. People who live where they grew up are expected to have the strongest ties and social capital. Those who have moved will have weaker ties. The more the recent move, the weaker the ties, since it takes time to build up social capital in one’s new location (Hagan et al. 1996). Also, family and household residential mobility means that relationships are disrupted, thereby reducing the social capital in the community. Therefore, the lower the chances for self-employment. This may be associated with loss of exchange networks with kin, friends and neighbors, memberships in clubs and other community organizations, information on jobs and businesses’ opportunities that are crucial for starting and maintaining a business of your own In contrast, those who have not moved, i.e., who have stayed longer in their community are more attached to their communities and, thus, more likely to have a self-employed member. It was expected that family residential mobility would decrease the likelihood of additional earners. Families with greater ties to their communities tend to have more friends and connections. Thus, they may possess important information for job opportunities or may serve as an employment referral system to their members. However, the effect of residential mobility on additional earners is found positive. This 150 finding suggests more research that would account for community net migration, i.e., an account for both in migrants and out migrants. People may move without losing their social capital, particularly, if they move within the same community or move to a community in which they have loved ones or fiiends. More importantly, what needed most is to know what ties and of what kind people have in a community. Both models of self-employment and additional earners control the economic capital of farrrilies and households. I find the importance of economic capital in shaping family/household work strategies. Families and households with greater economic capital are likely to have a self-employment business. They have investments to do so. More important, self-employment, and in general informal activities, varies by social class (Beneria and Roldan 1987). Nelson (1999) also shows that among “good” job households, informal economic activity most often took the form of an entrepreneur business clearly distinguished from, and subordinate to, regular work. In contrast, he finds that the members of “bad” job households are unable to develop independent, ongoing entrepreneurial businesses. Members of“ ” job households explicitly stated that they lacked sufficient capital to invest in the development of an on-the-side business. Instead, he finds that the later households are more likely, instead of ongoing entrepreneurial moonlighting, to pick up additional wage work Some members of the “bad” job households have the opportunity to bring in income through the independent sale of goods and services but on a very casual basis (p30-31). The study of self- employment requires to take into consideration social class. This would require to fully detail the different types of self-employment and understand the mechanisms through 151 which they relate to social classes. While households with more economic capital may be self-employed in addition to other activities, those with less income may be self- employed for survival or need to complement their incomes. Members of families and households with greater economic capital are also likely to have additional earners. Additional members of “high” class households, i.e., those with greater economic capital, especially adult children, may enter the labor force probably not to contribute to the family income but for personal needs such as going to movies or buying things on their own The other alternative is that parents may initiate and socialize them to the world of paid labor. Consistent with the embeddedness/capital perspective, it is argued that family and household work strategies vary depending on the family/household social structure. Besides the family/household structure, composition, life course stages, and economic capital, a series of control variables was introduced in the analysis to account for the variation in work strategies of families and households. The results indicate that Families and households with higher education are more likely to have a self- employed member. However, they are less likely to have additional earners. Minority families are less likely than non-Hispanic White families to have a self- employed member. They are also less likely to have additional earners. It is expected that families with higher education would have greater likelihood of self-employment They not only have the skills to manage their own businesses but also they can have access to investment money to start a small business than those with lower education It was expected that families with higher education would be less likely to have additional earners. Families with lower education may have additional earners to 152 complement their income while those with higher education may have additional earners for other reasons other than survival, most notably involving adult children in the paid labor force. This again indicates a class difference between families and households in the likelihood of self-employment and additional earners. As also expected, minority families are less likely than non-Hispanic White families to have a self-employed member. The lower likelihood of self-employment could be due to lower investments to start a self-employment business. Families and households work strategies not only depend on family and household factors but also on factors of labor market areas in which they live. From a social embeddedness perspective, family and household work strategies depend not only on family/household social structure but also on larger social context in which they are linked 51mm 521 tbs Results in 8218115111 in 1hr. 89521191111122 W Economic restructuring has had a devastating impact on the economies of places, especially those in anal areas. From a restructuring perspective and its associated uneven development, it is expected that metropolitan and non-metropolitan areas have different opportunities. Life in non-metropolitan areas is different from that of metropolitan areas for not only economic reasons, but also social, cultural, and historic ones. I proposed that non-metropolitan/metrOpolitan residential location determines work strategies (P21). As I hypothesized, I found that: Families living in non-metropolitan areas are more likely than those in metropolitan to have a self-employed member but less likely to have additional earners. 153 Non-metropolitan families and households face a wage/income structure that affords them low earnings and in which they have limited external resources to improve their financial situation Any work alternative that may provide additional sources of income is important. The main sources of income in non-metropolitan areas have been paid employment, public assistance, but also kin supports. Strong reciprocal networks of information and trust that characterize rural places may explain the greater prevalence of self-employment However, the lack of opportunity structure in non-metropolitan areas, compared to metropolitan areas, may explain the lower likelihood of families and household to have additional earners. One possible explanation for the differences between metropolitan and non- metropolitan areas in terms of work strategies is the opportunity structure of labor market areas in which families and households are located It was proposed that industrial structure of a labor market area affects work strategies (p22). It was hypothesized that the greater the prevalence of core industries in a labor market area the lower the likelihood of self-employment To account for the non-metropolitan/metropolitan differential in opportunity structure, it was hypothesized that the differences in self- employment between non-metropolitan and metropolitan areas are primarily due to their differences in the proportion of core industries. The following results were found: The greater the prevalence of core industries in a labor market area the lower the chances for self-employment. Families living in non-metropolitan areas are far less likely than their metropolitan counterparts to have a self-employed member if they have access to good jobs. 154 From the restructuring perspective, it was also hypothesized that the greater the prevalence of core industries in a labor market area the greater the likelihood of additional earners. The hypothesis is supported and I find that: The likelihood of having additional earners is enhanced by the quality of jobs in a labor market area. Other labor market area factors, other than the industrial structure, significantly influence the likelihood of families and households to have a self-employed member and/or have non-core members participate in the labor force. I proposed that residential stability of labor market area affects work strategies. (P23). Contrary to what I expected, I find that: The labor market area residential stability decreases the likelihood of self- employment. Finally, the economic disadvantage /inequality of a labor market area affects work strategies. (PZ4) . Hypotheses related to this proposition were tested and the following results are found: The labor market area economic disadvantage/ inequality decreases the likelihood of self-employment and additional earners. Families and households living in labor market areas that are disadvantaged may opt for self-employment or rely on the income of additional earners for survival reasons, i.e., in order to complement their poor incomes. In general those areas are characterized by low paying jobs and high unemployment rates. That may explain the lower likelihood of additional earners in those labor market areas. The likelihood of self-employment in disadvantaged areas may be limited because of the lack not only of financial resources to 155 start a small business but also because of the lack of people who would make those small businesses profitable, i.e., people with steady and relatively high incomes. More importantly, residents of disadvantaged areas tend to be isolated from other business communities outside their niche, thus may be less likely to start their own businesses. Findingsare summarized intable 11. Table 11. Summary of Findings Proposition ” *— ’ ‘ Fidimgs if Self-employment Additional Earners Family/household Level P 11. Family and household Married families are more Married families are less structure determines work likely than “non-married” likely than “non-married” strategies. families and households to families and households to have a self-employed have a self-employed member. member. Cohabiting households are Cohabiting households are more likely than single- less likely than single- headed households to headed households to have a self-employed have a self-employed member. member. F emale-headed Female-headed households are less likely households are less likely than male-headed than male-headed families households to have a self- and households to have a employed member. self-employed member. Families with children under 18 years of age are more likely than those without children to have a self-employed member. 156 Proposition Findings Self-employment Additional Earners Families with preschool children are more likely than those with school-age and/or adolescent children to have a self-employed member. Families with school-age children are less likely than those with adolescent children to have a self- employed member. P12. Family and household F arnily/household Family/household social ties to the labor residential mobility residential mobility market area affect work decreases the likelihood increases the likelihood of strategies. of self-employment additional earners. Level 2: Labor market area P21. Non-metropolitan- Non-metropolitan families Non-metropolitan families metropolitan residential and households are more and households are less location determines work likely than those in likely than those in strategies. metropolitan areas to have metropolitan areas to have a self-employed member. additional earners. P22. Industrial structure of The greater the prevalence The greater the prevalence a labor market area affects of core industries in a of core industries in a work strategies. labor market area the labor market area the lower the likelihood of greater the likelihood of self-employment. additional earners. Families and households living in non-metropolitan labor market areas with a greater proportion of core industries are less likely than their metropolitan counterparts to have a self-employed member. 157 Proposition Findings Self-employment Additional Earners P23 - Residential stability of Residential stability of a a labor market area affects labor market area is work strategies. negatively related to self- employment. P24 - Economic Economic disadvantage/ Economic disadvantage/ disadvantage /inequality of inequality of a labor inequality of a labor a labor market area affects market area is negatively market area is negatively work strategies. related to self- related to additional employment. eamers. E 1' I 1' . The findings for this study can be used to understand families and household differences in their responses to environment surrounding them. The changes in the structure of industries and occupations and associated management decisions have different effects on families’ economic resources and on the way they deal with such Changes. Limited opportunities in some labor market areas or opportunities that influence available options for families. Families in metropolitan areas, for example, differ from their non—metropolitan counterparts in employment opportunities and in other alternative options for survival and/or adaptations to changes in their opportunities. With the continuing changes in the global economy, some labor market areas will benefit fi'om new job opportunities while some others will continue to loose jobs and thus face periods of unemployment and layoffs and rising poverty and income inequalities. Policy aimed at helping families dealing with such macro changes should take into account differences 158 in Opportunities in a labor market area, not only in terms of number of jobs but also in terms of the quality of such jobs. In addition, the findings in this study reveal that the labor market residential stability influences family and household work strategies. However, a closer look of the ways in which the labor market area residential stability and family and household work strategies are related is finther needed If stronger and dense relationships within a community and between communities help families in dealing with the changes in their environment, a recommendable policy would be what Putnam has indicated, that is, to reinvest and reinforce social capital. Family and household strategies not only depend on larger social context in which they are embedded but also on family/household social structure. The findings in this Study indicate that families and household work strategies differ depending on the family/household structure. With the continuing changes in family structures, especially Single-headed and cohabiting families, policy aimed at helping families should take into consideration such changes given their differences in needs and in resources. In addition, the findings of this study indicate that families with children, especially young children, have different strategies than other families. They are more likely to have a self- employed member, probably because having young children is a constraining factor for other opportunities. In addition, they are less likely to have additional members in the labor force because often adult members, especially adult children, may help in family child rearing responsibilities. Any policy that would help alleviate the burden of child care, especially in early ages, would allow parents to adopt strategies they normally 159 would not think of if they had not access to good quality and affordable child care in their vicinity. Therefore, policies that reinforce work-family relations, especially in this ongoing period of economic restructuring, would give families many alternative work strategies. The findings of this study suggest that fanrily and household work strategies are function of their economic capital. The income disparities between families no doubt has social and economic consequences. At the same time other social statuses such as race or immigrant statuses affect family and household work strategies. These findings may require policies not only help reduce the gap between the rich and the poor but may increase alternative choices of the poor and other disadvantaged groups, in situations such us the relocation of a major employer, a cut in welfare subsidies, a loss or decrease of an important source of income, or a loss of family social capital. In summary, the findings have policy implications for rural development and for families. First, the labor market area social structrue affects the ways families and households work strategies. The quality of jobs in an area is an important factor that help explaining the differences between labor market areas in family/household work strategies. In addition, the metropolitan and non-metropolitan differential effect on family and household work strategies should be taken into consideration in policies. Second, the family/household social structure influences whatever families do to deal with surrounding environments. Differences in structure, family-based social capital, and class should be taken into consideration and translated into adequate policies that help families to improve their well-being The connection of families and households 160 with different institutions in their communities, including the world of work but also other institutions, will enhance the abilities of families and households to adapt to change in one of those the institutions. While this study addresses the issue of family and household work strategies, it also has several limitations. One important limitation of this study is the use of secondary data analysis. Qualitative data would inform more about the different strategies of families and households given variable social context in which they are embedded Families and households have different ways of adapting to whatever changes they are confronted with, given their different resources (economic, social, cultural). An in-depth interview on household work strategies would give more details of Wimt exactly people do in their small businesses and whose household member is involved and helping out. More importantly, what is defined as an adaptive strategy for one family/household may not be an adaptive strategy at all for another family/household As Wolf (1992: 17) indicates, "if people or households were followed through time and qualitative data were gathered, there would be more awareness of how inaction may at times be part of a strategy.” Also, Garrett et al. (1993) mention that "strategies are rarely defined as deliberate, self-conscious formulations of householders." This study informs about two related family and household work strategies, self- employment and labor force participation of non-core members. However, there are other related work strategies that need attention for future research With the restructuring of the economy, more and more firms are relocating their businesses in 161 places where they can gain some profits, leaving some families in critical situations. One of the strategies for families is to commute, sometimes traveling long distances for better job opportunities in other places. Yet the competition with companies abroad will continue to influence the decisions for relocation of businesses abroad in search of cheap labor and away from unions. Another household strategy in response to economic restructuring is moonlighting. Stinson Jr. (1990) indicates that the number of multiple job holders increased by 52 percent from 1980 to 1989 (p3). Nelson (1999) also finds that in addition to individuals’ regular jobs, at least one member of their respondents’ household engages in a second activity —— such a second activity is either an independent business selling goods and services for cash or in-kind exchanges or a second-waged activity (1328). Another related household strategy is “shift” work With the increasing number 0f dual-earner families, more research is needed to analyze non-standard work schedules including working evenings, nights, week-ends or rotating schedules and how these emerging patterns of work affect family life (see Pleck and Staines 1985; Presser 1987, 1984). Future research should analyze different sources of income and support for families. Future research is also needed to further analyze how gender affects family and household work strategies. Previous studies on informal and industrial homework activities also stress the importance of gender in explaining men and women’s differences in such activities (Roldan 1987; Fernandez-Kelly 1989; Bose 1987; Gringeri 1990,1994; Dangler 1996; Boris 1994; Boris and Prfigl 1996, Nelson 1999). More 162 research is still needed to account for the effect of gender on self-employment. Self- employment is a productive activity that depends on capital and is linked to external economic forces, which are gendered in their structure. The consideration of gender relations is important because traditionally women, more than men, bear the responsibility of child care and other household chores (Hochschild 1989, 1997) and sex segregation remains a persistence feature of the formal labor force (Reskin and Padavic1993). The multilevel-modeling approach no doubt has contributed significantly to the understanding of how labor market area characteristics in conjunction with family and household factors affect family and household work strategies. More analysis is still needed to account for the analysis of multilevel data and the specification of structural equations within and between labor market areas. See Kaplan and Elliott (1997) and Muthén (1994) for a review of such types of models. These models would allow to capture the complex social relations that exist within and between labor market areas. Finally, optimal measures ofsocial capital are further needed to better address the concept of social capital and how it influences work strategies. There are challenges associated with the use of social capital including its conceptualization and measurement Can social capital be measured? Social capital is a multidimensioml and a multilevel concept Many studies have identified useful proxies of social capital using ‘ indicators such as trust, solidarity, norms of reciprocity, networks, membership in associations or social groups, ethnic diversity, social mobility, civic involvement and community engagement, community ties, community integrity and social structure, the 163 presence of long-term elderly members in a community and social interactions in a community etc. Social capital is not only a multidimensional concept but also it is measured at different levels of analysis. At the family level, social capital is conceptualized as a resource for its members. At the community level, social interactions among neighbors, fiiends, and groups generate social capital. At the firm level or any other organization, social capital is used to build and sustain efficient organizations including relations of trust among its members and working together for the common goal. Ethnic groups and other social groups provide examples of people who share common values and culture can band together for the mutual benefits. At the society level, a civic society, the functioning of the state and/or government are central to the welfare of society. Thus, measuring social capital is challenging because it is difficult to measure things like trust or community civic engagement but it is possible when someone identifies “good” proxies. Any measurement of a concept depends on ones’ theory and conceptual framework. In this study, social capital is used at both the family and labor market area levels. Family and household structure and family ties to communities as measured by the family residential moves, are used as a proxy for social capital. Both family structure and residential mobility were used as proxies for social capital (Coleman (1988); Hagan et al. 1996; McLanahan and Sanderfur 1994). However, they are crude measures of social capital. Further research is still nwded to show how different family structures are endowed with different social capital and how they affect different work strategies. In this study, I find that, after controlling for other family characteristics, married families 164 are more likely than other families to have a self-employment member, cohabiting households have a greater likelihood of self-employment than single-headed households, and male—headed families are more likely than female-headed households to have a self- employment member. I also find that single-headed families are more likely than married and cohabiting families to have additional earners. This implies that intra- family / household relations have tremendous effect on self-employment and employment of its members. More research is needed to answer the following questions: What kinds of social ties exist in these different family structures? To what extent these social ties generate social capital that can be used by family members to do what they do best for their families? Families are sources of information for employment opportunities. They are also sources of investment money for family business. Family members are also used as cheap labor in family businesses. Family members are often the primary clients of family businesses’ products. Social capital resides in social relations within families and households that are beyond the presence or not of spouses or partners. It exists in the social interactions between family members, in their understandings and working together, in their social solidarity and trust, and in the long-run stability of the relationship. An emphasis here is not just on the presence of social ties but also the quality of those ties. An account of these social relations within families may explain why married families have greater social capital. A proper measure of social capital at the family level ought to account for these intra-family relations. 165 Families are not isolated entities, they are well connected to communities in which they live in, particularly to institutions whin those communities. According to Coleman (1988), Hagan et al. (1996), McLanahan and Sanderfur (1994), leaving a community tends to destroy established bonds, thus depriving family and children a major source of social capital (see also Portes (1998). However, family-based social capital tends to compensate the loss of community social capital due to family moves (Hagan et al. 1996). This implies the need to analyze social capital at both family and community level. Family moves, however, are not very a good proxy for social capital. Some families may move out from a community and gain community social capital, especially, if they move in a community in which they have fiiends and relatives, a community in which they grew up and consider home. Further research needs to distinguish family moves within the same community from moves outside the community. Moves within the same area may not involve changing schools or other losses of social capital (Hagan et al. 1996). Also, families may stay longer in a community and have greater social capital due to friends and relatives nearby but remain isolated to the rest of the community, thus be deprived of sources of vital information and employment opportunities (Sack 1974; Wilson (1987, 1996). Thus, what is needed is a model that accounts for different social networks in community and how those networks are linked or eventually isolated Another limitation of this study is the use of cross-sectional data. Longitudinal data would enrich the analysis of household work strategies. It would allow to analyze family and work strategies overtime. Although, the macro-micro contextual framework 166 enriches the analysis of household work strategies, it is yet limited It would be enriching to know when a particular strategy occurs and whether it is due to changes in family/household structure or to macro changes in the environment surrounding the family/household including not only economic but also political and social changes. Because social capital inheres in social structures and that social structures change overtime, a dynamic analysis of social capital is needed to fully understand its effects on work strategies in a given period The life course perspective (Elder 1994; Hagan et al. 1996) considers the long-term implications of the interdependency of family members’ lives, especially those of parents and children, and of the ways in which parental involvement in children’s lives can, in addition to its main efi‘ects, also condition or buffer the effects of losses of social capital a result of social change such as a family move. This is true, mutatis mutandis, of long—term relationships that can help families to deal with stressful events or adopt any strategy as a result ofchanges in other environments surrounding the family. Therefore, a longitudinal study is needed to not only capture when specific events, such as work strategies, occur, but also to capture the effect of time-varying predictors such as social capital. 167 REFERENCES Amsden, Alice H. 1989. Asian ’s Next Giant: South Korea and Late Industrialization, New York: Oxford University Press. Anheier, Helmut K., Jttrgen Gerhards, and Frank P. Romo. 1995. “Forms of Capital and Social Structure in Cultural Fields: Examining Bourdieu’s Social Topography.” American Journal of Sociology, 100: 859-903. Baca Zinn, Maxine and D. Stanley Eitzen. 1996. Diversity in Families. Fourth Edition. New York: Harper Collins College Publishers. Bane, Mary Jo, and David T. Ellwood 1986. “Slipping Into and Out of Poverty: The Dynamics of Spells.” The Journal of Human Resources, 21: 1-23. Baron, James N. and William T. Bielby 1980. “Bringing the Firms Back in: Stratification, Segmentation, and the Organization of Work” American Sociological Review, 45(5): 737-765. Bates, Robert. 1989. Beyond the Miracle of the Market: The Political Economy of the Agrarian Development in Kenya, Cambridge: Cambridge University Press. Becker, Gary S. 1975. Human Capital: A Theoretical and Empirical Analysis with Special Reference to Education, 2nd ed. , NY. Columbia University Press. . 1981. A Treatise on the Family. Cambridge, MA: Harvard University Press, Beneria, Lourdes and Martha Roldan 1987. The Crossroads of Class and Gender: Industrial Homework, Subcontracting, and Household Dynamics in Mexico City. Chicago: University of Chicago Press. Bloomquist, Leonard E. 1988. “Performance in the Rmal Manufacttning Sector.” In David L. Brown, J. N Reid, J. Bluestone, D. A McGranahan, and S. M Maxie (ed), Rural Economic Development in the 1980's: Prospects for the Future: A Summary. Rural Development Research Report no. 69. Washington DC. ERS, US. Department of Agriculture. Bloomquist Leonard E., Christina Gringeri, Donald Tomaskovic-Devey, and Cynthia Truelove. 1993. "Work Structures and Rural Poverty." Pp. 68-105. In Persistent Poverty in Rural America. Ed Rural Sociological Society Task F orce on Persistent Rural Poverty. Boulder, Colorado: Westview Press, 168 Bokemeier, Janet L. 1997. “Rediscovering Families and Households: Restructuring Rural Society and Rural Sociology.” Presidential Address for the 1996 Annual Meeting of the Rural Sociological Society. Rural Sociology, Vol. 62(1): 1-20. Bokemeier, Janet L., and Jean Kayitsinga 1997 'Underemployment in Rural America." In Gary A. Goreham (Ed) Encyclopedia of Rural America: The Land and People Vol. 2: Santa Barbara, Denver, and Oxford: ABC-CLIO Boris Eileen and Elisabeth Prugl 1996. Homeworkers in Global Perspective: Invisible No More. New York: Routledge. Boris, Eileen. 1994. Home to Work: Motherhood and the Politics of Industrial Homework in te United States. Cambridge: Cambridge University Press. Boris, Eileen and Peter Bardaglio. 1987. “Gender, Race, and Class: The Impact of the State on the Family and the Economy.” Pp. 132-151. In Naomi Gerstel and Harriet Engel Gross (eds). Families and Work. Philadelphia: Temple University Press. Borjas, George J. 1986. The Self-employment Experience of [migrants (NBER Working Paper No. 194). Cambridge, MA: National Bureau of Economic Research Bose, Christine. 1987. “Dual Spheres.” Pp. 267-285. In B. Hess and M. Ferree. Ed Analyzing Gender: A Handbook of Social Science Research, Thousands Oaks, CA: Sage. Bourdieu, Pierre. 1998. Practical Reason: 0n the Theory of Action. Stanford, CA: Stanford University Press. 1996. “On the Family as a Realized Category.” Theory, Culture, and Society 13(3): 19-26. 1993. Sociology in Question. London: Sage. 1990[1980]. The Logic of Practice. Stanford, CA: Stanford University Press. 1986. “The Forms of Capital.” Pp. 241-258. In J. Richardson ed Handbook of Theory and Research for the Sociology of Education, Westport, CT: Greenwood 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press. 169 Bourdieu, Pierre and J can-Claude Passeron. 1990[1970]. Reproduction in Education, Society and Culture. London; Newbury Park, CA: Sage. Boyd, Robert L. 1991. “A Contextual Analysis of Black Self-Employment in Large Metropolitan Areas, 1970-1980.” Social Forces, 70(2): 409-29. Burt, Ronald 1992. Structural Holes: The Social Structure of Competition. Cambridge, MA: Havard University Press. Breiger, Ronald L. 1995. “ Social Structure and the Phenomenology of Attainment.” Annual Review of Sociology, 21: 1 15-136. Brown, David L. and Thomas Hirschl. 1995. “Household Poverty in Rrual and Metropolitan Core Areas of the United States.” Rural Sociology, 60: 44-66. Brown, David L. and Kenneth L. Deavers. 1987. "Rural Change and the Rural Economic Policy Agenda for the 1980's" Pp. 1-28 in Rural Economic Development in the 1980's, edited by David Brown, J. Norman Reid, Herman Bluestone, David A McGranahan, and Sara M Mazie. Rural Development Research Report No. 69. Washington, D.C.:U.S. Department of Agriculture, Economic Research Service. Bryk, Anthony S. and Stephen W. Raudenbush 1992. Hierarchical Linear Models: Applications and Data Analysis Methods. Vol. 1. Advanced Quantitative Techniques in the Social Sciences Series, ed Sage. Newbury Park, London, New Delhi: Sage Publications. Bryk, Anthony 8., Stephen W. Raudenbush, and Richard T. Congdon, Jr. 1996. HLM: Hierarchical Linear and Nonlinear Modeling with the HLM/ZL and HLM/3L Programs. Chicago, IL: Scientific Software International, Inc. Campbell, Rex R, John C. Spencer, and Ravindra G. Amonker. 1993. “The Reported and Unreported Missouri Ozarks: Adaptive Strategies of the People Left Behind.” Pp. 30-52. In Thomas A Lyson and William F alk ed Forgotten Places: Uneven Development in Rural America, Kansas: The University Press of Kansas. Carr, Deborah 1996. “Two Paths to Self-Employment?: Women’s and Men’s Self- Employment in the United States, 1980.” Work and Occupations, 23(1): 26-53. Cheal, David 1991. Family and the State of Theory. Toronto: University of Toronto Press. 170 . 1993. “Unity and Difference in Postmodern Families.” Journal of Family Issues,V14(1):5-19. Christensen, Kathleen E. 1987. “Women, Families, and Home-Based Employment.” Pp. 478-490. In Naomi Gerstel and Harriet Engel Gross (ed) Families and Work. Philadelphia: Temple University Press. Clay, Daniel C., and Harry K. Schwarzweller. 1991. Household Survival Strategies. Vol. 5. Rural Sociology and Development, Greenwich: JAI Press. Colclough, Glenna and Charles M. Tolbert. 1993. “Divisions of Labor and Inequality in High-Tech Centers.” Pp 143-163. In Joachim Singelrnann and Forrest A. Deseran (eds.). Inequality in Labor Market Areas. Boulder: Westview Press. . 1992. Work in the Fast Lane: Flexibility, Divisions of Labor, and Inequality in H igh- Tech Industries. Albany: State University of New York Press. Coleman, James S., 1988. “Social Capital in the Creation of Human Capital.” American Journal of Sociology, 94: 895-8120. . 1990. Foundations of Social Theory. Cambridge, MA: Havard University Press. Collins, Jane L. 1990. “Unwaged Labor in Comparative Perspective: Recent Theories and Unanswered Questions.” Pp 3-24 in Martha Gimenez and Jane L. Collins (ed) Domestic Labor and Self-Employment within Capitalism . Albany NY.: State University of New York Press. Cromartie, John B. 1993. “Leaving the Country Side: Young Adults Follow Complex Migration Patterns." Rural Development Perspectives 8:22-27. Crow, G. 1989. "The Use and Concept of 'Strategy' in Recent Sociological Literature." Sociology, 2321-24. Curtis, Richard F. 1986. “Household and Family in Theory and Inequality.” American Sociological Review 51: 168-183. Dangler, Jamie Faricellia. 1996. Hidden in the Home: The Role of Waged Homework in the Modern World Economy. Albany: State University of New York Press. Davidson, Andrew. 1991. "Rethinking Household Strategies" Pp 11-28. In Daniel C. Clay and Harry K Schwarzweller (ed). Household Survival Strategies. Vol. 5. Rural Sociology and Development, Greenwich: J AI Press. 171 Deseran, Forest A, William W. Falk, and Pamela Jenkins. 1984. "Determinants of Earnings of Farm Families in the US.” Rural Sociology 49(2, Summer):210-229. Dill, Bonnie Thornton and Bruce B. Williams. 1992. "Race, Gender, and Poverty in the Rural South: African American Single Mothers." Pp 97-109. In Cynthia Duncan (ed) Rural Poverty in America. New York: New York Auburn House. Diprete, Thomas A, and Jerry D. Forristal 1994. “Multilevel Methods: Methods and Substance.” Annual Review of Sociology, 20: 331-57. Doeringer, Peter B. 1984. "Internal Labor Markets and Paterrmlism in Rural Areas." Pp. 272-289. In Paul Osterman (ed). Internal labor Markets. Cambridge, MA: The MIT Press. Duncan and Tickamyer. 1988. Poverty Research and Policy for Rural America. The American Sociologist, Fall: 243-259. Eardley, Tony and Anne Corden. 1996. Low-Income Self-Employment: Work, Benefits, and Living Standards. Aldershot, Brookfield USA, Hong Kong, Singapore, Sydney: Avebury. Easterlin, Richard. A. 1980. Birth and Fortune: The Impact of Numbers on Personal Welfare. New York: Basic Books. Elder, Glen H 1994. “Time, Human Agency, and Social Clumge: Perspectives on the Life Course.” Social Psychology Quarterly, 57:4-15. Elder, Glen H, and Avshalom Caspi. 1990. “Studying Lives in a Changing Society: Sociological and Personological Explorations." Pp. 201-247. In Murray, Henri Alexander and Albert 1. Rabin Studying Persons and Lives. New York: Springer Pub. Co. Ellwood, David T. 1988. Poor Support: Poverty in American Families. New York: Basic Books. England Paula. 1993. Theory on Gender/Feminism on Theory. New York: Aldine de Gruyter. Evans, Peter. 1995. Embedded Autonomy, Princeton: Princeton University Press. Falk, William and Thomas A. Lyson 1988. High Tech, Low Tech. No Tech. Albany: State University of New York Press. 172 Feldman, Shelly and Rick Welch 1995. “Feminist Knowledge Claims, Local Knowledge, and Gender Divisions of Agricultural Labor: Constructing a Successor Science. ” Rural Sociology 60: 23-43. Fernandez-Kelly, Patricia M. 1995. Social and Cultural Capital in the Ghetto: Implications for the Economic Sociology of Immigration.” Pp. 213-247. In Alejandro Portes(ed). The Economic Sociology of Immigration. New York: Russell Sage. 1994. “Towanda’s Triumph: Social and Cultural Capital in the Transition to Adulthood in Urban Ghetto.” International Journal of Urban and Regional Research, 1 8( l )2 88-2 12. 1989. “Power Surrendered, Power Restored: The Politics of Work and Family among Hispanic Garment Workers in California and F lorida.” Pp. 130-9. In Louise A Tilly and Patricia Guerin ed Women, Politics, and Change. New York: Russell Sage Foundation. F erree, Myra Marx 1990. “Beyond Separate Spheres: Feminism and Family Research.” Journal of Marriage and F amily, 52 (November): 866-84. Fink, Deborah. 1992. Agrarian Women: Wives and Mothers in Rural Nebraska 1880- 1940. Chapel Hill, NC: University of North Carolina Press. Fitchen, Janet M 1981. Poverty in Rural America: A Case Study. Boulder: Westview Press. --—--. 1991. Endangered Spaces, Enduring places: Change, Identity, and Survival in Rural America. Boulder, Colorado: Westview Press. ----—. 1995. “ Spatial Redistribution of Poverty Through Migration of Poor People to Depressed Rural Communities.” Rural Sociology, 60: 181-201. -—-—---. 1994. “Residential Mobility among the Rural Poor. ” Rural Sociology 59:416- 436. Flora , Cornelia B., and Jan L. Flora. 1993. “Entrepreneurial Social Infrastructure: A Necessary Ingredient.” AMVALS, AAPSS, 529:48-58. 173 Flora, Jan L. 1998. “Social Capital and Communities of Place.” Rural Sociology, 63(4):481-506. Gardner, J. M, and D. E. Hertz. 1992. “Working and Poor in 1990.” Monthly Labor Review, 115: 20-28. Garkovich, Lorraine, Janet L. Bokemeier, and Barbara Foote 1995. Harvest of Hope: Family F arming/F arming Families. Lexington, Kentucky: University of Kentucky Press. Garkovich, Lorraine and Janet L. Bokemeier. 1988. “Factors Associated with Women’s Attitudes toward Farming.” Pp. 120-140. In 0. Beaulieu (ed). Rural Crises in the South. Boulder: Westview Press. Garkovich, Lorraine. 1989. Population and Community in Rural America. New York: Greenwood Press. Garrett, Patricia et al. 1993. “Rural Families and Children in Poverty. ” Pp 230-58. In Persistent Poverty in Rural America ed. Rural Sociological Society Task Force on Persistent Rural Poverty. Boulder, San Francisco, Oxford: Westview Press. Gerstel, Naomi and Harriet E. Gross 1987 Families and Work. Philadelphia: Temple University Press. Gimenez, Martha E. 1990. “The Dialectics of Waged and Unwaged Work: Waged Work, Domestic Labor, and Household Survival in the United States." Pp 25-46 in Martha Gimenez Jane L. Collins (ed.) Domestic Labor and Self-Employment within Capitalism. Albany NY.: State University of New York Press. Glazer, Nona Y. 1990. Women ’s Paid and Unpaid Labor: The Work Transfer in Health Care and Retailing. Philaderphia, PA: Temple University Press. Goldscheider, Frances K and Linda J. Waite. 1991. New Families, No Families?: The T ransformation of the American Home. Berkeley: University of California Press. Goldstein, Harvey. 1995. Multilevel Statistical Models, Second Edition. London: Edward Arnold Gorham, Lucy. 1992. “The Growing Problem of Low Earnings in Rural Areas.” Pp: 21- 39. In Cynthia M. Duncan ed. Rural Poverty in America. New York: Auburn House. 174 Granovetter, Mark S. 1985. “Economic Action and Social Structure: The Problem of Embeddedness. ” American Journal of Sociology 91: 481-510. 1974. Getting a Job: A Study of Contacts and Careers. Cambridge, MA: Harvard University Press. Gringeri, Christina E. 1994. Getting By: Women Homeworkers and Rural Economic Development. Lawrence, KS: University Press of Kansas. Hagan, John, Ross MacMillan, and Blair Wheaton. 1996. “New Kid in Town: Social Capital and the Life Course Effects of Family Migration on Children.” American Sociological Review, 61(3): 368-3 85. Hanson, Susan, and Geraldine Pratt. 1993. “Dynamic Dependencies: A Geographic Investigation of Local Labor Markets. ”Economic Geography, v68(4): 373-405. Hao L. 1994. Kin Support, Welfare, and Out-of- Wedlock Mothers. New York: Garland. Hareven, Tamara K. 1991. “The History of the Family and the Complexity of Social Change.” American Historical Review, 96: 95-124. ——---. 1982. Family Time and Industrial Time. Cambridge, England: Cambridge University Press. Hartrnann, Heidi I. 1987. “Changes in Women's Economic and Family Roles in Post-World War 11 United States.” Pp 33-64 in Lourdes Beneria and Catherine R Stimpson (ed.) Women, Households, and the Economy. Rutgers University Press. . 1981. "The family as the Locus of Gender, Class, and Political Struggle: The Example of Housework" Signs: Journal of Women in Culture and Society 6: 366- 394. Harvey L., David 1989. Potter Addition: Poverty, Family, and Kinship in Heartland Community. New York: Aldine de Gruyter. Hill, R. 1970. Family Development in Three Generations. Cambridge, England: Schenkman. Hochschild, Arlie R 1997. The T me Bind: When Work Becomes Home and Home Becomes Work. New York: Metropolitan Books. Hochschild, Arlie R with Anne Machung 1989. The Second Shift: Working Parents and the Revolution at Home. New York, New York: Viking. 175 Hondagneu-Sotelo, Pierrette. 1992. “Overcoming Patriarchal Constraints: The Reconstruction of Gender Relations Among Mexican Immigrant Women and Men.” Gender in Society, 6(3): 393-415. Holstein and Gubrium 1995. “Deprivatization and the Construction of Domestic Life. ” Journal of Marriage and Family, 57: 894-908. Hoppe, Robert. 1993. "Poverty in Rural America, Trends and Demographic Characteristics." Pp 20-38. In Persistent Poverty in Rural America. Ed Rural Sociological Society Task Force on Persistent Rural Poverty Boulder, Colorado: Westview Press. Hossfeld, Karen 1990. “Their Logic Against Them: Contradictions in Sex, Race, and Class in Sillicon Valley. ” Pp. 149-178. In Kathryn B. Ward (ed). Women Workers and Global Restructuring. Ithaca, New Yorksz Press. Jacobs, Jane. 1961. The Life and Death of Great American Cities. New York: Random House 1961. Jensen, Leif and David J. Eggebeen 1994. Nonmetropolitan Poor Children and the Reliance on Public Assistance. Rural Sociology 59 (1): P45-65. Jensen, Leif and Marta Tienda 1989. “Non-Metropolitan Minority Families in the United States: Trends in Racial and Economic Stratification, 1959-1986.” Rural Sociology 54:pp. Jobes, Patrick G, William F. Stinner, and John M Wardwell. 1992. “Paradigm Shift in Migration Explanation.” Pp. 1-31. In Jobes, Patrick G, William F. Stinner, and John M. Wardwell Lanlmn Ed. Community, Society, and Migration: Non- economic Migration in America, MD: University Press of America Jones, J ., and J. Kodras 1990. “Restructured Regions and Families: The Feminization of Poverty in the US. " Annals of the Association of American Geographers 80: 163-83. Kalleberg, Arne L. 1989. “Linking Macro and Micro Levels: Bringing the Workers Back into the Sociology of Work" Social Forces, v67(3): 582-594. Kalleberg, A, and I. Berg. 1987. Work and Industry. New York: Plenum. . 1988. “Work Structures and Markets: An Analytic Framework.” Pp. 3-17. In G. Farkas and P. England (eds.). Industry, F inns, and Jobs. New Yorlc Plenum. 176 Kaplan, David and Pamela R Elliott. 1997. “A Didactic Example of Multilevel Structural Equation Modeling Applicable to the Study of Social Organizations.” Structural Equation Modeling, 4(1): 1-24. Kassab, C. 1992. Income and Inequality: The Role of the Service Sector in the Changing Distribution of Income. NY: Greenwood Press. Kertzer, D. 1., and Hogan D. P. 1989. Family, Political Economy, and Demographic Change: The Transformation of Life in Casalecchio, Italy, 1861-1921. Madison, Wisconsin: University of Wisconsin Press. Killian, Molly Sizer, and Thomas F. Hady. 1988. The Economic Performance of Rural Labor Markets.” Pp. 181-200. In David L. Brown, J. Norman Reid, Herman Bluestone, David A McGranahan and Sara A Mazie (eds.). Rural Economic Development in the 1980's: Prospects for the Future. Agricultural and Rural Economy Division, ERS. US. Department of Agriculture. Rural Development Research Report no. 69. Washington, D. C.: US. Government Printing Oflice. Killian, Molly Sizer, and Charles M. Tolbert. 1993. “Mapping Social and Economic Space: The Delineation of Local Labor Markets in the United States.” Pp. 69-82. In Joachim Singelmann and Forrest A Deseran (eds. ). Inequality in Labor Market Areas. Boulder: Westview Press. Kreft, Ita G.G., Leeuw, Jan de Leeden, Rien van der. 1994. “Review of Five Multilevel Analysis Programs: BMDP-SV, GENMOD, HLM, ML3, VARCL.” The American Statistician, v48, n4, p324(12). Lasch, Christopher. 1977. Haven in a Heartless World: The Family Besieged. New York: BasicBooks. Leontief, Wassily, and Duchin Faye. 1986. The Future of A utomation on Workers. New York: Oxford University Press. Lichter, Daniel T. and David J. Eggebeen. 1992. “Child Poverty and the Changing Rural Family." Rural Sociology 57 (2): 151-172. Lichter, Daniel T., and Janice Costanzo. 1987. Non-metropolitan Underemployment and Labor-Force Composition. ” Rural Sociology 52(3): 329-344. Lichter, Daniel T. and David Landry 1991 "Labor Force Transition and Underemployment: The Stratification of Male and Female Workers." Research in Social Stratification and Mobility 10: 63-87. 177 Lichter, Daniel T., Gail M Johnson, and Diane K McLaughlin 1994. “Changing Linkages between Work and Poverty in Rural America." Rural Sociology 59: 395-4 15. Lichter, Daniel T., and Diane K McLaughlin 1995. Changing Economic Opportunities, Family Structure, and Poverty in Rural Areas." Rural Sociology, 60(4): 688-706. Lichter, Daniel T. 1989. "The Underemployment of American Rural Women: Prevalence, Trends and Spatial Inequality." Journal of Rural Studies, 5(2), 199-208. Lie, John. 1997. “Sociology of Markets.” Annual Review of Sociology, 23:341-60. Light, Ivan. 1984. “Immigrant and Ethnic Enterprise in North America” Ethnic and Racial Studies 7: 195-216. Light, Ivan and Edna Bonacich. 1988. Immigrant Entrepreneurs: Koreans in Los Angeles 1965-1982. Berkeley, CA: University of California Press. Lobao, Linda M 1990. Locality and Inequality: Farm and Industry Structures and Socioeconomic Conditions. SUNY Series on the New Inequalities, ed A Gary Dworkin New York: The State University of New York Press. Lobao, Linda M. and Michael D. Schulman 1991. “Farming Patterns, Rural Restructuring and Poverty: A Comparative Regiorml Analysis.” Rural Sociology, 56(4):565-602. Longford, Nicholas.T. 1993. Random Coefiicient Models. Oxford (England): Clarendon Press; New York: Oxford University Press. Loury, Glenn 1992. “The Economics of Discrimination: Getting to the Core of the Problem.” Harvard Journal for Afi'ican American Public Policy Lyson, Thomas A and William W. Falk 1993. Forgotten Places: Uneven Development in Rural America. University Press of Kansas. Lyson, Thomas A, William W. Falk, Mark Henry, Joan Hickey, and Mildred Warner. 1993. “ Spatial Location of Economic Activities, Uneven Development, and Rural Poverty.” Pp. 106-135. In Persistent Poverty in Rural America Ed Rural Sociological Society Task Force on Persistent Rural Poverty. Boulder, San Francisco, Oxford: Westview Press. 178 McGranahan, David A 1988. Rural Workers in the National Economy." Pp 29-47. In David L.Brown, J. Norman Reid, Herman Bluestone, David A McGranahan, and Sara Mazie (eds.). Rural Economic Development in the 1980's: Prospects for the Future. McLanahan, Sara and Gary Sanderful 1994. Growing up with a Single Parent: What Hurts, What helps. Cambridge, Massachussets, London, England: Havard University Press. McLanahan, Sara, Karen Booth 1996. Mother—Only Families: Problems, Prospects, and Politics.” Journal of Marriage and the Family, 51: 557-80. McLaughlin, Diane K and Lauri Perman. 1991. “Returns vs. Endowments in the Earnings Attainment Process for Metropolitan and Nonmetropolitan Men and Women. ” Rural Sociology 56 (3): 339-365. McLaughlin, Diane K, and Carolyn Sachs. 1988. “Poverty in Female-Headed Households: Residential Differences. " Rural Sociology 53(3): 287-306. McMichael, Philip. 1996. “Globalization:Myths and Realities.” Rural Sociology, 61(1):25-55. Mingione, Enzo. 1991. Fragmented Societies: A Sociology of Economic Life beyond the Market Paradigm. Translated by Paul Goodrich. Goodrich, Cambridge, MA: Basil Blackwell Ltd . 1994. “Life Strategies and Social Economics in the Postfordist Age. " International Journal of Urban and Regional Research, 18:24-45. Moen, Phyllis, and Elaine Wethington. 1992. “The Concept of Family Adaptive Strategies.” Annual Review of Sociology, 18: 233-252. Morris. M, AD. Bernahardt, and MS. Handcock 1994. “Economic Inequality: New Methods for New Trends.” American Sociological Review, 59: 205-19. Muthén, Bengt O. “Multilevel Covariance Structure Analysis.” Sociological Methods and Research, 22(3): 376-398. Myles, John, and Adnan Turegun 1994. “Comparative Studies in Class Structure." American Review of Sociology, 20:103-124. 179 Nee, Victor and J imy M. Sanders and S. Semau. 1994. “Job Transitions in an Immigrant Metmpolis: Ethnic Boundaries and the Mixed Economy.” American Sociological Review, 59:849—872. Nelson, Margaret K 1999. “Economic Restructuring, Gender, and Informal Work: A Case Study of a Rural County. Rural Sociology, 64(1):18-43. Nord, Mark, A. E. Lulofli, and Leif Jensen. 1995. “Migration and Spatial Concentration of Poverty.” Rural Sociology, 60(3): 399-315. 0' Hare William P. 1988. The Rise of Poverty in Rural America. Number 15 (July). Washington, DC. Population Reference Bureau Pahl, RE. 1984. Divisions of Labour. Oxford, New York' Basil Blackwell. . 1988. On Work: Historical, Comparative and Theoretical Approaches. Oxford, New York: Basil Blackwell. Parcel, T L. and Menaghan, E G. 1994. Parents 'Jobs and Children ’s Lives. New York: Aldine de Gruyter. . 1994. Early Parental Work, Family Social Capital, and Early Childhood Outcomes. American Journal of Sociology 99: 97-1009. Piore, Michael J ., and Charles F. Sabel. 1984. The Second Industrial Divide: Possibilities for Prosperity. New York, NY:Basic Books. Pleck, Joseph and Graham Staines. 1985. "Work Schedules and Family Life in Two- Eamer Couples." Journal of Family Issues 6 (1, March):61-83. Polanyi, Karl, C. Arensberg, and H. Pearson 1957. Trade in Market in the Early Empires. New York: Free Press. Portes, Alejandro, and Sassen-Koob, Saskia 1987. “Making it Underground Comparative Material on the Informal Sector in Western Market Economies. " American Journal of Sociology, 93(1): 30-61. Portes, Alejandro, and Julia Sensenbrenner 1993. “Embeddedness and Immigration: Notes on the Social Determinants of Economic Action.” American Journal of Sociology, 98: 1320-50. Portes, Alejandro. 1998. “Social Capital: Its Origins and Applications in Modern Sociology.” Annual Review of Sociology, 24(1): 1-24. 180 1987. “The Social Origins of the Cuban Enclave Economy of Miami.” Sociological Perspective, 30:340-372. Portes, Alejandro and Min Zhou 1996. “Self-Employment and the Earnings of Immigrants.” American Sociological Review, 61(2): 219-230. Portes, Alejandro and Alex Stepick. 1993. City on the Edge: The Transformation of Miami. Berkeley: University of California Press. Presser, Harriet B. 1987. “Work Shifts of Full-time Dual-earner Couples: Patterns and Contrast by Sex of Spouse.” Demography 24(1): 99-112. . 1984. “Job Characteristics of Spouses and their Work Shifts." Demography 21 (4): 575-589. Putnam, Robert. 1995. “Bowling Alone: America’s Declining Social Capital." Journal of Democracy, 6:65-78. . 1993. “The Prosperous Community. Social Capital and Public Life.” The American Prospect 13:35-42. Redclift, Nanneke and Sarah Whatrnore. 1990. "Household, Consumption, and Livelihood: Ideologies and Issues in Rural Research." Pp 182-197. In Terry Marsden, Philip Lowe and Sarah Whatmore (eds) Rural Restructuring: Global Processes and their Responses. Critical Perspectives on Rural Change Series. London: David Fulton Publishers. Reskin, Barbara and Irene Padavic. 1993. Women and Men at Work. Thousand Oaks, London, New Delhi: Pine Forge Press. Robinson, Lindon J ., and Steve D. Hanson 1995. “Social Capital and Economic Cooperation ” Journal of Agriculture and Applied Econometrics, 27(1): 43-58. Rueschemeyer, Dietrich and Peter Evans 1985. “The State and Economic Transformation: Toward an Analysis of the Conditions Underiying Effective Intervention.” Pp. In Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol ed Bringing the State Back in. New York: Cambridge University Press. Rural Sociological Society Task Force on Persistent Poverty. 1993. Persistent Poverty in Rural America: Rural Sociological Society Task Force on Persistent Rural Poverty. Rural Studies Series, Boulder, San Francisco, Oxford: Westview Press. 181 Sampson, Robert J ., Stephen W. Raudenbush, and F elton Earls. 1997. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy.” Science, Vol. 277: 918-924. Salomon, Sonya 1992. Prairie Patrimony: Family, Farming, and Community in the Midwest. Chapel Hill, NC: University of North Carolina Press. Santi, Lawrence L. 1988. “The Demographic Context of Recent Change in the Structure of American Households.” Demography, 25(4): 509-519. Saraceno, Chiara 1989. “The Concept of Family Strategy and Its Application to the Family-Work Complex: Some Theoretical and Methodological Problems.” Marriage and Family Review, 14(1-2): 1-18. Sassen S. 1995. Immigration and Local Labor Markets. Pp. 87-127. In Alejandro Portes (ed). The Economic Sociology of Immigration. New York: Russell Sage. Sayer A. 1989. “Postfordism in Question” International Journal of Urban and Regional Research, 13(4): 666-95. Schumacher, F. 1973. Small is Beautiful, New York: Harper and Row. Sell, Ralph R 1992. “Individual and Corporate Migration Decisions: Residential Preferences and Occupational Relocations in the United States.” Pp. 221-54. In Jobes, Patrick G, William F. Stinner, and John M Wardwell Lanhan ed Community, Society, and Migration: Non-economic Migration in America, MD: University Press of America Shapiro, Isaac. 1989. Laboring for Less: Working but Poor in Rural America. Washington, DC: Center on Budget and Policy Priorities. Simpson, Harper Ida 1989. “The Sociology of Work: Where Have the Workers Gone?” Social Forces, 67(3): 563-581. Skolnick, Arlene S. 1995. The Intimate Environment: Exploring Marriage and the Family. Sixth ed New York: Harper Collins College Publishers. Smith, Mark H., L. J. Beaulieu, and A Seraphine 1995. “Social Capital, Place of Residence, and College Attendance." Rural Sociology, 60(3):363-380. Smith, Joan 1984. “Transforming Households: Working-Class Women and Economic Crisis.” Social Problems, 34(5):416-436. 182 Snipp, C. Matthew, and Leonard E. Bloomquist 1989. “ Sociology and Labor Market Structure: A Selective Overview.” Pp. 1-27. In William A. Falk and Thomas A. Lyson Ed Research in Rural Sociology and Development: Focus on Labor Markets, vol.4., Greenwich, CT: JAI Press. Soja, Edward W. 1989. Postmodern Geographies: The Reassertion of Space in Critical Social Theory. London, New York: Verso. Stack, Carol B. and John B. Cromartie 1992. “The Journeys of Black Children: an Intergenerational Perspective." Pp 363-89. In Jobes, Patrick G, William F. Stinner, and John M Wardwell Lanhan ed Community, Society, and Migration: Non-economic Migration in America, MD: University Press of America Stack, Carol B. 1974. All Our Kin: Strategies for Survival in a Black Community. New York: Harper & Row . Stacey, Judith. 1991. “Backward Toward the Postmodern Family: Reflections on Gender, Kinship, and Class in the Silicon Valley. ” Pp 17-34. In Alan Wolfe Ed. America at Century ’s End, Berkeley, CA: University of California Press. . 1990. Brave New Families: Stories of Domestic Upheaval in the late Twentieth Century America. New York: Basic Books. Sanders and Nee. 1997 “Social Capital, Human Capital, and Immigrant Self- Employment.” American Sociological Review, 61(2): Steinmetz, George and Erik Wright 1989. “The Fall and Rise of the Petty Bourgeoisie: Clmnging Patterns of Self-Employment in the Postwar United States. American Journal of Sociology, 94 (5): 973-1018. Stinner, W. F., Nithet Tinnakul, Stephen Kan, and Michael B. Toney. 1992. “Community Attachment and Migration Decision Making in Nonmetropolitan Settings.” Pp. 47-84. In Jobes, Patrick G, William F. Stinner, and John M Wardwell Lanhan ed. Community, Society, and Mgration: Non-economic Migration in America, MD: University Press of America Stinson, John F. Jr. 1990. “Multiple Jobholding Up Sharply in the 19805.” Monthly Labor Review, 113 (July): 3-10. Tickamyer, Ann R 1996. “Sex , Lies, and Statistics: Can Rural Sociology Survive Restructuring? (or) What is Right with Rural Sociology and How Can we Fix it.” Rural Soiology 61 (1):5-24. 183 Tickamyer, Ann R, Janet L. Bokemeier, Shelly Feldman, Rosalind Harris, John Paul Jones, and DeeAnn Wenk. 1993. “Women and Persistent Rural Poverty.” In Persistent Poverty in Rural America Ed Rural Sociological Society Task Force on Persistent Rural Poverty. Boulder, San Francisco, Oxford: Westview Press2201-229. Tickamyer, Ann R, and Janet L. Bokemeier. 1988. “Individual and Structural Explanations of Non-metropolitan Men and Women's Labor Force Experiences.” Pp 153-170 in Falk Lyson and Harry K Schwarzweller (eds) Research in Rural Sociology and Development, Vol. 3. Greenwich: J AI Press. Tickamyer, Ann R. and Janet L. Bokemeier. 1993. “Alternative Strategies for Labor Market Analyses: Micro-Macro Models of Labor Market Inequality.” Pp 49-68 in Joachim Singelmann and Forrest A. Deseran (eds.). Inequality in Labor Market Areas. Boulder". Westview Press. Tickamyer, Ann R and Cynthia M Duncan 1990. "Poverty and Opportunity Structure in Rural America." Annual Review of Sociology 16:67-86. Tickamyer, Ann R and Melissa Latimer. 1993. “A Multi-Level Analysis of Income Sources of the Poor and Near Poor. Pp. 83-106. In Joachim Singelmann and Forrest A Deseran (eds.). Inequality in Labor Market Areas. Boulder: Westview Press. Tickamyer, Ann R 1992. “The Working Poor in Rural Labor Markets: The Example of the Southeastern United States. ” Pp 41-62 in Cynthia M Duncan (eds) Rural Poverty in America. New York, Westport, Connecticut, London: Auburn House. Tickamyer, Ann R and Janet L. Bokemeier. 1989. “Individual and Structural Explanations of Non-metmpolitan Men and Women's Labor Force Experiences.” In F alk W. Lyson eds. Research in Rural Sociology and Development, vol. 4. Greenwich, Con: JAI Press, 1989. Tienda, Marta, Wilma Ortiz, and Shelly Smith. 1987. “Industrial Restructuring, Gender Segregation and Sex Differences in Earnings.” American Sociological Review 52: 195-210. Tilly, Louise A 1987. “Beyond Family Strategies, What?” Historical Methods, 20: 123-64. . 1979. "Individual Lives and family Strategies in the French Proletariat." Journal of Family History 4: 137-52. 1979. 184 Tilly, Louise A, and Joan W. Scott. 197 8. Women, Work, and Family. New York: Holt, Rinehart and Winston Tolbert, Charles M, and MS. Killian 1987. Labor Market Areas for the United States. Staff Report No. AGES870721. Agriculture and Rural Economy Division, ERS, US. Department of Agriculture. US. Bureau of the Census. 1990. “Household and Family Characteristics: March 1990 and 1989.” Current Population Reports, Series P-20, No. 447. Washington, DC: US. Government Printing Office. Uzzi, Brian. 1996. “Embeddedness and Economic Performance: The Network Effect.” American Sociological Review, 61(4):674—698. Voss, Paul R., Thomas Corbett, and Richard Randell. 1992. Interstate Migration and Public Welfare: The Migration Decision Making of a Low Income Population.” Pp. 111-48. In Jobes, Patrick G, William F. Stinner, and John M Wardwell Lanhan ed Community, Society, and Migration: Non-economic Migration in America. MD: University Press of America. Voydanoff, Patricia, and Donnelly, B. W. 1988. “Economic Distress, Family Coping, and Quality of Family Life. ” Pp. 97-116. In P. Voydanofi‘., P., Donnelly, B. W., and Fine M A ed Families and Economic Distress: Coping Strategies and Social Policy. Beverly Hills, CA: Sage. Voydanoff, Patricia 1987. Work and Family Life. Newbury Park, CA: Sage Publications. Wacquant, Loic J .D. 1998. “Negative Social Capital: State Breakdown and Social Destitution in America’s Urban Core.” Neth. J. of Housing and the Built Environment, l3(1):25-40. Wade, Robert. 1990. Governing the Market: Economic Theory and The Role of Government in Eastern Asian Industrialization, Princeton: Princeton University Press. Ward, Kathryn 1990. Women Workers and Global Restructuring. Ithaca, NY: ILR Press. Wallerstein and J. Smith 1991. "Households as an Institution of the World-Economy." In Race Lesser Blumberg, (ed) Gender, Family, and Economy. Sage Publications. 185 Wenk, DeeAnn, and Constance Hardesty. 1993. “The Effects of Rural-to-Urban Migration on the Poverty Status of Youth in the 19805. ” Rural Sociology, 58: 76- 92. Whitener, Leslie A. 1991. “The JOBS Program and Rural Areas.” Rural Development Perspectives, 7(2): 21-26. Wilson, Julius William. 1987. The Truly Disadvantaged: The Inner City, The Under Class, and Public Policy. Chicago and London: The University of Chicago Press. . 1996. When Work Disappears: The World of the New Urban Poor. New York: Random House. Wolf, Diane L. 1992. Factory Daughters: Gender, Household and Rural Industrialization in Java. California: University of California Press. -------. 1991. “Does Father Know Best? A Feminist Critique of Household Strategy Research.” Pp 29-43 in Clay and Schwarzweller (eds) Rural Sociology and Development, Vol. 5: Household Strategies . Greenwich, Connecticut. . 1990 “Daughters, Decisions and Domination: An Empirical and Conceptual Critique of Household Strategies.” Pp 43-74. In Development and Change Vol. 21. London, Newbury Park and New Delhi: Sage Publications. Woolcock, Michael. 1998. “Social Capital and Economic Development: Towards a Theoretical Synthesis and Policy Framework” Theory and Society 00: 1-57. Wright, Mareena McKinley. 1995. I never Did Any Fieldwork, But I Milked an Awful Lot of Cows. Gender and Society, 9: 216-35. Zukin, Sharon and Paul DiMaggio. 1990. Structures of Capital: The Social Organization of the Economy, New York: Cambridge University Press. Zhou Min 1992. New York ’s Chinatown: The Socioeconomic Potential of an Urban Enclave. Philadelphia Temple University Press. 186 avegxu Bufiowo 5:208:82 come— 28380 3&2 882 3:2qu «agav— oflo>fih v55 888m 59392—0 59630 £58m 8.5.5. «385 50 093302 news—32 81382 E3— 3:3 Sumo—2 Gong—QM «NH vNa vNH vNH vNH vNH vma MNH MNH MNH mNH MNH mNH MNH mNH NNH NNH NNH NNH NNH NNH NNH NNH HNH HNH mmamm mNHoN mHHmN mmomm mmowm hoomm Hoomm mmomm mnomm mmomm hvomm Hmomm mmomm maomm mooom mmawm hmaom mmawm HNHwN hHHmN Hmomm hooom mooom mmaom whomm gnu—=60 @8303 2880 093332 88002 832 oSmEaZ 03 £33k Begum @883 :25: 3336 an 8.8.2 .533. ease 2.6 329: 58m nous—U can? >§m yum—G75 m<fl¢< hang—<2 K85 HNH HNH ONH ONH oma ONH ONH oma ONH mHH mHH mHH mHH mHH mHH mHH mHH maa mHH baa baa baa mHH mHD—szh: mmomm maowm mwfiwm mmaom maflwm reflwm moamm Hoawm mmomm rmamm mvfimm HHHmN moomm Hmomm haomm Haomm mnowm swoon mmowm mwoom mvoom bmomm mmflmm eases? §33a£m 8:5 as .5 8.25 sacs: 8%: .6093 0303 8333 .838». 0.3% 0:352 cacao—m «ab—connem 02:80:02 033.52 8:382 3:4 .6: saga—05 8.09 332530 5w? gz E7560 mHH Hoawm mHH mmamm wHH HnHoN waa hvaom mHH mNHoN oHH mmomm oHH mmomm mHH hmomm wHH mvomm mHH Hmomm mHH mwomm mHH mmomm vHH mnomm vHH bmomm vHH mmHoN «Ha moamm «Ha moawm vHH hmowm VHH mmomm vHH anomm vHH mvomm «Ha Hvomm vHH mmomm vHH moomm .32324 wk:— 187 8.3a fifixfi 55.5 0302 09853 555 menu 915 £22 em =Sfih€4 creme: Saga unease «355 «33 Ex :33..— 3555 0252 53 8> .5952 8:33. 85. .88»? 828m mma hma hma nma FMH hMH mma omH wma oma mma mma mma mMH mma mma mma mmH mMH mma mma vma va vma «ma vma mma mma ovoma mvamm mmomm mHHmH hmoma mmoma bNomN Hmomm mvamfi HvaH mmoma Hmoma meamm HhHmm mmamm mmamm mmomm mmomm Hmomm mmomm mHHmN Hwamm emamm boamm Haomm moomm thmm hvamm eEO 85>» Exam Beebe—U can mmH mma MMH ama HMH HmH ama Hmfl HmH HMH Hma HmH HMH HMH Hma HmH Hma aMH HmH Hma mNH mNH mNH mNH mNH mNH FNH PNH FNH PNH mvamm mwomm mmomm mNNHN MNNHN mHNHN HHNHN mmHHN mbHaN anHN HHHHN moaam Hwoam mNoHN mnama broma Hmoma mvomfi mmoma maoma mmama bmHmH mHHmH bvomfl Hmoma mmoma moamm Hmomm mmomm haomm 565 was... coco—neon— .882 “53 883. :aflv :53 3.83 Feed seam 88m can? 5:5 acumen 32m 032m b25332 Egg 9&3 23sz 0880 8.28..— 3:5 8.36 30 $3.56 «888083— 32 2.32m cacao mazNH mmoam hNH macaw mNH whama wNH HonH wNH Hvomfi mNH mvamm mNH mmHmm mNH maamm mNH moamm mNH Hmomm mma abomm mNH hmomm mNH bvomm mma hmomm mNH bmomm mNH mmomm mNH Hmomm vma mvaN vNH HvaN vNH bmawm cm<§€H warm 188 £83 8.5: arm can? Bkgogh saw mos—om 8mg: :Boofiou :23 Stem 5562 8.3 8&2. x29 333 a??? 93an 8:333 25va mow—.55 fiaeoc§> coon—cam themed 8am baud 835 $35 5.335 Hma HmH HmH omH omH omH omH omH omH omH mvfl mvH ova mvfi mvfi mva mva mvH mva mefl nvfi va bvfi hva Fwd an nvfi th hva mmomm broom mvomm mmamm hmamm Hmamm mhomm bwomm Hmomm maomm emama HHHmH mmoma mhowfi mmoma mmHhH Hmaba mmeH Hoaha pvo>H meawa mmeH hvama mNHmH mNHmH mmama Hmowa hmowa mmeH 830 owaflo 08.82 €32 8853 2880 “81:5 32>» 853 8:33th aged 53:0 =05 860m 8.5—53’ ES eoE§> 53:5 6525 043m 980982 >20 269m «wan—5.. 89.8.. Boaefifiam Beam nae wvfi ova med oea ©VH ova ova mvfl mva mvfl mva mva mvH mva mva vvfi vwfl qu vva vva vva vvfi mvH mva mvH mva Nva NvH mHHmH baama moama Hoama mmomH mmowa bmomfi Hmama HmeH hmHmH mvomfi mmoma maoma hooma mmaba hmawa mewH mmama mmHmH HNHmH noamfi Hmoma mvama mnoma aboma mooma mvama mmama cam—02 e052 e852 303.com x828: 858:: 5595 88m $23 Eu? =3:on zen ..quugm :3 on :22 ESE 385 8&32 team “Baku Baku—on Bowman—m 33:. £315 BEE ass: v.80 ages gassed ”$22;n>h&R=HU Nva moamfl Nvfi hmoma Nva Hmoma NvH mooma NvH mmomfi Nva hmoma Nva maoma Nva HHomH Hva mmHmH flea thmH HVH Hmama HvH aroma avH mmoma HvH mmomH HvH moomH HvH Hooma ova mmamfi ova mmoma ova mmoma ova mmomfl ova mmoma oea mooma mma mmama mma Hmama mma moama mma bmoma mma haoma mma mmamfi mma mmomfl AXEJZA ma:— 189 0:38,: cook—yum 8x82 3332 53:33 5:5 ABM ion tuna—m 8.55 35> 3.5a ”225 ES «8.8m Began 6522 gwwaofio 33833— 5.38: omauwoo «wanna 2:33 .695 3:333 533— gag scam hfiau Rom Nam Nam Nam Nam aaN aaN aaN aaN aaN aaN 0am 0am oam oaN oaN oaN 0am 0am 0am oam oaN oam aha aha aha aha aha aha aha aha mNabN hmahm mmOhN mmonm mmamm maamm hoamm mmomm maomm moomm mmamm mmomm mmomm amomm avomm hoomm moomm amamm mmomm amomm mmomm maomm mmovm aaovm hmomm momam mmoam mmoam mvoam maoam .8952 3:333 3.55 38%: «53.530 cases. 3882 8.52 2835 €22 8:5 >33on E532 mad—05 55a wad—com azxsam vaaaim 03325 .832 332 «:8 80:3 .865 830m 83a 2.5 88.8.. «ma qwa vma vma vwa ooa oma oma ooa oma mma mma mma mma mma mma mma mma mma mma mma mma mma hma Ema bma hma hma mwomv mpomv mmamm mmomm mmomm mmamm haamm aoamm mmomm moomm mmamm mmamm hmomm mmomm mmomm mbomm mvomm mvomm awomm mmovm moamm mmomm moomm moamm mvamm avamm amamm mnomm «gm once—cm 3:62 8.3 $230 $235 a3n8a2 "5:2;n>h5:=!9 wma mma wma wma mna wma mma mma mma vma vma vma vma vma MMa mma mma mma mma mma Nma NMa Nma Nma Nma Nma Nma moavm mmovm mmovm amovm aaamm maomm mmovm moovm amomm hmamm aNamm maamm maamm mmomm hoavm moavm mmovm mnovm mmovn wmamm mmamm mmamm moamm mmomm mmomm mmomm hoomm eQ¢GAAH mafia 190 $32.53 .5533 egg 2.3 :85 03 .6 98m .855 38:5 832:82 3255 EM §> 08.82 8.8.12 303— magma“... $25 :5:— 8:3 «33 E555 33:38? 25832 853 Son mmm mmm mmm mmm mmm mmm mmm vNN qNN mmm mmm MNN mmm mmm mmm mmm mmm NNN NNN NNN NNN aNN aNN aNN aNN aNN CNN mmamm bMamm mmamm hmomm hqomm mmomm maomm haamm abomm mbama hhama mmama nmama mmama FOama aoama amoma maama moama mmoma aaoma mwama mmama maama moama amoma mmoma .355 :3qu Sagan 8.35 55325 582m «as: :85 «a: $3838 3&2 89.3. 533 goings? canons? =33: 333$ “—0855 22... 335.020 880 950 23%? 3:85 .3302 Sago Eon—02m ”Beam 03.8 «38— 33 320 ONN ONN omm ONN omm omm omm mam mam mam mam waN mam mam mam mam mam mam mam mam ham ham ham ham ham ham ham mam mam whoma mwoma mmoma mmoma maoma haoma Maoma amama hmama hmama mmoma mmama amama mmama amama mmoma awoma booma bmoma mmoma nmahm moahm mmowm averm thFN mvo>m mmorm amonm mmowm 8:0 :32 «:80 am 3.65 3%? 83:38? “afim has our— 02.3 :53 53:5: 89:5 3820 .595 5.22 3.88m Saga 83 2:: 58m 53.833 3033 28% 3E «ozoomz 33m 3 .565 53m 0:5 335 2: >558 mam amowm mam aoonm mam moamm mam mmomm mam abahm mam mmahm mam mmahm mam mmahm mam maahm mam moowm mam mmonm mam mmowm mam FMOBN mam mmowm mam machm mam moonm wam mvabm qam avahm vam mmonm vam moonm mam mmahm mam aoahm mam bvahm mam amahN mam moahm mam mnowm mam maonm mam maonm Nam M¢ahN Ema—h 191 Bow—v83 =okonah 8525 38m .882 3.9.32 0:3 3 303m zoom 58.3.. 8:50 4.5m 23...: zoom 8052 .95: 3.53 @8882 .85— 8:5 x30 84 QB: 8502 men gauge: «828$ 3.25 E 9.5.82 «5062 mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm bmm 5mm nmm nmm mmm mmm mmm omm omm 0mm mmm mmm mmm mmm momha mBHbH mmHhH mwaha mmaba mmaba mmora HHOFH monH hmoma mvoma thhH Hwaba Hmahfi mnowa hmaba moaha mocha fincha mvomm HHHmH hmomfl bmoma arena Poona mbaba hvaha mmaha mHHhH Rom 33:8 amigo 5322 5&5: 53 on 2%.. 853m 323% 33m van—£850 EomBEU 3.00 Eu 430 3.5.3 933509 Etna—U 081822 301:5 2. am @5202 88.382 «>5— 055 «38300 2333 G005 mmw mmm mmm vmm «mm vmm mmm mmm mmm mmm mmm mmm mmm mmm mmm mmm Nmm mmm mmm mmm Hmm Hmm Hmm HMN Hmm Hmm omm 0mm mmona HthH macha MHHBH moaha mmowa thFH ancha mvona vaPH mMOFH mmora mNOPH mmowa mmora Hmoma mmoma mvomH moomfl mmona Haamm moamn Fromm avomm mmomm amomm mwomm mvomm .55 .5320 85> 3:82 885 3 =85... 363m aofioqaufi. 3%.. 538—. 230 3m :55 «32530 camam .823 .895 83a... awaken 85.52 £854 £30 §< 2833a 9:80 858982 ”.333 8§u3oM xxx“ Esp—m m=z¢i~>h5R=HU omm mvomm omm mmomm mNN mmamm mmm Hmomm mmm mmomm mNN hmomm mNN mmonm mmm aNHmm mNN Hmomm mNN mmomm mNN mmomm mNN mmomn mNN haomm mNN Haomm mNN mwabm nNN Hvam hNN maamm hNN hmomm hNN mpomm hmm mmomm hNN mHomm mmm Hoomm oNN mHHmm mNN mmomm wNN whomm mNN hoomm mNN Hmomm wNN mmomm oNN moomm AKHJQA wir— 192 8.5m 5> 3on 9% >083 8:52 .852 E 883335 con—am 08305 €50 28.5 .855 m3 8: .852 £33 8?— 3.34... 5:83.82 acmBaE 39582 58:30 vaom 38m .0323 8535 382 ES Hmm Hmm Hmm amm Hmm Hmm Hmm Hmm Hmm Hmm Hmm HmN Hmm 0mm 0mm 0mm 0mm 0mm 0mm mvm mvm mvN mvN mom mvm mvm wvw mvw mvm mvm avam hmam mman mNHm Hoam mmom moon moon mvom mmom mHom moon moom mbamm mmamm hmamm HHHmN mvafia Hoona nmaba mHHPH bHHhH mmona macha mocha ahaba moahfi hmabfl hmahfl mmaha 5&3 2505 59.50 «80 Exam as $3 gm 353 v.33 am 3:30 am anwaoz 58:3 cabana ova—.836 nag—Sch Botany 59:33 :20 am 08.82 983233 magenta ..m 303250 Sm Pom cam—32 :9: £83 232 3.2: 352 .89 mvm mvm mvN mvm mvm hvm hvm bvm bvm th hvm bvm bvm hvm bvm how hvm hvm wvm mvm mvm owm mvN ovm mom mvm mvm mvm mvm hoaba HmOPH HNOhH FHONH mocha OHmmN mHNmN mmHmN mmHmN mmamm mfiamm mmomm mpomm abomm mmomm boomm mmaba mmHhH HNNmN bmamm mmamm finamm mmamm mmomm mhamm mmamm Hmamm mNHmN moomm eggs? 33823 gagem pagan €50 280m :5» 5302 03 33M 23M 035 :80 835 £959: .3833 85:38? 84:30 0333»; East»... 53313 :03. 0303. £88m mfil¢ifi>h5qzyu vvm aomha vvm mmaha vvm hhHFH vvm avaha vvm moaha vvm bMObH wvm mHOPH vvm FOOPH mvm hmaha mvm HHHFH mvm bacha mvm mmora mom mwODH mvm mmowa mvm mvcwa mvm HMObH Nvm Hmona Nvm mnona HvN mmamn HvN Hmamm HvN mmomm HvN mromm HvN mmomm va hmomm ovm hNHmm ovN moamm ovN Hoamm ovm .mmomm gag.“— 193 933: 883.820 Bah—um onom €333 aim 3.2.3. :55: genius .85 53m 35:3 .39 E580 =Ba§u Evan 55% 356:. 255 ma 53:3 3.05 ususm .0323 253 £33 mafifii 00.502 .8202 :03: commas: 5: bow wow now how oow oow oow oow oow oow oow oow oow oow oow oow oow oow oow How How How How How How How How How How How How noObw mwonw beehw ooonw mwaov oaaov moaov Hmoov mmoov Hooov mvoov ovoov nmoov Hmoow Hwoov mHoo¢ omomm ooabw HHObw Hawmw oowmw mmamw hmamw omflmw ohfimw ahamw bmamw mwamw Hwamw bfiamw oHHmw x29 €55 SEED €50 ~33 can? 283 885%? 33:8 3:3 am 383— 883 2.. Mo 93 8.3 9.30283— x80 8:30 33a.— £3.59 338332 2:83 2685 exam 35 ER: 302 Egg: hfiau 3326 a5 .23 53:8 How How How How How How How How oow oow oow oow oow oow oow oow wow wow mow bow now now now now now bow now now how moamw mbomw avomw mmomw Hoomw omHmH hHHmH booma Hmooo bmahw omahw finchw ouchw firebw HmOFw hHObw ooamw moomw mmoo mwwmw howmw Howmw Hmamw mvamw mmamw omomw Hmomw mwomw baomw Gems—nag}, 8:5 25% £33 382 EA Gems—£2. Goa—own sea: :35 :35 how—~53? saga? 8.52 Gauche». 855: GEESU $5.32 8308002 5.9.3: 03:30 bio—10 Eazam 23m 0882 35,—. nouns—m E0 comp—O =oBom mazno>H oow mocha oow mocha oow mocha oow moona oow mmana oow award oow Hocha oow ooowa oow hNObH vow hoHHw vow ovHHw vow mmaaw vow mmoaw vow omoaw ¢ow booaw vow Hoaba vow hwaha How onmw How mowmw How mofimw How mvfimw How Hmomw 9Q<§fiu mEEh 194 cog 335 madam Sm 9583 £88»: 8282 85>) Ea: beam 3.25 2°me «.8 85a: 2.55 0:85 .585 8:5 2835 9:5 v.80 =25 oaoom §=B< 2%? H8303 38:33 «E 6.38:: omw omw omw omw omw omw omw ohw ohw onw obw ohw ohw ohw ohw onw obw obw ohw ohw obw obw ohw ohw obw vhw vhw vhw vrw vnw vhw Hooam mvoam mmoam HwOHm mmama mmama mmHmH HmHmH oFHmH monH oonH monH monH HmeH broma mnoma mooma mvoma hvomfl mmoma mwomH hwomH oHomH mooma HoomH bmHmH hmHmH HonH Hoama mmomH HmomH gum—Ha: 8:805 583 39.5 88m nabs: zazfiam .zjxoia 5:349 fiafiim cozaagn >30 8302 >952 88.8.. 303:0on 5.2m 3380 8.:mb :93 2:232 333. cflsm xqqm 39¢.“qu :93 38:5 25 3.. 23 €55 «32330 8033 vhw vhw vbw whw mhw mnw mnw mbw obw mmw mnw www www www whw wbw wbw whw whw th th wa Hnw wa wa th th wa mow mboma omomH owomH HwomH Hmonw vaPw hvama moama moomH moomH avomfi ooahw Hoabw moonw mmonw onmH mvama HVHmH mHHmH mnabw Hoabw mmahw haabw mochw Hmobw mnoww noOPN mwonw moabw gnaw Bum—um 23an 885% 3:023 35 >305 v.80 8.5m :35» 5882 30 53>» 0.0on use—Em atom v5.5 8:960 9.3 com “=8 gum—HEEL Ema—z Baca—z 835— gz F5560 mow moahw mow mVth mow Hwabw mow Haabw mow hmObw mow Hoonw mow Hochw mow mmomm mow Hmomm mow whomm mow mmomm mow ovomm mow mmomm mow Hwomm mow baomm mow moomm mow hofibw mow hoahw mow thFw now mmomm bow hmomm bow Hmomm bow hoomm now omomm now maomm bow owflhw how mHth bow maarw how mmoww how mochw how moonw amddéd mhfim 195 03325 255902 83:28? 8&32 Exam 9585 §Eo> 5502 898—. 3:5 “300 now—am Homafiu 08.830 55.58 35,—. 2.3m 38.53 55m a»; 6.303 ion coo—23 bog—oi mom mom mom mom mom mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw hmw nmw mmw hmw Hvoov mHHmw mvan boom boom mHHov wamw mvamw hmomw boomw mmomw HHomw bmoow Hwoow HHoow memw mowmw moamw moomw mwwmw mwwmw honw moamw mmomw 2580 83:8 8:8 Sago owamo 5&82 30:82 552 29.8: 280 200 8950 $3230 28.5 8.5? H30 um ham SHE 36m %Bnooz 030334 8.0.52 gamma: :om b8: nag: muggy ado—on 8325 8:50 Eu bmw bmw mmw mmw omw omw omw omw omw omw omw omw omw omw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw mmw omw mmw mmw mmw mmw mmw bromw hoomw mmomw mvomw Hmamw Hvaw omamw Hmamw mmomw mmomw Hmomw mwomw hwomw maomw mwwmw nmfimw bramw monw mmHmw bvfimw hoamw Hoamw mmomw bwomw mmomw Hmomw mnomw moomw Hoomw 30mm 20mm 280 €325 unguam 380m 83m §22< 33.2 0:35»? :82 8: €o38>§ 88:2 Susan «Sam—~02 .8383 coon—am Em 3325 ”$8 9.80 835. o§a3§cm owam b28982 3:2 float—um 2882..— 2.3 8:5 ”53 ESE 22 E00 mmw bmomw mmw mwomw mmw Hwomw mmw mHomw mmw maomw mmw moomw omw moomw mmw mowow mmw Hwaow mmw befiow mmw moaow mmw Hmoow mmw mvoow mmw oooow wmw hnaam wmw omHHm wow mmaam wmw mmoam www mmoam www mwoam wmw thmH wmw mmHmH wmw mvaH wmw hmamfi wmw memH wmw mmomH wmw aboma wmw moomH omw wwfiov omw 32m omw 32m AXEJZA mafia 196 MICHIGAN STRTE U NIV. LIBRRRIES HI” llll H WIHIIWIINH\IMIIWIHII 9 582056 312 3020