-' W £24“ng I. L ABSTRACT AN ANALYSIS OF THE AGE-OCCUPATION MATRIX AS A CRITERION FOR VERTICAL AND HORIZONTAL MARKET DELINEATION by David Luther Wilemon 131 recent years, market researchers have attempted to improve the market delineation processes. Although other criteria can be employed as a basis for delineating markets, increasing emphasis has been focused on the use of single and nu11.t:i;>le socio-economic variables. Two methods fre- quently employed for delineating and analyzing consumer experxijxtlirwa patterns are social class and aggregated fam- ily life cycle analysis. Both methods, however, are sub- ject tx: IJrLique limitations in delineating markets. In so- cial class analysis, the family age variable usually is ignored within each class. Researchers delineating markets by aggregated life cycle analysis may not consider the within Class variability of incomes, occupations, social classes, and ages. Engea to the limitations of social class and aggre- gate lgifke <:ycle analysis, an alternative approach to mar— ket delineation is proposed and investigated in the research. The apprOaCh utilized employs both age and occupation of the family head as criteria for delineating markets. David Luther Wilemon {The research examined selected economic patterns of families in nine Standard Metropolitan Statistical Areas in the North Central and Eastern regions of the United States. The occupations investigated included salaried professionals and officials, self—employed, clerical and sales, skilled, semi-skilled, and unskilled. Bach occupa- tional category was disaggregated by the following age clas- ses: under 25, 25—34, 35-44, 45-54, 55-64, and over 65 yearws. flflue unpublished, cross-sectional data used were collected by the Bureau of Labor Statistics in 1960-61. A Least Squares computer program was used for determining the ixmcxonua elasticities, the marginal propensities to con— sume coefficients, the standard error of the coefficient, and statistical significance of each coefficient. {Irma study revealed the following general conclu- sions. First, incomes and expenditures vary considerably among"tYN3 ‘various age classes of each occupational cate— gory. Generally, the higher ranking occupations had higher outlays on the expenditures classified as "necessities“ (total. fkpcxfi, clothing, and total housing). By contrast, in the lower family age classes, the lower ranking occupa— tions spend absolutely and relatively more on house furnish- ings and automobile expenditures than do families in the higher‘ rwarHCing occupations. Second, expenditures generally vary more among the age classes of the higher ranking oc- cupations than for the lower ranking occupations. Third, David Luther Wilemon family Size and the number of children under 18 appear more influential in determining total food expenditures than does the family income level. Fourth, family disposable incomes reach maximum levels in the secondary age classes (45 years and over) for all the occupations. Disposable income lev— els generally follow the occupational ranking scale. Fifth, family savings levels do not completely follow the socio- economic ranking scale in the higher ranking occupations. Savings follow the ranking scale in the skilled, semi-skilled, and unskilled groups. Sixth, definite patterns were not always found for the relationship between occupational rank and the income elasticities for total expenditures. In most cases , however , the middle and lower ranking occupa- tional families had higher elasticities for total housing, total food, and clothing expenditures. The highest marginal propensities to consume usually were found in the primary age classes (under 44 years) of each occupational group, and generally were found in the clerical, skilled, and semi- skilled categories. In conclusion, the age-occupation variables appear potentially valuable as criteria for delineating markets. The approach utilized in the research demonstrates many of the advantages of the social class and family life cycle delineation methods. Furthermore, it eliminates many of the weaknesses of the social class and life cycle approaches. Two primary methods by which the age—occupation David Luther Wilemon approach can be utilized in market delineation are: (l) in identifying prime market targets for purposes of direct— ing the marketing effort; and (2) for researching socio- economic changes in market areas over time. IKNI.ANALYSIS OF THE AGE—OCCUPATION MATRIX AS A CRITERION FOR VERTICAL AND HORIZONTAL MARKET DELINEATION By .. {I "\E/ Q David Li Wilemon A THESIS Submitted to Michigan State University .in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1967 Copyright by DAVID L. WILEMON 1967 ACKNOWLE DGMENTS A research work of this nature is impossible with- out the assistance of others. The author wishes to acknow- ledge the assistance and cooperation of the following indi- vidual s and organizations: Dr. Bernard J. LaLonde, Chairman of the Dissertation Com— mittee, for the many hours devoted in aiding the author in conceptualizing the research problem. Dr. LaLonde's example and dedication were invaluable to the author in completing the research. Dr. LaLonde's influence goes much further than the bounds of this research. Dr. w. J . B. Crissy, Associate Dean of External Affairs and Professor, Department of Marketing and Transportation Administration, who continually provided patience, under- standing, and encouragement throughout the author's gradu- ate program. Dr. Crissy's assistance and suggestions in this research are deeply appreciated. Dr. Stanley C. Hollander, Department of Marketing and Trans- portation Administration, for sharing his knowledge and research skills. Dr. Hollander's excellent ability to organize knowledge and his sense of style undoubtedly made the research more meaningful. Dr. Thomas A. Staudt, Chairman, Department of Marketing and Transportation Administration, for providing much en- couragement and financial assistance throughout the author's graduate program. The Bureau of Labor Statistics and especially Mr. Thomas Lanahan for suggestions on using and interpreting the 1960-61 Survey of Consumer Expenditures. Several graduate colleagues at Michigan State Uni— versity aided the author in formulating and solidifying the general problem. Such colleagues provide a constant source of meaning and enjoyment to a graduate program. The author hopes such comradeship continues throughout his pro- fessional development. Finally, the author would like to express his grat— itude to his wife, Jane, for her devotion and encouragement throughout the entire graduate program. Without her help in typing and reading the many rough drafts, the research could not have been completed. Dedicated to Jane, my best friend, for her understanding, sacrifice, and encouragement Chapter I. II. III. IV. V. TABLE OF CONTENTS BACKGROUND AND PROBLEM DELINEATION. . . . . . Background. . . . . . . . . . . . . . . . . The Problem . . . . . . . . . . . . . . . . Method of Approach. . . . . . . . . . . . . Limitations . . . . . . . . . . . . . . . . Some Possible Contributions . . . . . . . . Organization. . . . . . . . . . . . . . . . REVIEW OF LITERATURE O O O 0 O O O O O C O O 0 Organization and Rationale. . . . . . . . . Influence of Selected Socio-Economic Variables on Family Expenditure Patterns. Evaluation of Selected Socio-Economic Variables as Criteria for Delineating Markets . . . . . . . . . . . . . . . . . RESEARCH DESIGN . . . . . . . . . . . . . . . Introduction. . . . . . . . . . . . . Source and Nature of Empirical Data . Variables Utilized in the Research. . Analysis of Data. . . . . . . . . . . Definitions of Terms. . . . . . . . . PRESENTATION OF FINDINGS. . . . . . . . . . . Introduction. . . . . . . . . . . . . . . . Occupational Socio-Economic Rank and Absolute Expenditure Relationships. . . . Occupational Socio-Economic Rank and Relative Income-Expenditure Relationships Occupational Rank and Family Disposable Income Concentrations . . . . . . . . . Family Absolute Savings Levels. . . . . . Marginal Income-Expenditure Relationships Summary . . . . . . . . . . . . . . . . . O O O 0 SUMMARY AND CONCLUSIONS . . . . . . . . . . . Introduction. . . . . . . . . . . . . . . . Evaluation of Hypotheses. . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . Page 18 24 27 3O 31 33 33 33 84 110 110 111 125 130 I37 141 141 142 160 174 178 195 198 202 202 202 220 Chapter Comparison of the Occupational-Age Approach with the Social Class, Family Life Cycle, and Disposable Income Prediction Methods. Implications of Findings for Market Delin- eation. . . . . . . . . . . . . . . . . . Suggested Areas for Further Research. . . . APPENDIX A. Socio-Economic Characteristics of the Occupational Categories. . . . . . . . APPENDIX B. Selected Historical Consumer Budget StUdieSo 0 0 O O O O O 0 O O O O O O O APPEDHDIX C. Other Selected Family Expenditures . . . APPENDIX D. Specific Steps in the Data Processing. . BIBLIOGRAPHY O 0 O O O 0 O O O O O O O O 0 O O O O O O Page 262 273 281 285 295 302 308 313 LIST OF FIGURES Figure Page 1.. Influence of occupation on the life style and consumption life style of the family . . . . l3 2. Schema of vertical and horizontal market de- lineation. . . . . . . . . . . . . . . . . . 15 3. Graphical illustration of the absolute income hypothesis . . . . . . . . . . . . . . . . . 51 4. Graphical illustration of the relative income hypOtheSj—s O O O 0 O O O O O 0 O 0 0 ° 0 O O 53 5. Graphical illustration of the permanent income hypOtheSis O O O O 0 O O O O O O O O O O O O 57 6. David's life cycle classification patterns . . 9O 7. Family disposable income profiles by occupa- tional category and family age class . . . . 233 8. Total family expenditure profiles by occupa- tional category and family age class . . . . 236 9. Total family food expenditure profile by occu- pational category and family age class . . . 238 10. Total family housing expenditure profiles by occupational category and family age class . 240 11. Family clothing expenditure profiles by occu- pational category and family age class . . . 242 12. Family house furnishings and equipment expen- diture profiles by occupational category and family age class . . . . . . . . . . . . . . 244 13. Family automobile expenditure profile by occu- pational category and family age class . . . 24S 14. Saving profiles for the professional, self- employed, and clerical families by percent- age of disposable income . . . . . . . . . . 248 Figure 150 16. 17. 18. 190 20. 21. Saving profiles for the skilled, semi-skilled, and unskilled families by percentage of disposable income. . . . . . . . . . . . . . Family tion Family tion Family tion Family tion Family tion Family tion disposable income and selected consump— profiles, professional families . . . . disposable income and selected consump- profiles, self—employed families. . . . disposable income and selected consump- profiles, clerical families . . . . . . disposable income and selected consump- profiles, skilled families. . . . . . . disposable income and selected consump- profiles, semi—skilled families . . . . disposable income and selected consump- profiles, unskilled families. . . . . . Page 251 254 255 257 258 260 261 LIST OF TABLES Table Page 1. Possible Number of Delineated Market Classes Using Several Socio-Economic Variables. . . . 6 2. Evaluation of Selected Socio-Economic Market Delineation Variables . . . . . . . . . . . . 12 3. Income from Wages and Salary of the Wife at Different Stages in the Life Cycle--Average All Income Classes, 1953. . . . . . . . . . . 64 4. The Median Income of Spending Units at Differ- ent Stages in the Life Cycle, 1952 and 1953 . 64 5. Comparison of Income, Stage in Life Cycle, and Percent of Families Purchasing Durable Goods, 1953 Survey of Consumer Finances. . . . . . . 67 6. Purchases of Household Appliances Over Four Year Period by Income Class of the Family . . 72 7. Purchases of Household Appliances Over Four Year Period by Social Class Position of the Family. 0 O O O O O O O O O O O O O O O O O O 72 8. Relationships of Food Expenditures for One Week to Family Size, Urban Families, United States Department of Agriculture Survey, 1948. . . . 8O 9. Selected Average Per Capita Expenditures by Family Size and Income Class, Large Cities in the North, 1950‘51 o o o o o o o o o o o o 82 10. Alternative Family Life Cycle Stages and Nomen- clature . . . . . . . . . . . . . . . . . . . 92 11. Department of Commerce Selected Personal Con- sumption Expenditure Estimates as Percents of Total Personal Consumption Expenditures, 1961, and Bureau of Labor Statistics Percent Dis- tribution of Expenditures for All Families in Urban United States, 1960-61. . . . . . . . . 115 12. Comparison of Percent Distribution of Expendi- tures for Current Consumption by Income-Age Class, 1950 and 1960-61 0 o o o o o o o o o o 116 Table 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Population Characteristics of Selected Standard Metropolitan Statistical Areas in the North— East and North Central Regions, 1960_6l o o 0 Number of Sample Families in Each Age Class, Occupational Category, and Standard Metro— politan Statistical Area. . . . . . . . . . . Family Age-Occupational Research. . . . . . . Cells Employed in O 0 O O O O 0 O O O O 0 Occupational Socio—Economic Ranking, Mean Edu- cational Attainment, Dollar Incomes, 1960. Family Total Expenditures by Occupational Socio— Economic Rank and Age and Mean Disposable Class I O 0 O O O O O 0 Family Total Food Expenditures by Occupational Socio—Economic Rank and Age Class . . . . . . Family Total Housing Expenditures by Occupa- tional Socio-Economic Rank and Age Class. Family Clothing Expenditures by Occupational Socio-Economic Rank and Age Class . . . . . . Family House Furnishing tures by Occupational and Age Class . . . . and Equipment Expendi- Socio-Economic Rank 0 O O O O O O O O O O C Family Automobile Expenditures by Occupational Socio-Economic Rank and Age Class . . . . . . Average Propensities to Total Expenditures by Economic Rank and Age Average Propensities to ties” by Occupational Age Class . . . . . . Average Propensities to Consume for Family Occupational Socio— Class O O O O 0 0 O O O Consume for "Necessi- Socio-Economic Rank and O O O O O O O O O O O O Consume for Total Food Expenditures by Occupational Socio-Economic Rank and Age Class. . Average Propensities to Consume for Clothing Expenditures by Occupational Socio-Economic Rank and Age Class. . 0 O 0 O O O O O C O O O Page 121 122 126 128 143 146 149 151 154 158 161 165 168 171 Table 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Average Propensities to Consume for Family House Furnishing and Equipment Expenditures by Occupational Socio-Economic Rank and Age Class . . . . . . . . . . . . . . . . . . . . Family Disposable Income Levels by Occupational Socio-Economic Rank and Age Class . . . . . . Family Savings Levels by Occupational Socio- Economic Rank and Age Class . . . . . . . . . Income-Expenditure Elasticities for Total Ex- penditures, by Socio-Economic Occupational Rank and Age Class. . . . . . . . . . . . . . Income—Expenditure Elasticities for Total Food Expenditures by Socio—Economic Occupational Rank and Age Class. . . . . . . . . . . . . . Income-Expenditure Elasticities for Family Clothing Expenditures by Socio-Economic Rank and Age Class . . . . . . . . . . . . . . . . Income-Expenditure Elasticities for Total Hous- ing Expenditures by Socio-Economic Occupa— tional Rank and Age Class . . . . . . . . . . Marginal Propensities to Consume for Total Ex— penditures, by Socio-Economic Occupational Rank and Age Class. . . . . . . . . . . . . . Variability in House Furnishings and Equipment Expenditures by Age Class, Skilled Workers. . Variability in House Furnishings and Equipment Expenditures Between the Average for All Occupations and Skilled Families. . . . . . . Absolute Ranking of Disposable Income and Se— lected Expenditures for All Occupations, 25—34 Age Class 0 0 O O O O O O O O O O O O 0 Summary of Disposable Income--Selected Expendi- ture Rankings by Occupational Rank, 25-34 Age Class O O O O O O O 0 0 0 O O O 0 O O 0 O O 0 Percentage Expenditure Indexes for House Fur— nishings and Equipment Outlays. . . . . . . . Page 173 176 179 183 186 189 192 196 266 269 271 272 276 Table 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. Family Socio—Economic sionals . . . . . . Family Socio—Economic Employed. . . . . . Family Socio-Economic Family Socio-Economic Family Socio—Economic Skilled . . . . . . Family Socio-Economic skilled . . . . . . 1874—75 Massachusetts Characteristics, Profes- Characteristics, Self- Characteristics, Clerical Characteristics, Skilled. Characteristics, Semi- Characteristics, Un- O O 0 0 0 0 O O O O 0 O O Expenditure Survey Con- sumption Patterns of Families and Wage Earn— ers in 15 Cities and Towns by Income Class. . 1888-91 Survey of Consumer Expenditures Con- sumption Expenditure of Families of City Wage and Clerical Workers of 2 or More Persons, Income Class. . . . by O O O O O O 0 O O O O O O 1901 Consumption Expenditure Study of "Normal Families” in Principal Industrial Centers in 33 States, by Income Class. . . . . . . . . . 1917-19 Consumption Expenditure Study of White Workers' Families with at Least One Child, in Cities of All Sizes, by Income Class . . . 1934-36 Consumption Expenditure Study of Fami— lies of Employed Workers in Cities of 50,000 and Over, by Income 1950 Study of Consumer Expenditures, Class 0 O O O O O O O O O Incomes and Savings of Wage-Earner and Clerical- Worker Families in Cities of 2,500 and Over, by Income Class . . 0 O 0 O O O O O 0 0 O O C Family Food Expenditures Away from Home by Occupational Socio-Economic Rank and Age Class . . . . . . . O 0 0 O 0 O O O O O O O 0 Family Alcohol Expenditures by Occupational Socio-Economic Rank and Age Class . . . . . . Page 286 289 290 291 292 294 296 297 298 299 300 301 303 304 Table 54. 55. 56. Family Tobacco Expenditures by Occupational Socio-Economic Rank and Age Class . . . . Family Recreation Expenditures by Occupational Socio-Economic Rank and Age Class . . . . Family Personal Care Expenditures by Occupa- tional Socio—Economic Rank and Age Class. Page 305 306 307 CHAPTER I BACKGROUND AND PROBLEM DELINEATION Marketing management for many years has attempted ‘to understand more accurately the characteristics of exist— ing or potential markets. In recent years marketing manage- ment has placed even greater emphasis on the study of com— plex consumer and market behavior. For the most part, the increase in the necessity for understanding consumer and market behavior has resulted from increased competitive activities in the market frequently reflected in narrowed profit margins, conglomerate market competition, and higher marketing costs. As a consequence, many firms have attempted to develop some basis of "differential competitive advantage" by improving their market delineation methods. The rationale has been that they are likely to direct the marketing process with greater certainty and efficiency if their markets are delineated more accurately. Background Basic Approaches to Market Delineation There are two broad fundamental methods of delineat- ing markets. These are described below. 1. By geographical boundaries. Many markets are delineated according to geographical tastes and habits. For some products there is an "Eastern market" or a "South— ern market.“ Another example where geographical market delineation is used is in retailing. The immediate environ- ment of a supermarket or department store may compose most of its market. From estimates of the needs and wants of the retail outlet's potential customers a product "mix" for the outlet can be developed. 2. By socio-economic characteristics of the con— sumer. The socio-economic approaches to market delineation are based on specific characteristics of consumers. Research evidence indicates that consumers with similar socio-eco— nomic characteristics often have similar expenditure pat- terns. The socio—economic approach also usually underlies the geographical market delineation method. Examples of the socio—economic approach to market delineation include delineating markets by income class, age, sex, social class, and family life cycle stage.1 The increasing awareness of the need for improved market delineation methods is noted below: As discretionary income rises and firms' competi- tive struggle to increase their share of this in— come intensifies, interest in how to make promotion- al efforts more effective is expanding. One approach lRobert Mainer and Charles C. Slater, "Markets in Motion," Harvard Business Review, XLII (March-April, 1964), 75-82; and Daniel Yankelovich, "New Criteria for Market Segmentation," Harvard Business Review, XLII (March-April, 1964), 83-90. toward this end is the use of market segmentation to "customize" a particular marketing mix for a particular segment of the market. The greatest shortcoming of this approach is the inability of sellers to identify homogeneous segments of con— sumers. Income is an effective basis of segmenta- tion for some products but not for others. Phys— ical location of consumers is of some help, but regional differences become less and less important in our peripatetic society. . . .1 The inherent advantages of delineating markets likely will have even more significant effects on future marketing strategy. For example, a market research executive, Ollie G. Crump, recently forecast that national advertising for many consumer goods will be deemphasized in the future.2 This potential trend, according to Crump, is the result of the increasing heterogeneity of consumers. As a consequence, more of the corporate effort will be devoted to delineating the pertinent market sectors. By investigating consumer behavior, market research- ers can determine whether or not consuming groups demonstrate expenditure patterns significantly different enough for market delineation.3 When groups having similar expendi- ture patterns are identified and the marketing effort is 1James M. Carman, The Application of Social Class in Market Segmentation (Berkeley: Institute of Business and Economic Research, the University of California, 1965), p. 1. 2American Marketing Association, News Bulletin, XXII, No. 1 (Chicago: American Marketing Association, July, 1965), 11. . 3John A. Howard, Marketing Management, Analysis and Planning_(Homewood, I11.: Richard D. Irwin, Inc., 1963), pp. 31-320 directed toward them, then, the firm's marketing process will likely be improved. A key problem, however, is deter- mining which consumers should be delineated and what vari- ables to employ in delineating them. The Function of Socio-Economic Classificatory Variables The basic function of socio-economic Classificatory variables is to aid in identifying, classifying, and dif- ferentiating consuming groups. It would be ideal if a firm could rely on a single variable to accomplish the above tasks. Rarely, however, is the single variable sufficient for market delineation since a multivariate problem exists. Consequently, many variables have to be studied in order to obtain a few variables that accomplish the desired dif— ferentiation of consumers. The classification process, if effective, identifies and delineates "homogeneous" groups according to predetermined socio—economic character- istics. Illustratively, the market researcher might use either a single socio—economic variable such as income class or a combination of several socio-economic variables, for example, age, income, and occupation. The market researcher's decision to employ single or multiple socio—economic Classificatory variables usually rests on several factors; for example, the decision may depend upon the nature of the firm's market. Usually the markets for "staple” goods are broad and encompass several heterogeneous segments of the general population, hence, broad market delineation may suffice. By contrast, the market for "luxury" goods may need to be defined with ex— treme precision. The number of Classificatory variables used may also depend simply upon how well the firm desires to define the markets for its products at a given point in time. Often the desire for such Classificatory know- ledge is weighed against the economic costs of delineating the market since market delineation costs tend to increase sharply as Classificatory variables are added. Still an- other consideration is that if too many variables are used, the differentiated sub markets will be too numerous and the potential customers in each too few for individual cul- tivation. Cummings, in employing an analogy applicable to market delineation, states: Statistical classification in general may be likened to the process of focusing a camera: the image becomes blurred equally by moving the focusing lens too close or too far from the focusing screen, and becomes clear at precisely one point only.1 An almost infinite number of unique consumer socio- economic classifications can be used to delineate markets. Illustratively, in the 1960 Census of Population, in addi- tion to other socio-economic and demographic variables, there are 2 sex classes, 7 age classes, 11 occupational lJohn Cummings, Fundamentals of Statistics (Scranton, Pa.: International Textbook Press, 1923), p. 9. categories, 11 income classes, 8 educational categories. The total number of delineated classes using age and sex would be 14 (2 x 7). If the age-sex classes were further disaggregated according to the 11 occupational categories, then the number of classes necessary to describe the popu— lation would be 154 (14 x 11). The consequence of continu- ally adding socio-economic variables is illustrated in Table 1. TABLE 1 POSSIBLE NUMBER OF DELINEATED MARKET CLASSES USING SEVERAL SOCIO-ECONOMIC VARIABLESa Number of Classes In Combination Socio—Economic For the with Preceding Characteristics Characteristic Characteristics Sex 2 2 Age 7 14 Occupation 11 154 Income 11 1,694 Education 8 13,552 aSource of socio-economic characteristics and num- ber of classes for the characteristic from U.S. Bureau of the Census. U.S. Census of ngulation: 1960 General Social and Economic Characteristicsi_Michigan, Final Report PC (1)424C (Washington: U.S. Government Printing Office, 1962). Table format adopted from John Cummings, Fundamentals of Statistics (Scranton, Pa.: International Textbook Press, 1923), p. 17. As illustrated in Table l, delineating markets by several socio-economic Classificatory variables can become exceedingly complex and costly. Therefore, the selection of the proper Classificatory variable or variables is critically important since some variables may provide a basis for delineating markets accurately, but cannot be operationalized economically. Therefore, a primary step in market delineation is the development of variables which are operational and meaningful. Develgpment of Market Delineation Criteria Generally, the socio—economic variables used in market delineation have been developed by using both micro and macro approaches. The micro approaches have usually been developed by behavioral scientists to study human be- havior and characteristics. From their research, various socio-economic variables have been identified to differen- tiate groups of individuals possessing relatively homogene- ous psychological, social, and economic characteristics. Frequently, the differentiated groups have been shown to exhibit similar consumption patterns. By generalizing from the micro findings of individuals and small groups, projec- tions on how larger aggregates of consumers behave have been made. A potential shortcoming of the micro approach is that the samples drawn may not be representative of a larger population. And large samples are often not econom— ically feasible for the sponsoring firm. It should be pointed out, however, that in some cases relatively small samples have proven to be reliable in predicting aggregate behavior. Another limitation is that the data collection instruments used may have been developed and utilized for other purposes. The Thematic Apperception Test, for exam- ple, has been used in studying the psychological makeup of consumers though it was developed for use in clinical psy- chology. The macro approach is concerned with aggregate ex— penditure patterns of large national and regional samples. A limitation applying to micro analysis applies to macro investigations but in reverse: The findings based on the aggregate may not be applicable to the individual consuming unit. In many cases, however, in market delineation or market sector analysis, the cost of investigating all indi- vidual expenditure patterns would be prohibitive. Market researchers, moreover, are primarily interested in aggre- gate behavior since the marketing effort usually is directed toward groups or market sectors. One approach to macro analysis, basically a deduc- tive one, examines "ex post" consumer expenditure patterns. By examining the expenditure patterns of different groups of consumers disaggregated by different socio—economic vari- ables, potential insights into macro consumer behavior can be gained. An important advantage to the marketer of the “ex post" approach is that knowledge of the complexities of consumer motivation is not always a necessary condition for predicting and observing market behavior. McNeal posits that market researchers usually do not need an explanation of the "whys" of consumer behavior, rather, attention should be devoted to examining the "patterns of consumer behavior." While this may hold at the macro level, some market research— ers would argue that the "why" of consumer behavior becomes critically important to the marketing strategy of the indi- vidual firm. A wealth of "ex post" descriptive consumer expendi- ture statistics has become available since 1950.2 From a marketing delineation standpoint, the statistics are poten— tially important for two primary reasons: (1) the statis- tics may provide information that can be employed to "test and operationalize" various socio-economic market delinea— tion variables, and (2) the statistics may also provide a framework which can be employed for market prediction. Evaluation of Socio-Economic Market Delineation Variables An evaluation of socio—economic variables as a means of stratifying population rests upon several criteria.3 lJames U. McNeal, "Consumer Behavior—-Introduction," Dimension of Consumer Behavior, ed. James U. McNeal (New York: Appleton-Century-Crofts, 1965), p. 6. 2University of Pennsylvania, Study of Consumer Ex— penditures, Incomes and Savings, Vols. I-XVIII (Philadel— phia: University of Pennsylvania, 1956); and United States Department of Agriculture, Food Consumption of Households in the United States (Washington: U.S. Government Printing Office, 1956); and Life, LIFE Study of Consumer Expenditures (New York: Time, Inc., 1957); and U.S. Department of Labor, Bureau of Labor Statistics, Survey_of Consumer Expenditure and Income, 1960-61. 3A more comprehensive discussion of evaluating cri- teria for the use of social class in marketing research can 10 Several evaluating criteria are described below: 1. Operational characteristics. Is the socio- economic variable practical in terms of implementation? Are data on the population's socio—economic characteristics and expenditure patterns difficult to obtain? What is the cost factor in implementing and operationalizing the cri- teria? 2. Monitoring of changes in the socio—economic variable. Is there monitoring of the socio-economic char— acteristics of the population to note important changes affecting market delineation? Income classes of the popu- lation are periodically monitored by governmental sources while social classes are not generally monitored. 3. Reflection of the influence of other variables. Does the variable reflect the influence of other socio- economic and demographic variables which influence expendi— ture decisions? Social class stratum usually reflects the influence of several factors on expenditure patterns, such as occupation, income, education, and dwelling area. How- ever, income class may not reflect the influence of other socio-economic variables affecting expenditure patterns. If a variable is orthogonal, that is, uncorrelated with other variables and relevant to classification, it is very be found in Social Class Definition in Market Research Ob- jectives and Practicei A Report Prepared by the Working Party on Social Class Definition (London: The Market Re- search Society, 1963), p. 22. 11 useful in market delineation; however, few orthogonal vari- ables exist. The next best alternative is to identify vari- ables which encompass other variables in a known way and which have Classificatory utility. Such variables make for economy of description, classification, and differen- tiation. 4. Homogeneous meaning of the variable. Does the socio-economic variable have a homogeneous meaning between market areas? Age, sex, and income classes have a homogene— ous meaning; however, social class strata may not. Conse- quently, social class data may be difficult to apply between market areas. 5. Predictive value. How accurately can the socio- economic variable predict present expenditure patterns? Can the variable be employed as a basis for estimating changes in future expenditure patterns? Preliminary Evaluation of the Age-Occupation Market Delineation Method Table 2 illustrates some of the potential strengths and weaknesses of several socio-economic market delineation variables. The evaluation table is most likely to apply best to large retailers and to multi-product firms. The evaluation of the several socio-economic market delineation variables reveals that the multi-variable approach-—age and occupation of the family head--appears to have several potential advantages in delineating markets. Many of the 12 .xuommumu HMCOHDMQDUUO am no £05m .omumomummmmflo ma :ddoum mdomcmmoeonz m umsuwcz paw mcumyumm mmaa >Haemw one memomuomMmHo op owns oonqu one does mpcmdmo mmHQMHum> UHEOdoumIOHUOm umzuo mo mucmSstH 03D Dumammu on mammamcm mHumo mmaa xaHEmm mo >DHHHQM when .HH umudmnu CH Namumadeou QMOE ommmDUmHo mum moaumflumuumumcu maumpfluu w>Hpm3Hm>m mru paw coaummoflamo umxume UflEOCOUMIOHuom Mo moonpme mSOHum> wrap mow mm» mow mm» mm» COHDMQSUUOI00< uHDUHMMHQ oz mmEHumEom oz Danuawwflo mmMHU Hmauom coauflcflmmo co mocmdmo oz Qmmeflpmeom mmEHumEom Daduamwflo mauxu mead maflemm moonumz mHQMHum>IHuHSZ .m mmefipweom wow wow >Ham5mo mm» coauMQduuo mmEHquom mow omeuowmflo 02 mm» xmm omuuoumflo mm» UODHOhmHQ mmEHumEom mm» ommm haflemm MO 004 omuuoumflo mm» ompuopmflo 02 wow mmmmmHU mEOUCH mpmomuoo< moospmz menmeum> memcam .< NDHHHQ< ocflcmmz mmHQMHum> mmHQMHum> coapmucmEdeEH coapmmcaamo coauUHomum msomcmooeom UHQQBMOOEmQ CH mo umOUmmUDHUCH umxumz UHEocoum ocm UHEOdoum mmocmru mo mUHumHuwuumumrU loauom Mo oonpmz loauom umruo OCHMODHcoz HmCOADmumdo muumawmm mSOSCHuCOU muwumflumpumumnu maumufluu m>aum5Hm>m mmqm ZOHBM m N mqm U > —— O -—I——-O ' I Horizontal Stratum E I AGE CLASS I Horizontal Stratum F V \V I AGE CLASS Figure 2. Schema of vertical and horizontal market delin— eation 16 than social class analysis.1 To disaggregate a population according to the usual social class methods normally requires interviewing a large sample. A few researchers have pointed out that differences in incomes, expenditures, savings, life styles, and family life patterns exist among the vari- ous social class strata.2 A weakness of social class anal— ysis, however, is that each stratum is not disaggregated by family life cycles. Such differences appear critically important in market delineation since a given social class stratum may have both younger and older families which could affect the validity of the variable for prediction purposes. A Similar limitation also holds when the general population is disaggregated into family life cycle classes. In delin- eating the general population into family life cycles, the social class factor is ignored.3 Therefore, in many cases to optimize the market delineation methods both the social class and stage in the life cycle of the family should be lForsocial class analysis in marketing see: Richard P. Coleman, "The Significance of Social Stratifi- cation in Selling," Proceedings of the 43rd National Con— ference of the American Marketing Association, ed. Martin L. Bell (December, 1960), pp. 171-84; and Pierre Martineau, "Social Classes and Spending Behavior," Journal of Market— Egg, American Marketing Association, XXIII (October, 1958), 121—30. The strengths and weaknesses of employing social class as a market delineation method are discussed in Chap- ter II. 2Ibid., pp. 174-84. 3For life cycle analysis see: Lincoln Clark (ed.), Consumer Behavior, Vol. II: The Life gycle and Consumer Behavior (New York: New York University Press, 1955). For other problems in the implementation of life cycle analysis see Chapter II. 17 considered. 6. Occupational categories are comparable between Standard Metropolitan Statistical Areas, cities, or other urban areas or regions since the commonly used occupational definitions and categories have a homogeneous meaning. 7. Occupation may also serve as a useful proxy variable for social class. In determining social class placement occupation of the family head is almost always employed as a key variable. Warner recognized the impor— tance of occupation in the determination of social class and found occupation highly correlated with his prediction of an individual's social class (r=.91).l Following Warner's investigation, Kahl and Davis also correlated occupation to social class placement and again found occupation highly correlated with social class estimates.2 But, in determining social class by the method gen- erally recommended for marketing and economic purposes, other factors are employed simultaneously with occupation. For example, source of income, housing type, and dwelling area are used in determining social class position by War- ner's Index of Status Characteristics method. According 1W. Lloyd Warner, Marcia Meeker, and Kenneth Eells, Social Class in America (Chicago: Science Research Asso— ciates, 1949), p. 168. 2 Joseph A. Kahl and James A. Davis, "A Comparison of Indexes of Socio-Economic Status," American Sociolggical Review, XX (June, 1955), 317—25. 18 to Warner, "the Index of Status Characteristics is, pri- marily, an index of socio—economic factors; but . . . it can be used with a considerable degree of confidence as an index of social-class position as well."1 A priori reasoning suggests that the variables, source of income, housing type, and dwelling area are in reality dependent upon the occupational category of the family. Therefore, why measure each of the variables since occupational rank is so highly correlated with social class prediction? As noted above, the socio-economic variables, age and occupation, appear to be useful in delineating the pop- ulation into relatively homogeneous groups. If the age- occupation concept is useful for market delineation pur- poses then it must reflect significant differences in fam— ily expenditure patterns since the concept of market delin- eation is based on the notion that many consuming groups can have different expenditure patterns. The Problem Statement of the Problem The primary objective of this research is to in- vestigate expenditure patterns, using both age and occupa— tion of the family head as market delineation variables. The potential utility of the age-occupation approach war- rants an investigation to determine whether or not income, lWarner, Meeker and Eells, op. cit., p. 39. 19 expenditure, and savings patterns do vary within and among family age classes of several occupational categories. In connection with the present investigation, sev— eral questions having potential market delineation impli— cations are explored. Some of the questions are: 1. How do income, saving, and expenditure patterns vary among family age classes of each occupational category? How do income, saving, and expenditure patterns vary among various age classes of different occu— pational categories? Should some of the occupational categories or fam- ily age classes be combined for market delineation and prediction purposes? How do income elasticities for different expendi- ture accounts vary among family age classes of an occupational category? How do income elasticities among family age classes for various expenditure accounts of an occupation vary from the elasticities of other occupational categories? How do family saving patterns vary within and among occupational-age categories? How do average propensities to consume and save vary among the age classes of various occupational categories? 20 By examining the preceding questions, the research can determine the influence of age and occupation of the family head on family expenditure patterns. Hypotheses The major hypothesis of the study is that the socio- economic classificatory variables——age and occupation of the family head—-ref1ect Significant differences in family expenditure patterns. The research is also designed to test the following specific hypotheses: I. Socio-Economic Occupational Rank and Absolute Expendi- ture Relationships A. In each family age class average total dollar ex- penditures are positively related to the socio- economic rank of the occupation.1 The higher the occupational rank, the greater the total expendi- tures are for families in each age class. B. In each family age class average total food expendi- tures are positively related to the socio-economic rank of the occupation. C. In each family age class average housing expendi— tures are positively related to the socio-economic rank of the occupation. D. In each family age class average clothing expenditures 1 . . . . An explanation of occupational SOCio-economic rank is presented in Chapter III. 21 are positively related to the socio-economic rank of the occupation. Family expenditures for housefurnishings and equip— ment reach maximum levels in the primary age classes for each occupational category.l Family expenditures for housefurnishings and equip- ment do not follow the socio-economic occupational ranking scale in any age class. Family automobile expenditures do not correspond to the socio-economic occupational ranking scale in any age class. Maximum automobile expenditures for each of the manual2 occupational categories are concentrated in the primary age classes while maximum automobile expenditures for each of the nonmanual3 occupational groups are concentrated in the secondary age classes.4 II. Socio-Economic Occupational Rank and Relative Income- Expenditure Relationships A. In each family age class the average propensity to lPrimary age classes are those in which the family head's age is under 45 years. skilled, 2Manual occupational classes include skilled, semi- and unskilled workers. 3 . . . Nonmanual occupations include profeSSionals, man- agers, and officials, the self-employed, and clerical and sales workers. Secondary age classes include those in which the family head's age is over 45 years. 22 consume for total family expenditures is inversely related to occupational socio—economic rank.1 The higher the occupational socio-economic ranking, the lower the average propensity to consume in each age class of the occupation. That is, the research assumes that occupations with higher socio-economic rank also will devote more income to savings. B. In each family age class the average propensity to consume for the total of expenditure accounts classified as "necessities" is inversely related to occupational socio-economic ranking. C. In each successive occupational age class the pro— portion of income devoted to total food expenditures increases in the primary age classes and decreases in the secondary age classes. D. In each family age class of all occupations the average propensity to consume for total food expendi— tures is inversely related to the socio-economic rank of the occupation. 1The average propensity to consume (APC) is the ratio of mean disposable income to a family expenditure category. 2In the study "necessities" are defined as food, clothing, housing, and fuel, light, refrigeration, and water. For an alternative definition see: Donald F. Blankertz, "A Marketing Analysis of Suburban and Urban Expenditure Patterns,“ Theory in Marketing, ed. Reavis Cox, Wroe Alder- Son, and Stanley J. Shapiro (Homewood, 111.: Richard D. Irwin, Inc., 1964), p. 297. III. IV. A. 23 In each successive occupational age class the pro- portion of income devoted to clothing expenditures increases in the primary age classes and decreases with family age in the secondary age classes. In each family age class of all occupations the proportion of income devoted to housefurnishings and equipment is inversely related to socio—economic occupational rank. Occupational Rank and Family Disposable Income Con— centrations Disposable income levels for all nonmanual occupa— tional families reach maximum levels in the second- ary family age classes. Disposable income levels for all manual working families reach maximum levels in the primary family age classes. Absolute Savings Levels and Family Age Class A. In each family age class of the nonmanual occupa- tional categories the family savings levels do not follow the socio-economic ranking scale. In each family age class of the manual occupational categories family absolute savings levels follow the socio—economic ranking scale. Socio-Economic Rank and Income—Expenditure Elasticities A. The magnitude of income elasticity coefficients for total expenditures within each age class is inversely related to the socio-economic rank of 24 the occupation. The magnitude of income elasticity coefficients for total food expenditures within each age class is inversely related to the socio-economic ranks of the occupation. The magnitude of income elasticity coefficients for clothing within each age class is inversely related to the socio—economic ranks of the occupa- tion. The magnitude of income elasticity coefficients for total housing expenditures within each age class is inversely related to the socio—economic rank of the occupation. VI. Marginal Income—Expenditure Relationships A. In each family age class, the marginal propensity to consume for total expenditures is inversely re- lated to occupational socio-economic rank. The higher the occupational ranking, the smaller the marginal propensity to consume for total expendi— tures. Method of Apppoach The research employs unpublished cross-sectional data to examine variations in income, savings, and expendi- tures of occupational family age classes. The unpublished data are part of the 1960-61 national urban Bureau of Labor Statistics' Survey of Consumer Expenditures. Specifically, 25 the data utilized in this research represent family expendi- tures in nine Standard Metropolitan Statistical Areas in the North Central and Eastern regions of the United States. Each Standard Metropolitan Statistical Area selected has a population that exceeds 1,400,000. Further, each Stand- ard Metropolitan Statistical Area is characterized as being located in heavily urbanized areas, and most have a heavy concentration of industry, similar climatic conditions, and a relatively similar occupational distribution. The data utilized in this research are disaggregated into six occupational categories. And each occupational category is further disaggregated into six chronological age classes. For each occupational—age classification several family expenditure accounts are examined. The sample utilized in the research contained 2,168 families of two or more persons each. Since the sample employed in this study was an ag- gregation of nine sets of Standard Metropolitan Statistical Area data, the data from each Standard Metropolitan Statis- tical Area had to be weighted and averaged. The initial step in constructing the sample entailed placing the orig- inal data from each Standard Metropolitan Area on data processing cards. The next step involved weighting each occupational-age class of each Standard Metropolitan Sta— tistical Area by the number of sample families in each occupational—age class. Then, the dollar value of each 26 family expenditure account in each occupational-age class was multiplied by the number of families in each occupational- age class. The next phase consisted of adding the totals of the number of families times the value of the family ex— penditures for each of the nine Standard Metropolitan Sta— tistical Areas and dividing by the total number of families in the occupation-age classes of the nine Standard Metro— politan Statistical Areas. The computations result in a sample representing families in all nine Standard Metropol- itan Statistical Areas. From the resulting weighted sample, computer programs for the marginal propensities to consume, average propensities to consume, and average propensities to save were constructed. To compute income elasticities for selected family expenditures within each occupational— age class, a computer program employing regression analysis was employed. First, the elasticity measures and the mar- ginal propensity to consume are used to determine the em— phasis families place on selected goods in the different age classes. Second, the average propensity to consume is computed to measure the relationship between average disposable income and each family expenditure account with- in each occupation-age class. The average propensity to save, the reciprocal of the average propensity to consume, also, is computed to measure savings or dissavings within each occupational-age class. Third, the absolute dollar 27 values of income and expenditures in each occupation—age class are examined to note the absolute differences in con- sumption patterns of each occupational—age class. Limitations Certain limitations of the study and data should be recognized. 1. The study examines the expenditure patterns of six broad occupational categories used by the Bureau of Labor Statistics. The grouping of several occupational titles into a few broad occupational classes can result in overaggregation. For example, in the occupational cate— gory, clerical and sales workers, a wide range of incomes is normally found which can affect the predictive value of the average incomes and expenditures for the aggregated occupational class. This variation must be considered in interpreting the findings of the study. 2. Certain limitations apply to chronological in— terpretations of cross-sectional data. That is, consumers may not always react in their economic life patterns as indicated by cross—sectional analysis. Perhaps a more ac- curate approach would entail investigating a large group of families for several years and noting how their income— expenditure relationship changes over time. However, such an approach is difficult, costly, and time consuming. Irving B. Kravis, in addressing the above limita- tion of cross-sectional budget analysis, states: 28 On the economic side we are faced with the all too familiar gap between the construct of pure theory and the statistical approximation that is available. We would like to know how the same family or group of families would behave at instantaneously alter- native levels of income. Budget data for any one period of time shows up how different families be— have at several income levels. The difficulty can be minimized by ensuring that all the families, at whatever income level, are homogeneous with respect to the social and demographic variables that influ- ence demand relationships.1 In the present study an attempt has been made to insure as much homogeneity within the groups as possible. Illustratively, the occupational categories have been dis— aggregated into six family age classes. This disaggrega- tion procedure is expected to increase the homogeneity of the groups studied. 3. The unpublished Bureau of Labor Statistics' data employed in this study are also subject to errors that may influence the findings. These limitations have been expressed by the Bureau of Labor Statistics: All data have been reviewed, edited, and screened to minimize processing errors. Chance variations due to sampling can be measured statistically and the BLS is planning to estimate the sampling error for selected items and groups of items. Approxi- mately 75 percent of the families in the 1960 sample furnished usable schedules and some of the nonre- spondents supplied limited information on family lIrving B. Kravis, "Expenditure-Income Relation- ships for Consumers Durable Goods and Problems in Their Derivation," Proceedipgs of the 115th Annual Meeting of the Business and Economic Statistics Section of the Amer— ican Statistical Association, ed. Walter F. Ryan (American Statistical Association, 1955), p. 106. 29 characteristics which will be used to evaluate the nature of the sample losses due to nonresponse. Among the participating families, inaccurate report- ing is a source of error despite continued research in schedule design and intensive training of the interviewers. Such inaccuracies result from memory errors, misunderstanding of a question or reluctance to answer it, and incorrect entries by the inter— viewer. Although the BLS has accumulated substan— tial knowledge about such reporting errors and will continue research in this field, these errors can not be quantified satisfactorily.l 4. The time between the Bureau of Labor Statistics' study and its subsequent release may also be a minor lim— itation. The data utilized in this study, although collect— ed in 1961 and 1962, were not released until 1965. However, the National Industrial Conference Board, in its publica— tion Expenditure Patterns of the American Family, utiliz- ing basic Bureau of Labor Statistics data, made the follow— ing statement. . . . the lapse of a reasonable period of time in a survey of this type does not significantly affect the usefulness of the results. Since the period in which the survey was in the field, average con- sumer expenditures increased but the evidence sug— gests that the difference in expenditure patterns among the various classifications of families . . . remained relatively constant. Change tends to be uniform. For instance, the variations in outlays for food by age of family head reported in this study are very similar to the variations reported in the Bureau's study conducted ten years earlier. . . .2 lU.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditures and Income Total United States, Urban and Rural 1960—61, BLS Report No. 237-93 (Washington: U.S. Government Printing Office, 1965), pp. 5-7. 2National Industrial Conference Board, Expenditures Patterns of the American Fami1y_(New York: National Indus— trial Conference Board, 1965), p. 153. 30 5. The study is not meant to be representative of the entire United States population, only to cities of 1,400,000 or more population in the North Eastern and North Central regions of the United States. However, it is reason— able to assume that the selected sample utilized in this study is generally representative of other urban areas in the United States. As a result, generalizations can be drawn from the findings of the study which are applicable to other urban areas in the United States. Some Possible Contributions The development of market delineation methods has generated considerable interest in recent years. As a re- sult, several approaches to market delineation have been suggested for improving the market delineation task. Never- theless, most of the approaches demonstrate significant limitations affecting implementation and utilization.1 The focus of the present research is to investigate an alternative market delineation approach which may elim- inate many of the implementation and utilization problems of other market delineation approaches. The age—occupation approach to market delineation is expected to be relatively accurate in disaggregating consumers into relatively homo— geneous socio—economic groups. If demonstrated that 1The limitations of several market delineation ap- proaches will be explained in Chapter II. 31 expenditures vary among the different occupational categories and between age classes of an occupational category, the market delineation process may be improved. By knowing the importance the different occupations and the age classes within an occupation place on various expenditures, several applications are expected. One of the most significant applications may be in building a model of consumption be- havior based on the occupation-age variables which can be usedftm predicting future macro market expenditure patterns.l By predicting and forecasting expenditure patterns, firms may more accurately plan marketing programs, advertising appropriations, sales effort allocation, and the firm's distribution networks. Finally, it is possible that the study will provide a basis for developing hypotheses of future market delinea— tion research. Organization The remainder of the study is organized into four chapters. Chapter II presents a review of the contributions of several researchers of consumption expenditures. The review also examines the influence of income and other socio-economic variables on family expenditure patterns. lU.S. Department of Labor, Bureau of Labor Statis- tics, "Uses of Family Expenditure Data," Report No. 238-13, August, 1965, (Washington: U.S. Government Printing Office), p. 9. 32 Finally, the review presents several of the most commonly utilized socio-economic approaches of market delineation to determine the strengths and weaknesses of each method. An analysis and evaluation of the data used in the study and the research design are presented in Chapter III. The findings of the research are presented in Chapter IV. Chap— ter V contains the summary and conclusions, as well as sug- gestions for future research. CHAPTER II REVIEW OF LITERATURE Orggnization and Rationale Since the present study is focused on consumer be- havior and market delineation, a review of the literature was made relating the effect of several socio-economic vari- ables on consumer expenditure patterns. The primary objec— tives of the review are twofold: First, there has been little research undertaken attempting to explain or illus- trate differences and similarities in expenditure patterns of different occupations disaggregated by age classes. Therefore, a review was made of several other socio-eco- nomic variables likely having similar influences on occu- pational expenditure patterns. This review is used in ex- plaining and understanding the findings of the present re- search. The second objective of the literature review is to evaluate the employment of several socio-economic vari- ables in delineating markets. Influence of Selected Socio-Economic Variables on Family Expenditure Patterns There are several socio-economic variables influ- encing family expenditure patterns. To understand consumer 33 34 behavior, the nature and effect of the different socio— economic variables on expenditure patterns must be consid- ered. Several studies have revealed in part the influence of socio-economic factors on consumption patterns. Usually, in order to determine the influence of socio-economic vari— ables on consumption patterns, the researcher constructs a "structural model."1 The function of the model is to relate the effect of several socio-economic variables on family expenditures. A model expressing the relationship of socio-economic and demographic variables on expenditures is:2 Y = f(xl,x2,x3, ... xn). The above model denotes the relationship between Y, the dependent variable, (expenditures) and the independent X variables X and Xn. The independent variables can 1’ 2’ include income, family size, family life cycle, occupation, race, age, and social class. The problem normally involved in determining the effect of socio-economic variables on expenditure patterns is determining what variables to ex- amine. In addressing the above problem, Robert O. Herrmann asserts that classical economic theory does not provide a lRobert Omer Herrmann, "Household Socio-Economic and Demographic Characteristics as Determinants of Food Expenditure Behavior" (Unpublished Ph.D. dissertation, Michigan State University, East Lansing, 1964), p. 9. 21bid., p. 10. 35 framework for the inclusion of all socio-economic variables, only income. Therefore, other disciplines frequently must also be utilized in determining what variables to examine. For marketing purposes, to understand consumer behavior and market delineation, the selection and determination of significant independent variables is of critical impor- tance. Based on past empirical research, the following socio—economic variables appear to significantly influence family expenditure patterns. 1. Family income 2. Change in family income 3. Age of household head and stage in family life cycle 4. Family size 5. Social class and occupation Much of the research determining the influence of the above variables on expenditures has been conducted by economists, market researchers, home economists and agri- cultural economists. The agricultural economists and home economists are primarily interested in the effects of socio— economic variables on food expenditure patterns. Several findings of prior empirical research on the variables are reviewed below. Income as a Determinant of Family_Expenditures Since the present study utilizes cross-sectional lIbid. 36 budget data, a review was made of several consumer budget studies. The primary purposes of the review were: (1) to examine the contributions of pioneering budget analysts and (2) to note the effect of income and other socio-economic variables on family expenditure patterns over time. The usefulness of cross-sectional data, for understanding con— sumer behavior, is increased if basic income-expenditure relationships remain relatively constant over time. There— fore, several cross-sectional budget studies are examined in order to determine the stability of family expenditures. Foundations of Family Budget Analysis Since society has usually been concerned with its standards of living, a number of "living standard" studies has been undertaken during the past two hundred years.1 Reviewing some of the studies through time primarily reveals two changes: (1) the purposes of budget studies have changed and (2) the scope of the studies has been enlarged. Most of the pioneering budget studies usually examine all broad occupational categories.2 Another significant development 1Much of the historical development of family bud- get analysis has been drawn from Carle C. Zimmerman, 92p: sumption and Standards of Living_(New York: D. Van Nostrand Company, Inc., 1936), Chaps. 1—17. As far as could be de- termined, LePlay's consumption studies have not been pub— lished in English; therefore, the background information on LePlay relies on Zimmerman's study. 2University of Pennsylvania, Study of Consumer Ex— penditures, Incomes and Savings, Vols. 1—18 (Philadelphia: University of Pennsylvania, 1956); and Life, Life Study of 37 in cross-sectional budget studies conducted by the govern— ment has been to provide a "benchmark" for the formulation of the Consumer Price Index aiding the measurement of the "cost of living." Some of the most significant contributors to family budget analysis have been Frederic LePlay, Ernst Engel, and Carroll Wright, each making unique contributions to consumer budget analysis and directly or indirectly aiding in the formulation of several "laws" of consumption.1 The "laws” not only relate income to consumption and saving, but indirectly reflect the significance of other socio- economic variables on consumption. Frederic LePlay LePlay's contributions to budget analysis and to the social sciences in general have influenced man's think- ing on societal well-being for nearly 150 years. LePlay and his followers conducted budget studies in several coun— . 2 . . . . tries. Zimmerman, in an extremely comprehenSive analySis Consumer Expenditures (New York: Time, Inc., 1957); and U.S. Department of Labor, Bureau of Labor Statistics, Sur— vgy of Consumer Expenditures and Income, 1960—61. 1 . . . . The ”laws“ of consumption as discussed in this study refer to income-expenditure generalizations based on empirical research. 2Zimmerman, op. cit., p. 418. According to Zimmer- man, the LePlay school conducted family budget studies in Algeria, Arabia, Austria, Bali, Belgium, California, Canada, China, Corsica, Egypt, Great Britain, the United States and several other countries. 38 of LePlay's works, reports that LePlay was primarily inter- ested in the economic and social well-being of a Society. To study the well—being of society LePlay used family bud- gaxsas a framework for his analysis. LePlay largely studied the relation of the stand— .ard of living to the social structure. He early formulated the conception that if one knew the en- tire family budget one could describe the kind of family. Another idea was developed from this: if one knows the type of family one could understand 1 the total social structure of which it was a part. LePlay's budget studies were primarily limited to working class families. To interview families about their expenditure patterns LePlay developed a comprehensive family classification system to record and classify expenditures. Zimmerman notes that LePlay's Classificatory systems "are still models in methodology."2 LePlay's classification system as presented by Zimmerman is noted below: 1. Place. This includes a discussion of the geo— graphical location of the particular family studied, and the type of industry and family in that area. In this section is presented the rationale for choosing the particular fam- ily and some discussion of its representative— ness. ~ 2. Civil state. This section lists the members of the household by age and sex, gives their age at marriage, and includes data concerning any children who have died. . . . 3. Religion and moral habits. Here the formal religion of the family is given, and the degree lIbid., p. 424. 21bid. 6-8 0 9-11 0 12-13 0 39 of actual adherence to the forms and tenets of the religion is discussed. Hygiene and health service. This discusses the physical condition of the members, the diseases which they have had, and their recourse to and costs of medical treatment. Rank of the famil . This deals with the posi- tion of the family in the social hierarchy and describes changes in the social standing of the family which tends to improve or depress its occupational (economic) status. These categories are devoted to a discussion of the means by which the family exists. Np. é’is devoted to property, which is divided into real property, domestic animals, and tools or working equipment. Each item is discussed and is given a money value. No. 7 covers subven- tions, by which is meant the individual's rights in connection with the property and services of the community, the employer, or the govern- ment. Each item is discussed and evaluated. No. 8 involves the work and industries carried on by the family. Details are given concern- ing the work of the husband, wife, children, and other family members. If the family car- ries on any "industry," such as the keeping of a cow or a garden, this is discussed in de- tail. These deal with the mode of existence of the family. No. 9 discusses the principal items of food, the number of meals partaken each day, the menus of typical meals, and the time of eating. No. 10 deals with the house, furniture, and clothing. It involves a physical descrip- tion of the home, a list and evaluation of the important items of furniture and equipment, and the clothing for each member. No. 11 ends the discussion of the modes of existence by a characterization of the most important forms of recreation of each member of the family. These deal with the history of the family. No. 12 discusses the principal phases of the family from the birth of the parents until the time of the study. It generally includes data concerning the social status and occupations Of the last generation of the family. No. 13 attempts to characterize the family according 40 to the mores and institutions which seem to ensure the physical and moral well-being of the family. . . . 14. Analysis of income for the year. This is al— ways divided into four sectionS-—property, sub- ventions, wages, and industries--and each one of these is always divided into two parts--the sources of income and the income. . . . 15. The expenses. These are divided into five sec- tions-—food; household; clothing; moral, recre— ational, and health needs; and industry, debts, taxes, and insurance-~with an additional state- ment on yearly savings. Each section lists the expenditure in money or the value of the goods consumed if they were not purchased. The first section concerns food, which is divided into that consumed at home and that away from home. The foods are classified into cereals, fats, milk and eggs, meat and fish, legumes and fruits, condiments and stimulants, and fermented drinks. . . The section on the household includes items of lodging, furniture, heat, and light. . . . The next section concerns moral, recreational, and health needs. This lists detailed expenditures for religion, education of the children, char- ity, recreation and ceremonies, and medical service. Section five lists the expenditures for industry, debts, taxes, and insurance. . . . 1 The total contributions of LePlay are not difficult to evaluate. His comprehensive family Classificatory sys- tems undoubtedly have influenced the present Bureau of Labor Statistics methodological procedures. Illustratively, the Bureau of Labor Statistics is concerned with the geograph- ical location of the consumer, the demographic character- istics, the rank of the family (the occupational category) and the consumption standards of the family. LePlay's fam- ily consumption expense categories are also quite Similar lIbido, ppo 420-320 41 to the one presently employed by the Bureau of Labor Sta- tistics. LePlay's studies, however, were somewhat broader in that they also investigated additional characteristics of the consumer, such as the religion of the family, data concerning deaths in the family, family hygiene, holdings of property, and so forth. LePlay's studies have provided an important bench— mark for economists, sociologists, and anthropologists studying the well—being of society at different points in time. Although LePlay's working class Classificatory system is not completely comparable to the present occupa— tional classification system, it has undoubtedly provided an important benchmark for examining and comparing consump— tion standards.l Ernst Engel Engel was strongly influenced by LePlay's earlier work.2 Much like LePlay, Engel believed that a valid ap— proach of appraising society's well-being was through cross- sectional analysis. Engel, especially interested in apply- ing the statistical method to the social sciences, made several contributions toward understanding consumer behavior. His interest in analyzing previous budget studies, while the director of the Saxony Statistical Bureau from 1850 lIbid., pp. 418-19 and 424. 2Ibid., p. 374. 42 to 1858, led him to observe some systematic relationships between income and expenditures.1 Wright reports that Engel compared the expenditure patterns of three different population classes to note the relative amount of expenditures devoted to different ex- penditures. His population groups were: (1) "a tolerably well—to-do member of the working class, (2) . . . a man whose income is double that of the former, and lastly, (3) . . . a person in easy circumstances."2 From the in- vestigation, reports Wright, Engel found that the "working man“ spent approximately 95 per cent of expenditures for subsistence, clothing, lodging, and firing and lighting. The “man of the intermediate class" spent 90 per cent of total expenditures for the accounts while the “person in easy circumstances" spent 85 per cent of his expenditure budget on the accounts. From the above expenditure rela— tionships Engel formulated his now famous law relating in— come to food expenditures: The poorer an individual, a family, or a people, the greater must be the percentage of the income necessary for the maintenance of physical susten- ance, and again of this a greater portion must be allowed for food.3 lIbid., p. 39. 2Massachusetts Bureau of Statistics of Labor, Sixth Annual Report of the Bureau of Statistics of Labor, Public Document 31 (Boston: Wright & Potter, State Printers, 1875), pp. 437—38. 3Zimmerman, op. cit., p. 101. 43 It is reasonable to assume that the lower income groups will spend a greater percentage of their incomes on neces- sities. But, does the same relationship hold for other expenditures? Following Engel's presentation of his food law, others began to investigate if the law was generally valid for other expenditures. One of the first to do so was Carroll Wright, in 1876, while serving as Director of the Massachusetts Bureau of Statistics of Labor. Carroll Wright As a contemporary of Engel, Wright reinterpreted Engel's Law and published the following four statements which have generally become known as "Engel's laws." 1. That the greater the income, the smaller the relative percentage of outlay for subsistence. 2. That the percentage of outlay for clothing is approximately the same, whatever the income. 3. That the percentage of outlay for lodging, or rent, and for fuel and light, is invariably the same, whatever the income. 4. That as income increases in amount, the percent- age of outlay for "sundries" becomes greater. Zimmerman notes, however, that the above statements of Wright were "an erroneous interpretation of Engel's law" since Engel's law applies pply_to food expenditures. Un— fortunately, several researchers have empirically attempted to invalidate Engel's law by using the law to predict other lMassachusetts Bureau of Statistics of Labor, pp. cit., p. 438. 44 income-expenditure relationships. Several consumer budget researchers believed that Engel's law also would hold for other family expenditures. As Zimmerman notes: On the basis of this [Wright's] interpretation, it has become popular in many studies to abridge and reinterpret Engel's law. Thus, F. H. Streigh- toff found that in New York City clothing percent— ages increase with increasing income while percent— ages for fuel, light, and housing expenditures de- crease. W. F. Ogburn, in a statistical study of families in the District of Columbia, showed that the percentage of expenditure for clothing increased and the percentage of expenditure for rent, fuel, and light decreased with increasing income. Ogburn mistakenly attributes such formulations to Engel, just as do two-thirds to nine-tenths of all studies, outside of German literature, dealing with this particular subject.1 However, Wright made the conclusions by examining the data of Engel's study. In reality, it appears that Engel formulated his income-food expenditure "law" while Wright, examining the same data source, only expanded on the income-expenditure relationships rather than misinter— preted Engel's findings. Therefore, Wright Should be given credit for formulating the “laws" regarding "clothing," "lodging or rent," "fuel and light," and “sundries." Wright made many contributions to the study of fam- ily budget analysis. As Director of the Massachusetts lZimmerman, op. cit., p. 101. The "mistake" in the nomenclature and interpretation of Engel's law is found frequently in the economics and marketing literature. See: William A. Berridge, Emma A. Winslow, and Richard A. Flinn, Purchasing Power of the Consumer: A Statistical Index (New York: A. W. Shaw and Company, 1925), p. 168; and Paul A. Samuelson, Economics: An Introductory Analysis (New York: McGraw-Hill Book Company, Inc., 1964), p. 209. 45 Bureau of Statistics of Labor, he concentrated primarily on studying laborers' living conditions. In 1875, he studied both skilled and unskilled workers and compared their con- sumption habits in value, quantity, and qualities. Wright's comprehensive methods created a foundation for most of the subsequent budget studies. As Zimmerman stated in 1935, methods of budget analysis had not advanced greatly.l Development of "Laws" of Consumption LePlay, Engel, Wright, Zimmerman, and other budget analysts formulated several consumption "laws“ in an attempt to explain specific income-expenditure relationships. The consumption "laws" are presented below by specific expendi- ture category and the formulator of the "law."2 Food consumption "laws" Engel: The poorer an individual, a family, or a people, the greater must be the percentage of the income necessary for the maintenance of physical susten- ance, and again of this a greater proportion must be allowed for food.3 Wright: That the greater the income, the smaller the rela- tive percentage of outlay for subsistence.4 lZimmerman, op. cit., p. 471. 2Ibid., pp. 51-53. Zimmerman presents several other consumption "laws." Many of the "laws" presented in Zimmer- man's writings duplicate the 1aws presented in the research and have been deleted from the study. 3Ibid., p. 51. 4Massachusetts Bureau of Statistics of Labor, pp. Cite, po 4380 46 Clothing “laws" Wright: The percentage of outlay for clothing is approxi- mately the same, whatever the income. Halbwachs: Labor spends a little less per adult unit for cloth— ing than employers in the same income group. Offi— cials spend a great deal more per adult unit for clothing than employees and laborers in the same income class. Rent, fuel, light, and housing "laws" Wright: The percentage of outlay for lodging, or rent, and for fuel and light, is invariably the same, what- ever the income. Chapin, Streightoff, and Ogburn: With increasing incomes, the percentages for fuel, light and housing decrease. Miscellaneous or sundry “laws“ Wright: AS the income increases in amount, the percentage of outlay for “sundries" becomes greater.5 Evaluation of Consumppion ”Laws" The consumption "laws” were originally formulated lIbid. 2Zimmerman, o . cit., p. 52. Massachusetts Bureau of Statistics of Labor, pp. cit., p. 438. 4Zimmerman, op. cit., p. 52. 5Massachusetts Bureau of Statistics of Labor, pp. cit., p. 438. 47 to explain relationships between income and consumption of specific expenditures. However, since many of the "laws" were developed 50-100 years ago, several important changes have occurred in our society which may affect the validity of the "laws.“ As an example, when the "laws“ were being developed there was little consumer credit, especially for the working class families. Therefore, the incomes of fam- ilies usually placed rigid restrictions on family expendi- tures. By contrast, in our society consumer credit is com- monly used, and as a result, current family income does not place such a rigid parameter on spending behavior. The majority of the pioneering budget studies also dealt with families spending most of their incomes usually on subsistence expenditures.l And the families of Engel's and Wright's period came much closer to the conception of “economic man“ than do current families since most current families have discretionary income. Most of the pioneering budget analysts did not consider the proposition that some goods have an “economic" significance while others may have "socio—psychological" significance. Illustratively, eco- nomic goods are consumed for the satisfaction of physiolog- ical needs while socio-psychological goods are demanded for 1Ibid., p. 440. Wright estimates that in 1875 skilled workers in Massachusetts devoted 93 per cent of their in- comes for subsistence (food), clothing, rent, and fuel. In this study “necessities" are defined as total food, housing, clothing, and utilities. 48 their use as symbols of status and mobility.1 Other changes having possible effects on current family expenditure patterns are the existence of Social Security benefits, minimum wage laws, guaranteed annual wage contracts, annuities, retirement plans, and subsidized medical care. The general validity of the consumption "laws" over time can be evaluated by reviewing several American cross— sectional budget studies. By reviewing the proportion of income spent for various commodities in major budget studies conducted from 1875 to 1950 several relationships are noted below tending to validate some of the above consumption laws. The budget studies are found in Appendix I. l. The proportion of income spent for food or sub— sistence generally decreases as income increases in all the budget studies "validating" Engel's and Wright's "laws." 2. Wright's “law," "the percentage of outlay of clothing is approximately the same, whatever the income," is not “validated" in the budget studies since the percent— age of income spent on clothing increases with increase in income. 3. Wright's "law," "the percentage of outlay for lodging, or rent, and for fuel and light, is invariably the same, whatever the income," is somewhat difficult to -r lPierre Martineau, "Social Classes and Spending Behavior," Journal of Marketing, Vol. 23 (October, 1958), pp. 123-25. 49 "validate." However, in the 1874—75, 1881—91, 1901 studies, utilities plus housing or rent expenditures do remain rela- tively constant in spite of changes in income. However, in the 1934-36 and 1950 studies, fuel, light, refrigeration, and housing tend to decrease proportionately as income in- creases. The latter two studies of 1934-36 and 1950 cor— respond to Chapin's, Streighloff's, and Ogburn's statement that "with increasing incomes, the percentage for fuel, light, and housing decreases." 4. Wright's “law," "that as income increases in amount, the percentage of outlay for 'sundries' becomes greater," is also "validated" by the study. Change in Income as a Determinant of Expenditures Another variable likely having a significant influ- ence on family expenditure patterns is the change in income.1 Unfortunately, the effect of the change in income variable frequently has been ignored in the marketing literature. To examine the effects of change in income a review was made on three contemporary income hypotheses explaining the effect of change in income on family expenditure and saving patterns. In recent years, the influence of a change in in— come on consumption patterns primarily has been studied by 1Elizabeth W. Gilboy, "Income-Expenditure Relations," Review of Economic Statistics, Vol. XXII, 1940, pp. 115-21. 50 economists and agricultural economists. By analyzing change in income—expenditure relationships, three hypotheses have been posited: (1) the absolute income hypothesis, (2) the relative income hypothesis, and (3) the permanent income hypothesis. Each hypothesis attempts to explain how con- sumers adjust consumption to new income levels. All the hypotheses can be utilized in determining and interpreting O O O O l cross-sectional income—expenditure relationships. Absolute Income Hypothesis Keynes, one of the first to comprehensively examine the relationship between change in income and consumption, developed the absolute income hypothesis primarily because classical economic doctrine did not adequately explain in- come—aggregate consumption behavior. As a result, Keynes posited a revolutionary approach for explaining income— consumption relationships. When Keynes wrote the General Theory about the only data available for analyzing income—consumption relation— ships were cross-sectional studies.2 From the cross-sec- tional budget studies Keynes noted wide variations in con— sumer expenditure patterns. Keynes observed, for example, 1A comprehensive examination of the three hypotheses is presented in: Robert Ferber, “Research on Household Behavior,“ American Economic Review, 52:19—63, 1962. 2J. M. Keynes, The General Theopy of Emplpyment Interest and Money (New York: Harcourt, Brace and Company, Inc., 1936). 51 that consumption generally decreases at an increasing rate as income increases.1 From such observations, Keynes de— veloped his consumption function and the absolute income hypothesis. As Hansen noted in his interpretative analysis of Keynes: Keynes' most notable contribution, however, was his consumption function. The psychological pro— pensities of consumers plus the institutional be- havior patterns of the community (notably those of business firms) are such, he argued, that (1) some part of income (except at very low levels) is saved and (2) of any net addition to real income, some of the increment is saved. Accordingly, the behavior patterns of the community are such that a gap exists (which gap widens absolutely as real income increases between the amount the community wishes to consume and the output the community is capable of producing.2 The Keynesian relationship among income, consump— tion, and saving can be illustrated in Figure 3. 45° INCOME = CONSUMPTION I. CONSUMPTION FUNCTION .1 CONSUMPTION INCOME A denotes dissaving B denotes saving Figure 3. Graphical illustration of the absolute income hypothesis lIbido, pp. 96—970 2Alvin H. Hansen, A Guide to Keynes (New York: McGraw-Hill Book Company, Inc., 1953), p. 27. 52 Keynes posited several hypotheses attempting to explain the slope of the consumption function. Keynes be— lieved the slope partially was determined by social and psychological influences on the consuming unit. The social factors included income distribution and occupation while the psychological influences were determined mainly by con— sumer "expectations." In Figure 3, to the right of the equilibrium point (where consumption equals income), income increases both absolute1y_and relatively as compared to the increase in consumption expenditures. Keynes expressed the fundamental relationship between consumption and income in terms of the marginal propensity to consume expressed as the change in consumption M°P°C' = the change in income Under the Keynesian assumption the marginal propensity to consume, the consumption function slope, must be less than unity. The reciprocal of the marginal propensity to con- sume, the marginal propensity to save, asserts under the Keynesian analysis that, as income increases absolutely, saving will also tend to increase absolutely and relatively. Relative Income Hypothesis Due to a significant limitation of the Keynesian income-consumption hypothesis, notably that a stable and predictable relationship did not always exist between con- sumption and income, an alternative hypothesis was sought 53 which would aid in explaining the instability. Empirical investigations revealed that increases in consumption tended to lag behind increases in income; decreases in consumption lagged behind decreases in income. Duesenberry maintained consumption was determined in part by former living stand— ards.l That is, if income increased, families would not immediately increase consumption expenditures since the expenditure patterns of families were conditioned to pre— vious standards of living. Conversely, if income decreased, the family would not immediately decrease consumption ex— penditures. Duesenberry's hypothesis is illustrated in Figure 4. CONSUMPTION EQUALS INCOME 45° LONG RUN CONSUMPTION FUNCTION 5 SHORT RUN H CONSUMPTION E FUNCTION (2) 2 3 SHORT RUN z CONSUMPTION 8 /"“"r, FUNCTION (1) o INCOME Figure 4. Graphical illustration of the relative income hypothesis 1J. S. Duesenberry, Incomey Saving and the Theory_ of Consumer Behavipg (Cambridge, Mass.: Harvard University Press, 19527, Chap. 5. 54 In the “short run," as income increases or decreases consumption increases or decreases but not as rapidly as income. Families usually exhibit a short run consumption function with increases in income. However, in the "long run" consumption patterns will generally adjust to the new income levels. A neoclassical addition to the relative income hy- pothesis introduced the wealth effect. With the addition of the wealth effect, variations in income, saving, and consumption patterns could partially be reconciled and ex- plained.1 For example, when family income decreased, con- sumption usually did not immediately decrease. The neo— classical approach explained the 1ag in consumption was due to the ”wealth“ of the family, such as stocks, bonds, and other assets. Permanent Income Hypothesis AS noted previously, the literature reveals a di- versity of opinion on the effect of income change on ex— penditure patterns. Ruth Mack decided the direction of lJ. Tobin, "Relative Income, Absolute Income, and Savings,“ in Money, Trade and Economic Growth, Essays in Honor of John H. Williams (New York: Macmillan Company, 1951), pp. 135-56. Tobin found that the income-consumption relationship could be explained by other factors than "rela- tive income.“ For example, he studied the difference in saving habits between Negro and white families and found that although Negro and white families may have equal in- comes, the white family usually saved less. Tobin main- tained that the difference was due to the greater amount of wealth (assets) white families possessed. 55 change in income should be regarded as a determinant of expenditures.1 By utilizing Bureau of Labor Statistics data gathered in 1942, Mack found the effect of changes in income of families on expenditures. The families were separated into the following three "income-change" classi— fications: 1. Families whose income in the first quarter of 1942 (converted to annual basis) was within plus or minus five percent of their 1941 income. 2. Families whose income had increased by more than five percent. 3. Families whose income had declined by more than five percent between 1941 and the first quarter of 1942.2 After plotting the income-consumption regression for each of the income-change classifications, Mack made the following observations. 1. At each income level, expenditure is consider— ably higher for fallen income families than those whose income was stable or rising. 2. The income spending patterns of risen and stable income families do not appear to be markedly different although there does seem to be a faint tendency for rising income families to spend slightly less than stable income families at all levels of income except the top one. 3. Expenditures for the full sample seem to shift somewhat less with a given shift in income class (the regression is flatter) than for lRuth P. Mack, "The Direction of Change in Income and the Consumption Function," The Review of Economics and Statistics, 30:239-58 (November, 1948). 2 Ibid., p. 240. 53 any of the income-change groups. This tendency, which disappears in the highest income class, is caused by the fact that falling incomes were relatively more common in the lower income clas- ses and rising incomes in the higher classes.l Mack further reports that other studies confirm that families whose income falls into a lower income Class tend to spend more than the average for the lower income class group. Further, in cases where family income rose to a higher income class the level of expenditure tended to be somewhat lower than the average for that income class. In both cases Mack believed the expenditure behavior was due, in part, to the level of prior expenditures. In one sense, part of the family's expenditure patterns are "con- tractual." Expenditures for automobiles, owned housing, rent and utilities are not as easily adjusted upward or downward with changes in income. These findings led some economists to believe that present consumption is related primarily to the consumer's expectations of what their "permanent" income is rather than current income. Moreover, short term variations in income were thought to influence savings more than expendi- tures: as income increases in the short run consumption increases more slowly, and saVings increase proportionately greater than consumption. lIbid., p. 2410 2Ibid., p. 258. 57 Friedman's version of permanent income hypothesis is best known although others, such as Modigliani and Brum- berg, have formulated similar hypotheses.l Friedman's hy- pothesis maintains that the family's perception of permanent income is the most significant variable in explaining a family's consumption patterns. Although a family's income may fluctuate in the short run, consumption is usually sta- ble. For example, if income of a family suddenly increases or decreases there likely will be little change in the ex- penditures of the family unless the family perceives that the income change is permanent. Figure 5 illustrates the permanent income hypothesis. LONG RUN CONSUMPTION FUNCTION SHORT RUN CONSUMPTION z FUNCTION 3 SHORT RUN CONSUMPTION E C FUNCTION 2 2 / SHORT RUN CONSUMPTION D a Cl.__ __ __.__.__._#:=;=.————"’ FUNCTION 0 U Y1 Y2 INCOME Figure 5. Graphical illustration of the permanent income hypothesis ' 1Milton Friedman, A Theory of the Consumption Func— tion (New York: National Bureau of Economic Research, Prince- ton University Press, 1957), pp. 20—37; and Franco Modigliani and R. Brumberg, "Utility Analysis and the Consumption Func- tion--An Interpretation of Cross-Section Data,“ in K. K. Kurihara (ed.), Post—Keynesian Economics (New Brunswick, N.J.: Rutgers University Press, 1954), pp. 388-436. 58 In Figure 5, for example, consumers have an increase in income to income level (Y2). Under the assumptions of the permanent income hypothesis, consumers would not adjust consumption to the new level (C2) unless the new income level was perceived as permanent. It is likely, however, consumers usually make some minor adjustments with increases or decreases in income. The Income Hypotheses and Family Savings The Keynesian absolute income hypothesis points out as income increases savings increase both absolutely and relatively. By contrast, the relative income hypoth- esis states savings are employed to "adjust“ to a new con- sumption level. When income changes to a new level, con— sumption usually does not adjust automatically. The family savings are used, for example, when a family adjusts to a new lower income level. The permanent income hypothesis reflects a similar relationship, according to the hypoth- esis savings are used to help balance consumption and the . . . l consumer's perception of his permanent income. lGeorge Katona, "Effect of Income Changes on the Rate of Saving," Review of Economics and Statistics, 31: 95-102, 1949. Katona, using data from the Survey of Con— sumer Finance, found in the short-run savings usually in— crease proportionately more than income with increases in income and with decreases in income dissavings by families is more likely to occur than with consumers having a lower stable income level. Katona explains the income-saving relationship two ways: (1) consumers develop slowly chang— ing consumption habits and (2) expectations of the future income levels of families. 59 Using the permanent income hypothesis to explain income-saving relationships, Eizenga studied the saving patterns of various occupational groups classified by in- come classes. Using the 1950 Bureau of Labor Statistics' Study of Consumer Expenditures, he found that occupational groups do not save the same amount at similar income levels. He also found the occupations which dissaved the most at lower income levels tended to save more at higher income levels. Some of Eizenga's findings are presented below: 1. Self—employed people had greater savings at income levels above $8,500; however, up to $4,500 the self— employed dissaved. 2. Professionals save less than the self-employed at higher income levels and dissave less at lower income levels. 3. Professionals and skilled wage earners exhibit similar saving patterns at all income levels. The profes- sionals save only a little more than skilled wage earners at most income levels.1 Eizenga explains that the occupational income-saving relationships can be explained using Friedman's permanent income hypothesis: All the occupation groups dissave at lower income levels and save at higher income levels. And they lW. Eizenga, Demographic Factors and Savings (Am- sterdam: North-Holland Publishing Company, 1961), p. 70. 60 dissave more at lower and save more at higher in- come levels, the higher the permanent income that can roughly be approached by the average income of each occupational group. The average income of the self-employed people and the salaried pro- fessionals is considerably higher than that of other Occupation groups. And according to the permanent income hypothesis, consumption tends to be a function of permanent income. There still remains the question: how can the dissaving REE? sibility_be explained. As far as the older-retired people are concerned, those just draw on their pre- vious accumulated savings. But what about the young- er people in high-income occupations with low in- comes? They have confidence in the future returns of their present activities. This, however, only explains the willingness to borrow money. But such people have a big chance of finding a lender. Financial intermediaries, considering such appli- cations, will base their decision on the expected income develOpment. And both groups, although par- ticularly the self-employed people, might be able to get credit. Self-employed people have other forms of security than do salaried professionals. Indeed, material wealth is a better basis for credit than non-material wealth. This might be the reason why self-employed people dissave more at lower in- come levels than salaried professionals with almost the same average income.1 Stage in Family Life Cycle as a Determinant of Expenditures The observation that families have rather distinct social and economic family life stages has generated con- siderable interest in marketing and economics. Rural soci- ologists, however, were the first to study and delineate the family life stages. They were of the opinion that the family life cycle could delineate differences in urban and rural cultures; show changes in the socio—economic status lIbid., pp. 71-72. 61 of the family; and that it could measure family productiv— ity.1 Economists and market researchers began developing serious interest in the life cycle concept in the early 1950's to determine what specific effect, if any, the fam- ily life cycle had on expenditure patterns and whether the family life cycle could be employed in delineating markets and predicting consumer demand.2 Fisher maintains the family life cycle concept is a useful framework for explaining variations in the family's perception of its income and consequently consumption be— havior. Fisher believes the younger family frequently has a different perception of income and saving than an older family. Fisher also believes the perception of price changes is influenced by age of family. A crucial question arising from Fisher's investigation is whether age of family or the "economic correlation of age" is more important in af- fecting the family's perception of income, consequently lPitirim A. Sorokin, Carle C. Zimmerman, and Charles J. Galpin, A Systematic Source Book in Rural Soci— ology (Minneapolis: University of Minnesota Press, 1931), II, pp. 3-32; Charles P. Loomis, “The Study of the Life Cycle of Families," Rural Sociology, June, 1936, pp. 180-99; and J. A. Beegle and C. P. Loomis, "Life Cycles of Farm, Rural-Nonfarm, and Urban Families in the United States as Derived from Census Materials," Rural Sociology, March, 1943, pp. 70-74. 2Lincoln Clark (ed.), Consumer Behavior, II: The Life Cycle and Consumer Behavior (New York: New York Uni— versity Press, 1955). 62 influencing expenditure patterns.l Lansing and Kish believe the family life cycle should be considered as an independent variable which can be sub- stituted for chronological age in many types of consumer research since the stage in the family life cycle may have greater "explanatory" powers. The authors, like Fisher, maintain the stage in the family life cycle embodies other socio-economic variables which influence expenditure pat— terns. As Lansing and Kish state: It is well known that changes occur in people's attitudes and behavior as they grow older, but many of these changes may be associated less with the biological process of aging than with the influence of age upon the individual's family memberships. Thus, the critical dates in the life of an individ- ual may not be his birthdays so much as the days when a change occurs in his family status, for ex- ample, when he marries, or when his first child is born. On the other hand, Lansing and Morgan have utilized the family life cycle as a variable for explaining consumer . . . 3 . incomes, saVIngs, and expenditure patterns. USIng cross- lJanet Fisher, “Family Life Cycle Analysis in Re— search on Consumer Behavior," Consumer Behavior, ed. Lincoln Clark, Vol. II (New York: New York University Press, 1955), ppo 28-360 2John B. Lansing and Leslie Kish, "Family Life Cy— cle as an Independent Variable,“ American Sociological Re- ‘view, October, 1957, pp. 512-13. 3John B. Lansing and James N. Morgan, "Consumer .Finances Over the Life Cycle," Consumer Behavior, ed. Lincoln Clark, Vol. II (New York: New York University Press, 1955), Pp~ 36-51. 63 sectional data from the Survey of Consumer Finances in 1952 and 1953, Lansing and Morgan analyzed variations in income, savings, and expenditure patterns in the economic life his— tory of families. They pointed out in their study that family income increases, then decreases over the life cycle. However, the findings were somewhat surprising since family income peaks were normally reached in the "young married childless class." Lansing and Morgan also explained the family income patterns by analyzing the wife's contributions to the family income stream. In the "young married child- less stage" approximately 63 per cent of the wives worked for salaries or wages. In the next stage, “young, youngest child under 6,” only 23 per cent of the wives received in— come which indicates that as the young families have child- ren the income stream of the wives diminishes and conse- quently total family income declines. The next stage, "young, youngest child six or over," reveals that incomes again rise. Further, the percentage of wives who work again increases. From the above stage to the "older, no children under 18" stage income and the percentage of wives working decreased. The relationship of the wife's contri- butions to the family's income stream is illustrated in Table 3. Unfortunately, Lansing and Morgan do not explain the relative importance of the wife's contribution to fam— ily income which depends upon how much both the husband and wife make. Table 4 illustrates the family income—life cycle 64 TABLE 3 INCOME FROM WAGES AND SALARY OF THE WIFE AT DIFFERENT STAGES IN THE LIFE CYCLE-—AVERAGE ALL INCOME CLASSES, 1953 Percentage of Wives Earning Wage or Family Life chle Stage Salary Income Young, childless 63% Young, youngest child under Six 23% Young, youngest child six and over 38% Older (over 45), with children 24% Older, no children under eighteen 23% Source: John E. Lansing and James N. Morgan, "Con- sumer Finances Over the Life Cycle," Consumer Behavior, ed. Lincoln Clark, Vol. II (New York: New York University Press, 1955), p. 40. TABLE 4 THE MEDIAN INCOME OF SPENDING UNITS AT DIFFERENT STAGES IN THE LIFE CYCLE, 1952 AND 1953 Median Income Median Income Family Life gycle Stage 1952 1953 Young, single $2,600 $2,600 Young, married, childless 4,600 4,800 Young, married, youngest child under six 4,200 4,300 Young, married, youngest child six or over 4,500 5,000 Older, married (over 45), with children 4,400 4,300 Older, married, no children under eighteen 3,600 3,900 Older, single 1,700 1,700 Source: John B. Lansing and James N. Morgan, sumer Finances Over the Life Cycle,“ Consumer Behavior, Lincoln Clark, Vol. II (New York: 1955), p. 39. New York University Press, "Con- 65 relationship. In Table 4 family median income in both 1952 and 1953 are relatively high in the "young, married, child- less" stage, as compared to the other life cycle stages, which corresponds to the high percentage of wives working (Table 3). However, in both years median income decreases in the stage "young, married, youngest child under 6" and only 23 per cent of the wives worked. An increase in median income is noted in the "young, married, youngest child six or over" stage which corresponds with the percentage of wives receiving income (38 per cent; see Table 3). From that stage, median income for families tends to decrease as does the percentage of working wives. Lansing and Morgan explain further that the family life cycle offers a vehicle for studying durable goods spend— ing patterns. Although Lansing and Morgan's analysis is limited to durable goods, several findings are presented. The research noted family expenditures for durable goods follow a "priority system"; as a result, families frequently tend to allocate income to durable goods at varying rates, depending on the life cycle stage and the type of durable good. The highest percentage of families buying houses occurred in the category "young, married, youngest child under 6" and gradually declines over the life cycle. In new automobile purchases the percentage of fam- ilies buying new automobiles reaches a peak in the "young, married, youngest child over 6“ stage and then decreases. 66 By contrast, the percentage of families buying used cars remains relatively high in both the "young, married, child— less" and the "young, married, youngest child under 6" stages. In the purchase of furniture two peaks are reached: first, in the "young, married, childless" class and then in the "young, married, youngest child over 6." The highest per- centage of families purchasing refrigerators is found in the “young, married, childless" class. A similar relation- ship is presented noting the percentage of families purchas- ing stoves. By contrast, the highest percentage of families purchasing washing machines is found in the "young, married, youngest child under 6" class. The purchase of television sets also is higher in the "young, married, youngest child under 6“ class. The relationship among stage in family life cycle, median income, and percentage of families purchasing select- ed durable goods is presented in Table 5. It was further demonstrated by Lansing and Morgan that the family life cycle could be utilized to examine changes in assets and debts. The findings of Lansing and Morgan indicated the following relationships. 1. The percentage of families with any liguid as- sets remains relatively constant in all family life cycle stages. 2. The median liquid assets per life cycle stage initially increase, reach a peak, decrease, and gradually increase. 67 .mpooo OHQMMSO paormmSOC maeumad Imu pcm mcflmmruuSd Mom mmomum wauxu UMHH Thu macaw >UCODMHchU Hmwmcmm m we OD mummd Idm mumra .moooo Mahmuso How mcuOPDMQ mmmruudd :Hmwucmddmm: HMHHEHM ocsow .MOUCMCHm umesmcou mo >m>u3m cmoaruaz mo xpflmum>flco omma Thu Eoww mumo moan: .ummmwrm .msleea .dd .Admma .02flmcmq ummm .huflmum>flco mumpm cmoHQUHz .COHuMDHOMMHo .Q.cm OOQMHHQDQCSV :maumu mmaq Thu Op twpmamm Mahmupmm muzuflocmdxm ocm OEOUCH: .ummmmrm .3 Houmu. Q .ma .a ..penfl .mpoom manmusp 02A IMOQUHSQ mmaaflemm wo Demo MOQ mmm .d .Ammma .mmmwd mpflmum>flco xuo» zmz "show Bmzo HH .Ho> .xmeU CHOUCHA .om .uoa>mrmm MOE5chU :.OHU>U OMHA may um>o mmucmcam umEschU: .cmmuoz .z mmemn ocm ocamcmq .m crop Eoum mmumeflpmw OEOUCH cmaomz "muusom .MOHHHEMM mo pcmu uma m.o carp mmmq u + O n ma ma ma mm ea m COHMH>OHOB H v m 0 ma OH H OCHSUMS Unfizmmz v s m s 0H ma m O>Oum m s m NH NH ea H uowmummflummm m Ha ma em mm mm oa musuflcusm a HH ea ma mm mm Ha umu owns m m Ha om Ha Ha m umu zmz + m a s m o + Omoom . "mooow Mahmuso oos.em oom.mm oom.em ooo.mm oom.em oom.em oop.mm meoucH spasm: OHmCHm ma COHOHHQU o um>o m umono mmmHoHHrU mamcam umpao umpcp spas pesto peflro pmauumz pesos COHOHHQU oz Ame um>oo unmocsow unmocso» ocsow poaunmz omfluwmz omeuumz omauumz uwpao HOOHO dado» mesow ..mmozmem mmme .maooo mqmrisidered as an accurate predictor of social class place— n“fatrlt. And it directly influences the consumption life 53‘t1)rles of the family. However, there is a general paucity (2’35 .research illustrating the direct influence of occupation ‘:>ru, family consumption patterns. Usually, the occupation ‘J—Eiltiable is employed in combination with other socio— EEC:Dnomic variables in determining social class placement. :EIj- Warner's research, for example, on social stratification terEE occupation of the family head was the most accurate $erlgle socio-economic variable for predicting the social <:;J~€iss of a family.2 Warner makes clear, however, that the \ 13229.. 2W. Lloyd Warner, Marcia Meeker, and Kenneth Eells, §€9-Sg;al Class in America (Chicago: Science Research Asso— Q'j‘ei’mes, Inc., 1949), pp. 176- 85. 78 prediction accuracy of social class placement is improved .if other socio-economic variables are used in conjunction unith the occupation of the family head. An important ques- tjxon, for market delineation purposes, is whether the in— czliision of several other variables is necessary for predict- iJlg; social classes. For example, the amount of income, (inveLlling area, source of income, house type, and education éil?€e all correlated closely with occupation. The answer t3C> 'the question lies in the desired degree of accuracy in j—Cientifying social class. To illustrate the differences j~I1 class placement accuracy, Warner correlated occupation ES-ES a social Class predictor with his social placement re- E51—11ts for Jonesville and found that the coefficient of cor- Next, Warner Jc’eellation was .91 with a standard error of 1.8. CZWCerelated various “sets" of socio-economic variables to \ - lWarner frequently employed the following occupa- 1:--'1-mbination of socio-economic variables produced greater 17€=1¢iability in predicting social class placement. Therefore, although occupation is not as accurate j4r1 ;predicting social class as the combination of occupation Eililci other socio-economic variables, it is nevertheless a Key indicator of social class placement. And as a primary 'j~r1d§ family Size posits large families do spend more dollars IFCDJE food than smaller families, but, the amount spent is r1 . . <:>t: proportionate to the ratio of the number of persons. \ 1U. S. Department of Agriculture, Income and House— ‘GSE£Ld.Size, Marketing Research Report No. 340 (Washington: ‘ 53. Government Printing Office, 1959), p. 40. ‘7‘ r‘,—‘ t i" .‘T'xf‘fk 80 An example of family size-expenditure relationships is il- lustrated in Table 8. TABLE 8 RELATIONSHIPS OF FOOD EXPENDITURES FOR ONE WEEK TO FAMILY SIZE, URBAN FAMILIES, UNITED STATES DEPARTMENT OF AGRICUL- TURE SURVEY, 1948 _§Li:ze of Household Per Family Per Family Member Per cent Per cent Dollars Index Dollars Index 2 persons 20.18 100 9.66 100 3 persons 24.64 122 8.50 88 4 persons 28.15 139 7.49 78 55 <5r more persons 32.06 159 6.34 66 Source: United States Department of Agriculture, Consumption of Urban Families in the United States, U.S. .______d ‘AKQJEiculture Information Bulletin No. 132 (Washington: crvernment Printing Office, 1954), p. 6. The above relationships are considered to be the Ireisult of several variables influencing food expenditures. I?first, economies of scale are present in the preparation ()1? food. Second, many of the larger families include more czllildren than the smaller families. As noted, the four ID'EErson family spends only 39 per cent more for food than c1(Des the two person family. Third, large families gener- es'~:'Lly consume food of lower quality. Fourth, the smaller dELEMnilies usually eat away from home more than do large fam- 1 lies.1 \ lIbid., pp. 6-7. 81 A similar relationship also is noted in Table 9 for other family expenditure accounts. Using data from the 1950 Bureau of Labor Statistics' survey cross-classi- fied by family size and income class, the following per capita consumption relationships are noted. Table 9 gen- erally indicates as the family Size increases per capita eXpenditures generally decrease. One may again posit there are economies of scale in purchasing and consuming. As r1C:~‘t:.ed in the family size-per capita expenditure relation- SI‘Aips, expenditures do not vary greatly from the per capita e>ahmccmm mo xpflmum>HCD "macaamomaflcmv HH .Ho> .mmcH>mm cam mmEOUCH «mmusuflocmm Ixm umezmcou mo mosum .mflcm>ammccmm mo >DHMHO>HCD mcu Eoum omuamo< «muusom 82 OOH opm sma mm mam mmm.a oom.suooo.m« mm saw mNH Hm mmm ems.a ooo.muooo.mw am sad oaa em was mam.a ooo.m:ooo.s« ms mma as 88 was mHN.H ooo.auooo.mm hHflEmm COmuwm mmuca Has mom sow ems mas oos.m oom.suooo.pm sea Hmm «ma mma Hos mmm.m ooo.mnooo.mm mas omm spa am 088 mmH.~ ooo.mnooo.sa pm msa QHH ms Nmm smo.a ooo.anooo.mm NHHEmm GOmumm 038 mos ems mm msa mop.a mmo.m oom.snooo.ma oma Hos mmm omm mom.a maa.v ooo.puooo.ma mm mmm pp Nam oms.H mos.m ooo.mnooo.vm pm mmm ms 08H Hmm Nos.m ooo.anooo.mm um85mcou OHOCHm mumu mafinuoHU pcmamwfiom COHDOHOQO mummum>mm mmusuflocmmwm HOUHODZ cam oaocmmzom cam ooom ucmuusu OCHEMHCHDm Havoa mmmau OEOUQH cam mnam hafiemm Hmuomma .memoz use zH mMHaHO mom< Omaumqmm m mqmwever, showing differences in social class expenditure pBitterns, has been conducted on consumer durable goods. The Chicago Tribune Study noted the lower social classes DJ~aced more emphasis on large expensive appliances and house— thd furnishings than did middle class families. The find- illgs of Rainwater generally agreed with the findings of the Chicago Tribune Study. 84 The review of the effect of family size on family (expenditure patterns generally noted family size does in- iiluence expenditures, but not in direct proportion to the .niimber of people in the family. The family size—expenditure Irefilationship was explained by the existence of economics cxf’ scale in purchasing and utilization. Evaluation of Selected Socio-Economic Variables as Criteria for Delineating Markets In reviewing the effects of several socio-economic ‘Vréixriables on family expenditure patterns, several socio— e3<2<3nomic variables were examined influencing family expendi— ‘tILLre patterns. Market researchers are fundamentally inter— €E=ESted in determining socio-economic influences on consump- ‘tljuon patterns for purposes of market delineation. Never- tlfleless, although a variable can readily differentiate con- Esllmer expenditure patterns, the variable may have several (jiisadvantages for market delineation utilization. In the evaluation of socio—economic variables for InEtrket delineation, several methods of delineating markets Eil?e examined and evaluated. The two basic criteria employed t1<> evaluate the socio-economic variables for market delin- GE’Eition utilization are: (l) the operational characteris- 1blecs of the variables which include the cost and complexity (2’15 implementation and (2) the degree the variable reflects $3‘1‘—><:io-economic homogeneity of the differentiated groups. 85 Market Delineation by_Family Income Class The income class of the family frequently is a com- Inonly used method to delineate markets. For some types caf’goods and services family income serves as a reliable .puredictor of purchase patterns. To employ the income class rnesthod, markets are usually classified according to the Idlinnber of families in various income classes. From such ‘ £1 czlassification, the market is usually given some type . ‘ <>:E quantitative or qualitative rating for market evaluation I E>lerposes. The fact of families having parabolic income pat- tZEBrnS is a fundamental problem encountered in employing 3f?€1mily income classes. Incomes for families increase, then 'k1:sually decrease as families move through family life cycles. IPOI: example, in the income class of $5,000-5,999, both young- GEIT and older families are present and, as a result, errone- <3115 market conclusions may result. The younger families 11finally have a larger family and possibly both the husband Eirld wife are working. On the other hand, an older family innéiy be in the same income class with no children. Undoubt- €3<flly, the expenditure patterns of each family would differ ‘“’deely and likely would affect the reliability of employing S‘trictly income classes in market delineation. Using 1950 Bureau of Labor Statistics data, Kurt IVIEiyer posits somewhat of a different assumption regarding income class. Mayer asserts there is "an impressive 86 uniformity of spending patterns" regardless of occupational By examining aggregated family expenditure pat— category. only terns cross—Classified by occupation and income class, snuall differences were noted in the percentage of total expenditures devoted to food, housing, Clothing, and trans- ;xozrtation within each income-occupation class. Unfortué naitxely, Mayer does not consider the effect of the family age variable. In rechecking the source of Mayer's data, a. vvjmde range of family ages was found to be included in eeaxzru occupation—income category. For market delineation ENJlrgaoses, Mayer can be positing a dangerous conclusion. Moreover, the highest income class utilized by Mayer ranges OT134y' from $6,000 to $7,500, eliminating differences in fam— :13~)’ .spending patterns in higher income-occupation categories. I11 3.950, for example, in the 35-44 year age class for large czj—tliues in the North, over 18 per cent of the "self-employed" far“ilies and over 18 per cent of the "professionals" had higher incomes than $7,500. The omission of the higher :LrICRDne classes neglects important market sectors. Coleman maintains in the purchase of automobiles \ "Diminishing Class Differentials in lKurt Mayer, Interna- t: tzgie3 United States," K klos; Zurich, Switzerland: cbnal Review of Social Sciences, 12:605-26, 1959. 21bid., p. 622. 3University of Pennsylvania, Study_of Consumer Ex- 39 iirndituresJ Incomes and Savings, Vol. 18, Table 4-2, p. Iailadelphia: University of Pennsylvania, 1956). 87 income class does not appear the primary determinant. "status aspirations" of consumers determine what Rather, Similarly, .Orand and model of an automobile purchased. Scxzial Research, Incorporated's study of the adoptions of cxalxor television reveals that at the time of the study color .se1:s were being purchased by all income classes. Another factor possibly distorting the value of eungbldoying income class alone is that the lower income ranges azwa disproportionately represented by unskilled and semi- Slcfi.lgled workers, retired families, students, and in some Cases the young professionals and managerial workers. All 015 tflnese consuming groups likely have widely varying expen— diture patterns. A primary advantage of employing income classes 153 tnne usual ease in obtaining data and the simplicity of tilEB .income criteria. Met Delineation by Stage in %iy Life Cycle The life cycle as a basis for market delineation rleisi generated considerable interest in recent years since i. t: offered an approach that would potentially reduce some C) if the errors in other market delineation approaches, es- E’Ctially the income method. A fundamental problem in implementing family life Q j§r. 34 (Ann Arbor: University of Michigan, 1964), p. 11. 90 Others Head is single Head is married living with wife (no other adults in the family) Head of family No Children is younger Young children /////” (under 45 yrs.) Older children Aklfil Families \\\\1 Head of family Children is old (45 yrs. No children old or older) Head is married living with wife (no other adults in the family) Head is single (living alone) Others F“iglire 6. David's life cycle classification patternsl \ 1Martin H. David, Family_Composition and Consump— “‘~\£§Q, Contributions to Economic Analysis, No. 25 (Amsterdam: N<31rth-Holland Publishing Company, 1962), p. 22. 91 Under age 45 Single No children Children Married, no children Married, children Married less than 10 years Married 10 years or longer Youngest child under 6 Youngest child 6 or older 45 years of age or older Married, children Married, no children Single Children No children Most of the research on family life cycles has been CSDriciucted on aggregated socio-economic groups, possibly ‘NEEEiktening the findings for market delineation purposes. wrlGather different socio—economic groups have different life c:Et'CTJLes has not been widely investigated. Another limita- tliJDII in operationalizing the life cycle concept occurs since S‘Drnee researchers separate family life cycles into several S‘pecific stages while others separate only into three or :563LLr-and, as a result, some confusion exists regarding what 55t1519es are the most appropriate for marketing purposes. PiCDVvever, a comparison of the various family life cycle $51:éages reveals that there is not a great amount of diver- ESi-‘ty in the various approaches. In Table 10 a comparison 5‘53 made of the different life cycle stages which have been (iiidscussed in this section. As can be seen, with a few ad— j Listments there is a very close relationship among the al— tleirnative approaches. The family life cycle classification 92 sumpao no m OHficu pmmmcsowz :ummcoH no mummx 0H omfluumz: :COHUHHSU :GOHUHHEU mcHDHOQQSMIMHmm :cmucaflcu :cmucaflcu cpw3 .cmHu :OHHSU pmma OHOE no Ono >HHEOM .omauuozs uwoao= Inna umoao: mo rouam: spas maosous ompoaoaou :w HOUCS UHHEU Pmmmcso> :me Moos: [Imummm OH awn» oaflcu Dmmmcsom mama omauuozs .cmuoaasu somuoaaro omuoaaro :cmuoaflcu :cmuoHHEU cpflz .cmau :oHch umuflm muOE no mco meow .omauumzs mono»: nuns mono»: mo roman: are: maosous .ooauuoz :OUCODMHXO scouoHHrO usoosmomooa :cmuoaacu :OOHMMOE on .cmHu maauumum umsn cmuoaflcu oc .omfluumz: we comm: IHOE OGDON: :mmmfluumz: OHQSOU cmfluumzz on .Omaunmz :cmu uoaaro maocam: soaocan :EOMOHH50 ma ommm: oc mamcwm: whom» :OHOCHM mo mom Moon: mv noon: 0:50»: mamCHm umucmu cuummmmm UH>mQ Gmmuoz MHEOOA aflmwmw wwmmum THUNM >m>u3m cmmflcuaz cam cam .GMEHOEEHN OMHA haflemm mo xuwmum>HGD mcamcmq mammmm .cflxouom mm5<8zmzoz 9:. $33 386 ES 3:5 “3354834 OH mqmmao oo 6 me me o 6 ma oomoaro mm m h w m N ea cowmom mummm emnmm am RH ma HN RH H ma nHmDoe ma m a m N a N mason .Dm Ha N N N o o m sousnnouaa oa N N N N o N maroaoopaaro Na a m o N o o smmumo 302 .2 Ha m a N a o a snow 302 a a o a a o a Baouumo m a N a m o a ocoao>mau m m m m a o o ommoflcu a o o a N o a copnom whom» mm umUCD nmoao mo< omeaarm omaaarm oneaarm noaom ouwmamem numomcmz mHO NN 8 NH NH N N NH omnuHrO «a N N s N N NH counom mummh fimlmv NNN om HHH 84H Hm Na NoH NHNDoe ON HH NH NH N N NH NHooH .»m «N N NH NH e N a rouannuuHa NN OH OH NH s N NH eronomHHrm NN 0H N HH N N NH monumo 3oz .2 am HN HH NH 0N N NN xuow :02 NN 0H NH «H e N NH DHouumo oN s 6H HN NH N RN oomHo>oHo RN HH NH NN HH N aH oomoHro ON a N NH N O NH sounom mummw ddlmm NNNHO mm< omHHHxN omHHHHN ooHHHxN NOHNN omonosm numomcmz «Nam N oHo OH N H N H H N oNNuHro N N O O N N H copnom Hm>0 UCM mm NNN NN HN ON NN NN NN NHNDoN 4N N N N a N H NHsoH .HN No N N HH a N N BNHOENDDHN NN a H N N N N eraHoomHHrN Na HH a N N N HH Honumo 3m2 .2 ON NH NH N OH a HH snow 3oz NN N N N a N N uHouuoO Na N o OH N N N oooHo>mHo Na N N NH N N N ONNOHBO NN N N N N N NH counom mummm vmlmm NNNHo 6N4 omHHHHm omHHHHN omHHHHN NNHNN OmNonBm Nuwmmcmz HDCMUHMHcmHm pom mmuSuHUchxm. mmmm omsv .mmom vamm momm mmom some maem mmmm nmaaflxmc: .Asmw 0000 came HHHB .omvm mmsm Naom svoo snow nmaaflxmuwsmm .vmmv .mvmo macs meow .mmmo .svmm mmmm mmvm .mmmo nmaawxm boom vamo make 0005 mmsm vmmm .Nmmm omao Nome Hmoflumao ammo .mmam mmmm mmmm News ammk ommm mmva osom emsofiaemlmamm ommmm mosmm mamoam momma mummm mmmsm Heamm mmmma momma HmQOHmmmmoum um>o cam amumm vmnme mammmao v¢:mm wmnmm mm mmmmmau mmmmmao xz< mmmum>< mmmu®>< IOHUOm mumpcoumm NUMEHMQ mommmHU mm< mmapcmuawacmam uoc mmHSDHUchxm. .mawa mmma .amma mmma avma mama mmoa meda mmva pmaaaxWCD mava «mma amma omma .mama mmwa mmma amma mama omaaaXmlaEmm mmma vvma amma mmma .mmma mmma .amma vmma .mama omaaaxm mmma mvma .mmma mmma mmma mmwa amma mmma mmma amoaumau mmma amam .fimma maom .mmom .mfima mmma mmma .mmma pm>0aQEmeamm ommam mamam mmomm maomm aaamm mmmam momam mmmam m¢mam amcoammmwoum um>o ocm vmlmm vmlmv mommMaU vvlmm lemm mm mmmmmaU mommMaU xz< mmmum>< mmmum>< IOHUOm MMmocoumm NMdEaum mommMaU mm< mmo as. emumm amlmv mommmao vanmm amumm mm mmmmmao mmmmmao xz< mmmum>< mmmum>< IOHUOm Numocoumm wumEaum mommmau mm< mm0aQEmlmamm maoam wmmm mmaam moaam oaoam ommm ome mmmm vmm m amcoammmmoum um>o ocm vmimm vmlmw mmmmmau wwlmm vmlmm mm mmmmMaU mmmmmau xz< mmmam>< mmmum>< IOHUOm mumocoumm MMMEaum mmmmmau Umfl mmo 62m emumm amume mommmao valmm amumm mm mmmmMao mmmmmao xz< mmmam>< mmmam>< Ioauom Nwmpcoumm JNumEaum mommmau om< mm0aasmnmamm mmmam mmoam mmmam ammam amaam mmoam ammm mmoaw mmaam amcoammmmoum um>o mam amnmm amnmv mommmao aaumm emumm mm mmmmmao mmmmmao x2< mmmam>< mmmum>< IOHUOm Numocoumm NmMEaum mmmmmau Qm< mmo 6cm amumm amume mmmmwau eeumm amnmm mm mmmmmau mmmmmao xz< mmmum>< mmmum>< Ioauom Numocoumm Mameaum mmmmmau mm< mm< mm QAQo 6cm amumm amume mmmmmao «enmm emumm mm mmmmmao mmmmmao mz< mmmam>< momum>4 IOHUOm Numocoumm MHmEaum mmmmMaU 0m4 mm4AU m04 Q24 224m UHZOZOUMIOHUOm A4ZOHB4QDUUO Mm mom QZDMZOU OB mmHBHmzmmomQ mo4mm>4 :MMHBHmmmUmZ: vN QAQ4B (I) (I) () (D (3‘ 166 In the secondary age classes, the average propen- sity to consume, for all occupations, generally tends to be considerably lower than in the primary age classes. In the 45—54 age class those in the clerical category have the highest average propensity to consume for necessities, followed by the semi—skilled, unskilled, skilled, self- employed, and professional families. The occupations in the 55-64 age class generally have lower income—expenditure ratios for necessities than do families in the 45—54 age class. The professionals and self-employed have the low— est average propensity to consume and they are followed by skilled, semi—skilled, clerical, and unskilled. In the 65 and over age class, the average propensity to consume generally continues to decline for all the occupational groups except for the unskilled workers who have a higher income—expenditure ratio. The professionals have the low- est ratio followed by the self—employed, skilled, clerical, semi-skilled, and unskilled. The data for the averages of the primary age clas- ses generally reveal that with decreases in the occupational rank the average propensity to consume for "necessities" increases. An exception to the pattern occurs with those in the skilled category who have a slightly lower average propensity to consume. A similar relationship exists for the averages of the secondary age classes. 167 Total Food Expenditures Two hypotheses were set forth relating the propor— tion of income devoted to family total food expenditures. The first hypothesis states that in each successive occu— pational age class, the proportion of income devoted to total food expenditures increases in the primary age classes and decreases in the secondary age classes. The second hypothesis states that the average propensity to consume for total food expenditures in each age class of all occu- pations is inversely related to the socio—economic rank of the occupation. That is, the lower ranking occupations will demonstrate higher average propensities to consume than will higher ranked occupational categories. The data for the average propensities to consume for families in the primary and secondary age classes are presented in Table 25. The findings for the average propensity to consume for the primary age classes points out that the average propensity to consume for the professionals, clerical, and skilled groups increases with family age. However, it is also noted that self-employed and semi—skilled groups de— crease, then increase the proportion of income spent for total food. The unskilled, on the other hand, decrease the proportion of income for total food from the 25-34 to the 35—44 age class. In the secondary age classes, the professionals, 168 om. mm. mm. mm. mm. mm. mm. mm. mm. vmaaaxmc: am. mm. mm. mm. mm. mm. om. mm. mm. umaaaxmuasmm om. mm. em. mm. mm. mm; am. mm. em. vmaaaxm mm. mm. «m. mm. mm. am. am. mm. mm. amuaumao ma. ma. ma. ma. mm. ma. mm. mm. om. um>OaaEmuwamm ma. ma. ma. ma. am. om. om. am. ma. amcoammmmoum um>o 6cm amnmm amnme mommmao aaumm «mumm mm mmmmmao mmmmmao xz4 mmmum>< mmmum>4 IOHUOm (Numodoumm [NumEaum mwmmmau 004 mm4AU mw4 Q24 224m UHEOZOUQIOHUOm A4ZOHB4QQUUO Nm mmmDHHszme QOOh A4808 mom QEDmZOU OB mMHBHmzmmomm m04mm>4 mm mam48 169 clerical, and skilled families generally decrease the pro— portion of income for total food in the secondary age clas— ses. The self—employed increase, then decrease the rela— tive amount of income for total food, but the semi—skilled families decrease, then increase total food expenditures. The unskilled workers decrease the proportion of income for total food, then increase the proportion in the 65 years and over age class. The uniform pattern of an increasing average pro- pensity to consume with decreasing occupational rank is not strictly maintained in any of the family age classes. In the nonmanual occupational groups in all the age classes, the self-employed have both higher and lower average pro- pensities to consume than the professional group. The oc— cupations in the manual categories follow a more uniform pattern; however, some variations do occur in the under 25 age class. The most uniform patterns are found in the averages for the primary and secondary age classes. Clothing Expenditures The hypothesis relating the income-expenditure ratios for family clothing expenditures states that in each successive occupational age class, the proportion of income devoted to clothing expenditures increases the primary age classes and decreases with family age in the secondary age 170 classes. The average propensities to consume for clothing are presented in Table 26. An examination of the data for the primary age clas— ses reveals that the average propensity to consume increases with increasing family age for the professionals, semi- skilled, and unskilled workers. The self-employed families decrease, then increase the proportion of income for family clothing. The clerical families devote the same proportion of income in the under 25 and 25—34 age classes, then in- crease the proportion. The skilled workers initially in— crease the proportion of income, then maintain the ratio in the 25—34 and 35-44 age class. In the secondary age classes, all the occupational groups decrease the proportion of income for clothing with increasing family age except for the semi-skilled group which maintains the same proportion in the 55—64 and 65 and over age class. The data for the averages of the primary age clas— ses reveal that the professional category devotes the small— est proportion of income toward the purchase of clothing. By contrast, the self—employed and clerical groups have the highest average propensity to consume and are followed by the skilled, semi—skilled, and unskilled groups. The data for the average of the secondary age clas- ses disclose that all the occupations devote a similar pro— portion of income for clothing except for the unskilled families who devote slightly less income. 171 mo. mo. 0a. mo. Na. aa. 0a. aa. 0a. wmaaaxmc: mo. mo. aa. oa. Na. oa. mo. aa. aa. Umaaaxmlaemm mo. mo. aa. oa. oa. oa. mo. oa. oa. omaaaxm mo. mo. ma. oa. ma. oa. oa. Na. aa. amuaumau mo. mo. aa. oa. va. mo. ma. ma. oa. ommoademlmamm mo. mo. oa. oa. oa. mo. mo. mo. oa. amcoammmwowm um>o paw vmlmm vmlmv mommMaU walmm vmlmm mm mmmmMaU mmmmMaU 224m mm mm< smug: mm< mm< awm€aum aa4 uazozoum mmmum>4 mmmum>4 mmmum>4 IOHUOm mumocoumm NMMEHMQ mmmmMaU mm4 mm4AU mw4 Q24 224m UHSOZOUMIOHUOW A4ZOHB4QDUUO Mm mmmDBHQZMQXm 02H$BOAU mom QZDmZOU OB mmHBHmZQQOMQ mm4mm>4 mm mam48 172 House Furnishings and Equipment Expenditures The average propensity to consume for house furnish- ings and equipment expenditures reveals a different pattern than for some of the other income—expenditure relationships. The hypothesis formulated states that in each family age class, the proportion of income devoted to house furnish— ings and equipment expenditures is inversely related to socio-economic occupational rank. The data for the aver— age propensity to consume for house furnishings and equip— ment expenditures are presented in Table 27. The data indicate that for average of all age clas— ses, there is very little difference in the average propen— sity to consume for house furnishings and equipment expen- ditures. However, in the primary age class, specifically in the under 25 age class, considerable differences are noted. In the under 25 age class, for example, those in the unskilled category devote the highest proportion of income to house furnishing and equipment expenditures. The unskilled are followed by the semi-skilled, profes- sionals, skilled, clerical, and self-employed. The skilled families in the 25-34 age class have the largest average propensity to consume. However, in the 35-44 age class the professionals, self—employed, clerical, and skilled workers have identical average propensities to consume. The semi—skilled and unskilled workers have slightly higher average propensities to consume. 173 ao. mo. vo. mo. mo. mo. ma. mo. mo. cmaaaxmcs 40. mo. we. we. mo. mo. oa. mo. mo. nmaaaxmnasmm mo. mo. ao. ao. mo. mo. mo. mo. mo. wmaaaxm mo. mo. mo. so. mo. mo. oa. mo. mo. amuaumao mo. so. mo. mo. mo. ao. mo. mo. 40. Umsoaesmnwamm mo. go. «O. we. mo. mo. mo. mo. mo. amcoammmwoum um>o 62m vamm emnme mmmmmao aaumm amumm mm mmmmmao mwmmmao zz4 mmmum>4 mmmum>4 IOHUOm humocoumm (NMMEaum mmmmMaU mm4 mm4QU m04 Q24 224m UHEOZOUQIOHUOW A420HB4QDUUO 2m mmmDHHQZQQXm EZQEQHQOM Q24 UZHSWHszm mmbom NAHS4M mom QEDmZOU OB MMHBHmzmmomm m04mm>4 mm mam48 174 In the secondary age classes, there is no identifi- able pattern for the average propensity to consume. For example, in the 45—54 age class the professionals, skilled, semi-skilled, and unskilled workers have identical average propensities to consume. Although slightly lower, the self— employed and clerical families also devote an identical proportion of income toward the purchases of house furnish- ings and equipment. In the data for the averages of the primary age classes, the professionals and self—employed distribute the smallest proportion of income for house furnishings and equipment. Following the professionals and self-em- ployed are the clerical, skilled, and semi-skilled families who have identical average propensities to consume. The unskilled have the highest average propensity to consume for the age class. The data for the averages of the sec— ondary age classes do not follow such a uniform pattern. As an example, the professionals have a higher average pro- pensity to consume than do the self-employed. Occupational Rank and Family Disposable Income Concentrations Two hypotheses were formulated with respect to fam— ily income patterns. The first hypothesis states that in the non-manual occupational categories (professionals, self- employed, and clerical families), disposable income levels are highest in the secondary age classes. The second 175 hypothesis states that in the manual occupational classes (skilled, semi—skilled, and unskilled) disposable income levels reach maximum levels in the primary age classes. Table 28 illustrates the income concentration for both the non-manual and manual groups. Non-manual Occppational Categories The average disposable income levels for the non— manual occupational categories (all age classes combined) follow the general ranking scale except that the self—em— ployed category has a slightly higher average income than the professional group. Considerable differences exist between the clerical, professional, and self—employed groups. In the primary age classes, average disposable in- come for the occupational categories does not follow the socio-economic ranking scale in all the family age classes. In the under 25 age class, clerical families have higher disposable incomes than do those in the self-employed group. The data for the 25-34 age class show that the self—employed have higher disposable incomes than either the clerical or professional families. But, in the 35-44 age class, the family disposable income levels follow the socio-economic ranking scale. In the secondary age classes the disposable income levels follow the occupational ranking scale in the 45-54 and 65 and over age classes. In the 55-64 age class, how- ever, the disposable income level of the self—employed exceeds 176 .coaHODawu mm mnu Op mcaououum Dcmwmmwap haucmuamacmam uoz. mmmw mamm mmvm mmmm mmmm ammv momv amam mamm UmaaHXmCD .amom ammm mamm mmmm mmmm mmmm mvmv amom movm Umaaaxmlaemm .mmmm .mmmm mmmm mmmm .mmam mmmm mmmm mmmm .voam Umaaaxm ommm mmmm mmam mamm ommm mmmm ammm mamm moam amoaumau vmmm mmama .mmmaa oomaa mamm mamm vvfiv mmmm .mmmoa Um>OadEmlmamm mmmmam ommoam ovmaam omvaam ommmm ammmm momma mmamm vmaoam aMCOammmmoam wm>o ocm vmlmm vmnmv mommMaU velmm vmlmm mm mmmmMaU mmmmmau 224m mm mm< amass mo< wm<. a4 mommm>4 mmmwm>4 IOHUOW Numocoumm NHMEaum mmmmMaU mm4 mm4qU m04 Q24 224m UHZOZOUQIOHUOW A420HB4QDUUO Mm m4m>m4 mm mqm48 MZOUZH mam4m0mmHQ NAHE4M 177 that of the professionals. The data also point out that the disposable income levels of the self-employed and cler- ical groups decline with increases in family age. However, the disposable incomes of those in the professional cate- gory increase in the 65 and over age class. The data for the average disposable incomes for the primary age classes follow the ranking scale. However, the averages of the secondary age classes do not since the self—employed have a slightly higher disposable income level. Manual Occupational Categories The average disposable income for the manual work— ing classes (all age classes combined) follows the socio— economic ranking scale. In the primary age classes, income increases with family age for all the occupations. Further, with increases in family age, the differences in disposable income between the skilled and semi-skilled workers remain relatively stable. However, the differences in disposable income between the semi-skilled and unskilled groups tend to increase. In the secondary age classes, the family disposable incomes again follow the occupational ranking scale. The data also reveal that all the families experience sharp declines in disposable income. The average income levels for both the primary and secondary age classes correspond to the socio-economic rank of the occupations. 178 Family Absolute Savings Levels Two hypotheses were formulated relating family sav— ings levels with the age and occupation of the family head. The first hypothesis states that in each family age class, the family savings levels of the non—manual occupational categories do not follow the socio-economic ranking scale. The second hypothesis states that in each family age class the family saving levels of the manual occupational cate- -—..—+ d.“ .. 1 I gories do follow the socio-economic ranking scale. ! The definition of total family savings used in this study is broader than the definition of "liquid cash hold— ' ings"; however, it gives a more realistic view of the total family savings patterns. Savings are defined as: + the increase in assets — the decrease in assets — the increase in liabilities + the decrease in liabilities + personal insurance In Table 29 the absolute savings patterns are pre- sented for both the non-manual and manual occupations. Non-manual Family Savings Patterns Those in the self-employed category have the high— est average savings (all age classes combined), followed by the professional and clerical workers. Family saving patterns in the under 25 age class follow the socio-economic ranking scheme. However, in the 25-34 age class, the self- employed families demonstrate the highest saving levels, 179 osm omm mmm som mma mu mmm: moa osm umaaaxmca mmm msm mmm mam mma mma man msa mam smaaaxmnaEmm msm mam omm oam mam kma moa mms smm nmaaaxm msm mmm mss mmm mom mmm ssa mmm mmm amuaumau amsm amma ossm mmam ama mmmm mmm moma mmma ammoaQEmumamm mmm a mmmaa smmaa Nomad mmaaa maoaa mamas mmaam mmaaa amcoammmmoum um>o saw smumm smums mommmau ssamm smumm mm mmmmmau mmmmmau xzsm mm mo< umncb 664 66¢ a4 mmmwm>4 mmmum>4 IOHUOm humpcoumm NMMEaum mmmmmau 004 mmma muza>ma mo. mrp um ucmuawacmam. .pcmauamwmou muauaumMam mnp mo wound oumvcmumm mmmmmau um4 mma. mom. mma. mom. 0mm. mom. uouwm pamocmvm mmm. .omm. .vmm. .mmm. .mmm. .mma.a owaaameD mmm. mmm. mma. mma. mmm. moa. wouum pwmocmum mmm. .mmm. .mmm. .vmo. .smm. .mmw. owaaaxmlaemm mmm. mma. maa. mmm. mmm. mam. uouum pumpcmum amm. .mam.a .mmm. .ssm. .mmm. .mom.a omaaaxm mmm. sma. mam. 0mm. ama. mmm. aoaum pamUCMDm mmm. .mmm. .mvo. .mmm. .msv. mmm. amuaumau moa. amo. omo. amo. maa. ooo. aouwm pamocmum .kmm. .mmm. .mmm. .mom. mma. ooo. sm>0aQEmuwamm ama. msm. oma. mas. sma. oaa. uouwm pampCMpm .mmm. .mmm. .ovm.a mow. .mma.a mam. mamcoammmmoum um>o Dam mm vmlmm swims sslmm vmlmm mm umUCD xcmm Mumocoumm NWMEaam amcoauMQduuo uancoumloauom mm4qu QO4 Q24 224m A4ZOHB4QDUUO UHZOZOUQIOHUOm Mm hmmmDBHQZQQXm Q4808 mom mMHBHUHBm4AQ mmDBHQZQQXQIQZOUZH Om mdm4B 184 have the lowest elasticity coefficients and are followed by the semi—skilled. It is noted, however, that the elas— ticity coefficient for the clerical category is not signif— icantly different at the .05 level. In the 25-34 age class, the self-employed families demonstrate the lowest elasticity coefficient. They are followed by the clerical, unskilled, semi—skilled, skilled, and professionals. In the 25-34 age class, however, the elasticity coefficients for the self-employed are not statistically significant. The mag- nitude of elasticity coefficients in the 35—44 age class are inversely related to occupational rank except in the unskilled classification. The unskilled have a lower elas— ticity coefficient than do the semi—skilled. It is noted that the professional elasticity coefficient does not meet the .05 statistical level. In the 45—54 age class, several diverse patterns are noted in the relationship between occupational rank and total expenditure elasticities. The self—employed families demonstrate the lowest elasticity coefficients. They are followed by the semi—skilled, unskilled, skilled, clerical, and professional families, respectively. Both the professional and clerical families demonstrate elastic expenditures. In the 55-64 age class, the professionals have the lowest elasticity coefficient for total expendi- tures. The professionals are followed in the magnitude of income elasticities by the clerical, unskilled, semi- 185 skilled, self-employed, and skilled families. The skilled families are the only occupational group in the age class which demonstrates elastic expenditures. In the 65 and over age class, the elasticity coefficients for the cler— ical, skilled, semi-skilled, and unskilled do not meet the .05 statistical levels. The unskilled families have the lowest elasticity coefficients. The unskilled families are followed in order of increase of elasticity coeffici— ents by the self—employed, professional, clerical, skilled, and semi—skilled families. All of the occupations demon- strate inelastic expenditures for total expenditures in the 65 and over age class. Total food elasticities The second hypothesis states that the magnitude of income elasticities within each occupational age class is inversely related to the socio-economic rank of the occupations. In the under 25 age class, the only families dem— onstrating elastic expenditures are in the skilled group. Elasticity coefficients could not be computed for the self- employed in this age class. The professional families dem- onstrate a unique income-total food expenditure relation— ship. As noted in Table 31, the elasticity coefficient signifies that total food expenditures are an inferior good. Inferior goods are those which decline with increases in family incomes. Of the remaining family groups in the under 186 .am>ma mo. wnu um pcmuamacmam. .ucmauawmmou wuauaumMam 03p mo uouum pumocmumm mma. mma. mom. mmm. mmm. mam. wowwm pumpcmum .mom. .mmm. .Nmm.a mmm. .amm.a mmm. UmaaaXmCD mmm. mmm. mam. smm. mmm. mmm. uouwm oumocmpm mmm. mmm. mss. .mmm. mmm. sdm. omaaaxmlaemm oam. msm. amm. amm. mmm. msv. uowum pumocmvm omm. .maa.a oms. oas. .mmm. .moa.a omaaaxm mso.a vva. was. sma. mma. mmm. aoaum oumocmum mmm.a .mmm. sva.a .msm. mmo. mmm. amuaumau mmo. Nma. mmo. oso. mma. coo. uoaum pumocmum .smm. .smk. .mmm. .amm. aom. ooo. smsoaaemnwamm ama. mmm. mmm. msv. mam. mam. muowwm pumpcmum .mam. mmm. .mmm. mmm. .msm. asm.l mamcoammmmowm am>o ocm mm smimm swims sslmm vmlmm mm umpda Mama Numocoumm mamEaum amaoapmdzuuo uancoumloauom mmmmmau 004 . . mm4AU m04 Q24 224m Q4ZOHB4QDUUO UHZOZOUQIOHUOW Mm mmmDBHQZQQXm QOOh Q4808 mom mmHBHUHBm4dm mmDBHQZMQXmlmZOUZH am QQQ4B 187 25 age class, the semi-skilled have the lowest elasticity coefficient. The semi—skilled families are followed by the unskilled, clerical and skilled families, respectively. However, only the skilled families have elasticity coeffi- cients which meet the .05 significance criterion. In the 25-34 age class, the unskilled families have the only elas— tic total food expenditure coefficient. The ranking of occupational families from lowest to highest income elas- ticities is as follows: clerical, self-employed, profes- sional, semi-skilled, skilled, and unskilled. However, only the elasticities in the professional, skilled, and unskilled groups meet the statistical significance cri- terion. In the 35—44 age class the professionals have the lowest elasticity coefficients. They are followed by the self—employed, skilled, unskilled, clerical, and semi—skilled families in order of increasing income elasticity coeffici— ents. The elasticity Coefficients for the professional, skilled, and unskilled families, however, do not meet the statistical significance criterion. In the 45—54 age class, the clerical and unskilled families are the only ones demonstrating elastic total food expenditures. The skilled families have the lowest total food elasticity coefficients. The skilled families are followed by the semi—skilled, self—employed, professional, clerical, and unskilled families. All the elasticity co- efficients meet the .05 significance level, except for 188 clerical, skilled, semi-skilled, and unskilled families. In the 55—64 age class, the ranking of occupations accord— ing to the magnitude of the elasticity coefficients, from lowest to highest, is as follows: professionals, clerical, semi—skilled, self—employed, unskilled, and skilled fami— lies. All the elasticity coefficients meet the significance level except for the professional and semi—skilled families. In the 65 and over age class, only the clerical families demonstrate elastic expenditures for total food expenditures. The skilled families have the lowest elasticity expenditures and are followed by the unskilled, self-employed, profes- sional, semi-skilled, and clerical families. All the elas- ticity coefficients meet the .05 significance level except the one for the skilled families. Clothing elasticities The third hypothesis states that the magnitude of income elasticity coefficients for clothing is inversely related to socio-economic ranks of the occupational groups within each age class. Again in the under 25 age class, the elasticity coefficients for the self—employed clothing expenditures cannot be computed. The elasticity coeffici— ents in Table 32 reveal several diverse patterns. In the under 25 age class, the clerical and skilled groups both demonstrate elastic expenditures. However, the other groups have inelastic clothing expenditures. The lowest clothing elasticity coefficient occurs in the professional group. 189 .am>ma mo. wry um pcmuamacmam. .ucmauawwmoo muaoapmMam mnu mo aouam oumocmumm amm. mom. mmm. amv. mmm. mmm. uouam Unmocmum amm. amm. .mam.a mmm. mmo. wmm. omaaHcha smm. mmm. 0mm. mmm. mmm. mas. uouum Unmocmum .msm.m mom. mmm. mov.a .mmm.a mmm. UmaaaxmlaEmm mmm. mma. mas. mas. aam. mmm. uouum pumpcmum aoa. .amm.a .mom.a mvm. ass. .mmo.m Umaaaxm mmm. ama. mmm. was. mms. vmm° wouam UMMUCMDm 0mm.a .msm.a msm.a .omm.a omm. omm.a amuauoau amm. ova. oma. mma. mmm. ooo. woaam pumpcmum .mmm.a .mmo.a .mmm. .mmm. mmm. ooo. sm>Oaasmnwamm ssm. mmm. mmm. mmm. mss. vmm. muoawm UMMUCMbm .mmm. .mma.a ssm. mmm. mmm. sma. mamcoammmmoum um>o paw mm smlmm swims svlmm vmlmm mm amps: xcmm Mummcoumm . NMManm aMCOHDMQSUUO UaEOCOUdIoauo mommmau mw4 . . m mm4qu mw4 Q24 224m UHZOZOUQIOHUOW Mm mmmDBHQZQQXQ DZHEBOQU NAHZ4® mom mmHBHUHBm4Am mmDBHQZMQXQImZOUZH Nm mdm4B 190 In increasing elasticity magnitudes, the professionals are followed by the unskilled, semi-skilled, clerical, and skilled groups. Only the skilled families, however, have elasticity coefficients which are statistically significant. In the 25-34 age class, only the semi—skilled fami— lies demonstrate elastic clothing expenditures. The low- est elasticity coefficients are found in the unskilled group. The unskilled families are followed by the self—employed, clerical, skilled, professionals, and semi—skilled families. Only the elasticities for the semi—skilled groups are sta- tistically significant in the 25-34 age class. In the 35- 44 age class, the clerical and semi—skilled families both have elastic clothing expenditures. The lowest elasticity coefficient is in the unskilled group. The unskilled fami— lies are followed by skilled, self—employed, professional, clerical, and semi-skilled families, respectively. Only the self-employed and clerical groups have elasticity co— efficients which are statistically significant in the 35—44 age class. In the 45~54 age classification, the clerical, skilled families, and unskilled families demonstrate elas— tic clothing expenditures. The lowest elasticity coeffici— ent is found in the professional category. The profession- als are followed by semi-skilled, self—employed, clerical, Skilled, and unskilled families. The elasticity coeffici- ents are statistically significant in the self-employed, 191 skilled, and unskilled families. The coefficients are elas— tic for all the occupational categories in the 55-64 age class, except for semi-skilled and unskilled groups. The semi-skilled families have the lowest elasticity coeffici— ents, and they are followed by those in the unskilled, self- employed, professional, clerical, and skilled groups. All the coefficients are statistically significant, except for the semi—skilled and unskilled families. In the 65 and over age class, the self-employed, clerical, and semi-skilled groups have elastic clothing expenditure coefficients. The skilled have the lowest elas— ticity coefficient. The skilled families are followed by those in the professional, unskilled, clerical, self—employed, and semi—skilled families. The elasticity coefficients for clerical, skilled, and unskilled families are not statis- tically significant at the .05 level. Tptal housing elasticities The fourth hypothesis states that the magnitude of elasticity coefficients for total housing expenditures within each age class is inversely related to the socio- 8conomic occupational rank. The total housing elasticity coefficients presented in Table 33 illustrate the diversity Of the responsiveness of expenditures to changes in income. In the under 25 age class, the skilled and unskilled groups demonstrate elastic total housing expenditures. The total housing elasticity coefficients cannot be computed 192 .am>ma mo. mes um pemuamaeeam. .ucmauammmou >pauapmMam may mo uouum oumocmumm mma. mom. aom. 0mm. mmm. mmm. wouwm pumpcmum .ams. mmm. .mmm. .mmm. mmm. ama.a Umaaaxmsb mmm. mmm. mmm. oom. msm. amm. uoaam pwmocmum mmm. .mam. .amm. mmm. .mmm. amm. UmaaHXmlaEmm mmm. som. 0mm. amm. smm. omm. wouwm pampcmum mmm. ovo.a ssm. msm. .mmm.a .mmm.a poaaaxm mum. smm. ams. msm. mma. msm. uouum sumeeeum mmm. .mms.a smm. mmm. .mmm. mma. amoaumau mmm. mmo. moa. smo. com. 000. uoaam pwmocmpm mmm. .mmm. .mmm. .mmm. mam. eee. emsoaeemueamm mmm. mmm. mmm. mmm. mmm. mmm. muouum pummcmam mmm. mms. .amm.a .mmo.a .mmm.a mmm. mamcoammmmoum um>o Ucm mm smlmm smlms sslmm smlmm mm amps: Mcmm MMmpcoumm Numsaum amcoawmdsuuo uancouwloauom mmmmmau um4 . . mm4QU m04 Q24 224m UHZOZOUQIOHUOm Nm mmmD8HQ2mmXQ 02Hm302 Q4808 mom mMH8HUH8m4Qm mmD8HQ2mQXQIQZOUZH mm mam48 193 for the self-employed group since not enough sample fami— lies are in that occupational-age category. The clerical families have the lowest elasticity coefficient. They are followed by the semi—skilled, professionals, unskilled, and skilled families, respectively. The elasticity coef- ficients for the professional, clerical, semi—skilled, and unskilled are not statistically significant at the .05 level. In the 25-34 age classification, the unskilled fam- ilies have the lowest total housing expenditure elasticity coefficient. They are followed by the self—employed, semi— skilled, clerical, skilled, and professional families. The elasticity coefficients are statistically significant for all the occupations, except for the self-employed and unskilled groups. In the 35—44 age class, the only occu- pational group having elastic total housing expenditures are the professional families. The other occupations ranked according to the magnitude of income elasticities are cler- ical, self-employed, semi-skilled, unskilled, and the pro— fessionals. The clerical, skilled, and semi—skilled total housing elasticity coefficients are not statistically sig- nificant. The elasticity coefficients in the 35-44 age class, ranked according to occupation with the lowest to the high— est elasticity coefficients, follow the following sequence: clerical, self-employed, semi—skilled, skilled, unskilled, and professionals. The professional families are the only 194 occupation demonstrating elastic expenditures. In the 35- 44 age class, the elasticity coefficients for the clerical, skilled, and semi—skilled, and semi-skilled groups do not meet the .05 level of significance. The professional families are the only occupational groups demonstrating elastic total housing expenditures in the 45-54 age class. The unskilled have the lowest in— come elasticity and are followed by those families in the self-employed, skilled, semi—skilled, clerical, and profes— sional groups. The elasticity coefficients for the cler— ical and skilled families do not meet the .05 level of sig— nificance. In the 55-64 age class, both the clerical and skilled families have total housing coefficients which are elastic. The lowest elasticity is found in the professional occupa— tional category. The professional families are followed by the unskilled, semi—skilled, self-employed, skilled, and clerical groups. The coefficients for the professional, skilled, and unskilled families are not statistically sig— nificant at the .05 level. In the 65 and over age classification none of the occupations demonstrate elastic total housing expenditures. The lowest elasticity coefficients occur in the skilled families. They are followed by the professional, self- employed, unskilled, semi-skilled, and clerical families. Only the unskilled group demonstrates elasticity coefficients 195 which are statistically significant at the .05 level. Marginal Income—Expenditure Relationships One hypothesis was formulated to relate the rela— tionship between a change in income and a change in expendi~ tures. The hypothesis states that in each family age class the marginal propensity to consume for total expenditures is inversely related to occupational rank. The marginal propensity to consume coefficients in Table 34 reveal sev— eral diverse patterns. In the under 25 age class, for ex— ample; the lowest marginal propensity to consume for total expenditures appears in the professional category. Follow“ ing the professional families in the magnitude of the mar— ginal propensities to consume are the clerical, semi-skilled, skilled and unskilled families. As in the case of the in— come-expenditure elasticities, the marginal propensity to consume for the self—employed families cannot be calculated. The marginal propensity to consume coefficients for the professional and clerical families are not statistically significant at the .05 level. In the 25—34 age class the self—employed families have the lowest marginal propensity to consume for total expenditures. They are followed by the clerical, unskilled, semi-skilled, skilled, and professional families in the magnitude of the marginal propensity to consume. The mar— ginal propensity to consume coefficients are not statistic- ally significant at the .05 level for the self-employed, 196 .am>ma mo. may um semuawaemam. .pcmauammmou oESmcou ou xpamcmdoud amcamume msu wo wowuw pwmocmumm mma. mma. mma. mma. mmm. mma. uowwm pumpcmum msm. .amm. .mmm. .omm. msm. .mmm.a UmaaameD aom. mom. ama. ama. mmm. soa. uoaam Uumocmum mmm. .omm. .mmm. .mvo.a .mmm. .omm. UmaaaxmlaEmm mmm. mma. mmo. mmm. mmm. smm. uowwm pamocmpm mmm. .moa.a .amm. .mmm. .mmm. .sma.a omaaaxm amm. ama. mmm. mom. mma. amm. uoawm pumpcmum mom. .mmm. .aso.a .amm. mmm. mmm. amuaumau moa. sso. mvo. oma. moa. ooo. aowum Uumpcmum .mmm. .mee. .sem. .mme. ewe. oee. em>0aeemuwamm smo. mmm. mma. mmm. mma. amo. wouum pumpcmum .mmm. .msm. .mma.a mas. .Omm. mmo. mascoammmmoum um>o pcm mm smlmm swims sslmm smlmm mm meCD xcmm Namosoumm wumEaum ascoapmdsuuo mmmmMaU mm4 uancoumloauom mm4du m04 Q24 224m A420H84QDUUO UHZOZOUQIOHUOm Mm .mmmD8HQ2mmxm A4808 20m mZDmZOU O8 MMH8Hm2mmomm 442me42 flm mdm48 197 clerical, and unskilled groups. The magnitude of the marginal propensity to consume coefficients generally follow the occupational ranking scale in the 35-44 age class. The professional families have the lowest marginal propensity to consume of any occupation. They are followed by self-employed, skilled, clerical, un- skilled, and semi-skilled families. All of the marginal propensities to consume are statistically significant, ex— cept for the professional families. In the 45-54 age classification, both the profes- sional and clerical families have marginal propensities to consume which are greater than one. However, the low— est marginal propensity to consume is found in the self— employed category. The self-employed families are followed by those in the semi—skilled, skilled, unskilled, clerical and professional occupations. All the marginal propensi— ties to consume are statistically significant in the 45—54 age class. The professional families have the lowest marginal propensities to consume in the 55-64 age class. The pro— fessional families are followed by the unskilled, self- employed, clerical, semi—skilled, and skilled families. All of the marginal propensities to consume are statistic— ally significant in the 55—64 age class. In the 65 and over age class, the marginal propensi— ties to consume are all smaller than those in the 55-64 age 198 classes. The lowest marginal propensity to consume is found in the professional category. The professional families are followed by those families in the unskilled, clerical, self-employed, skilled and semi—skilled group. Only the marginal propensities to consume for the professional and self—employed groups are statistically significant at the .05 level. Summary The purpose of Chapter IV was to present the find— ings of the research. The findings based on the hypotheses of Chapter I illustrated the influence of the age—occupa— tion variables on family expenditure patterns. In present— ing the findings several relationships were revealed between the age—occupation variables and absolute expenditures, relative expenditures, disposable income levels, absolute savings levels, income—expenditure elasticities, and the mar— ginal propensities to consume for total family expenditures. The findings pertaining to absolute expenditures revealed several important relationships. The data for family total expenditures, for example, revealed that ab— solute expenditures generally followed the occupational ranking scale. Several variations between occupational rank and total expenditures were noted, however, in the under 25 age class. The relationship between occupational rank and total food expenditures was not as clearly defined as for total family expenditures. Only in the 45-54 age 199 class did absolute food expenditures completely follow the ranking scale. The expenditure sequence for total housing expenditures usually followed the occupational ranking scale although considerable expenditure variation occurred in the under 25 age class. Further, variations in expenditures were found between the professional and self-employed fam— ilies in the 55-64 and 65 and over age classes. Several variations between family clothing expenditures and occu— pational rank were found to exist in the various occupation— age classes. In most of the age classes, the self—employed had higher clothing expenditures than did the professional families. Several variations in clothing expenditures also were noted in the under 25 age class. Several other varia- tions also were found in tne house furnishings and equip- ment expenditures. The most significant variations were found in the under 25 age class, where clerical and semi— skilled families demonstrated the highest expenditures. In most of the other age classes, the lower ranked occupa- tions had higher expenditures than the clerical group. Considerable variations in automobile expenditures also were found. In the primary age classes, high absolute automobile expenditures were found in the skilled and semi- skilled groups. Variations in automobile expenditures also were found in the secondary age classes. The findings regarding the average propensity to consume for total family expenditures indicated that the 200 proportion of income devoted to total expenditures gener- ally increases with the lower ranking occupations. Further, the average propensity to consume generally decreases with family age in the secondary age classes. Similar relation— ships also hold for the "necessity” expenditure accounts. The average propensities to consume for total food expendi— tures also usually increased with decreases in occupational rank. Relative expenditures for family clothing outlays remained relatively constant in most of the occupational- age classes. However, with increases in family age in the primary age classes the proportion of income devoted to clothing generally increases. By contrast, in the second— ary age classes the proportion of income devoted to cloth— ing expenditures generally decreases. The findings also revealed that the proportion of income devoted to house furnishings and equipment expenditures generally is higher for the lower ranking occupations. The greatest variation in the average propensities to consume in the lower rank— ing occupations were found in the under 25 age class. Only in the 45-54 age class did the family dispos— able income levels completely follow the socio—economic ranking scale. Usually, however, the disposable income levels did follow the socio—economic ranking scale. The findings dealing with the absolute family sav— ings levels also revealed considerable variations among the occupational groups. Generally, the self-employed 201 exhibited higher savings than the professional group. And, on the average, the skilled families generally had higher savings than the clerical families. The remaining occupa— tional groups generally followed the socio—economic ranking scale. The findings on the income—expenditure elasticities for total expenditures revealed that in most of the age classes the lower ranking occupations have higher income- elasticities than the higher ranked occupations. Most of the income—expenditure elasticities for total food expendi— tures are inelastic. However, a definite pattern for the relationship between socio-economic rank and the total food elasticities was not always present in each age class. Several elastic family clothing expenditures coefficients were found in the 35—44, 45-54, 55—64, and 65 and over age classes. Although a definite pattern was not clearly es- tablished for any age class, the clerical and skilled groups generally had higher clothing elasticities. The income— expenditure elasticities for total housing expenditures indicated that total housing expenditures generally were inelastic. However, the professional, clerical, skilled, and unskilled families had elastic expenditures. The data on the marginal propensities to consume for total expenditures revealed that the highest ratios generally were found in the 35-44 and the 45—54 age classes. The marginal propensities to consume usually decreased with an increase in family age in the secondary age classes. CHAPTER V SUMMARY AND CONCLUSIONS Introduction The purpose of this chapter is to present the sum- mary and conclusions of the research. The chapter is di— vided into five sections. Section one evaluates the hypoth- eses presented in Chapter I. The conclusions of the research are presented in the second section. The family economic profiles presented in the conclusions section are based on data presented in Chapter IV. The third section com— pares family expenditures by occupation and chronological age to expenditures by social class, family life cycle stage, and income class. The implications of the occupation-age approach are presented in the fourth section. Finally, in the fifth section the suggested areas for further research are discussed. Evaluation of Hypotheses Socio-Economic Occupational Rank and Absolute Expenditure Relationships 1. In each family age class average total dollar expenditures are_positively_related to the socio—economic rank of the occupation. 202 203 From the data presented in Table 17, the hypothesis cannot be accepted for two reasons. First, in the under 25, 25—34, 55-64, and 65 and over age classes, total family expenditures do not completely follow the socio—economic ranking scale. In the under 25 age class, the clerical, skilled, semi—skilled, and unskilled occupational groups follow the socio-economic ranking system; however, the pro— fessional and self—employed categories do not. In the 25— 34 age class, all the occupations follow the ranking system except the skilled and clerical categories. The profession— als and self—employed and the clerical and skilled workers do not follow the ranking system in the 55—64 age class. In the 65 and over age class, however, all the occupations follow the ranking scale except for the skilled and semi— skilled workers. The second reason the hypothesis must be rejected is that the differences in expenditures among the occupa— tions in the 35-44 and 45-54 age class are not judged to be significantly different according to the 5 per cent criterion. 2. In each family age class average total food expenditures are positively related to the socio—economic rank of the occupation. On the basis of Table 18, the hypothesis related to each family age class cannot be accepted. In each of the age classes, families in the higher ranked occupational 204 categories do not always demonstrate higher total food ex— penditures. In the under 25 age class, the self-employed spend more than the professionals, and the semi-skilled workers spend more than the clerical workers or skilled groups. The data in the 25—34 year age class reveal that the self-employed spend more than the professionals and that the skilled and semi-skilled workers spend more than the clerical groups. The data for total food expenditures in the 35-44 age class reveal that the skilled workers spend slightly more than the clerical workers. However, the dif— ference is not significant according to the 5 per cent cri— terion. In the 45—54 age class, expenditures for total food do follow the socio-economic ranking scale. However, the differences in expenditures among the various occupations are not always significant. In the 55—64 age class, the total food expenditures do not follow the socio—economic ranking scale since the self—employed have higher outlays than do the professionals. The skilled group also has higher expenditures than the clerical group. Finally, in the 65 and over age class, semi-skilled and unskilled workers spend more on total food expenditures than either the clerical or skilled occupational groups. 3. In each family age class average total housing expenditures are positively related to the socio—economic rank of the occupation. 205 On the basis of the data presented in Table 19, the hypothesis cannot be accepted. In only two family age classes do total housing expenditures correspond to the occupational ranking scale. The 25-34 and the 45—54 age classes correspond to the occupational ranking scale; how— ever, in none of the age classes does a significant differ- ence in expenditures exist. 4. In each family age class family clothing gxpgnditures are_positively related to the occupational socio-economic rank. The data presented in Table 20 indicate that none of the expenditure data in the age classes completely fol— low the occupational ranking scale. In each age class, except the 65 and over group, the self—employed spend more than professionals for clothing. In the skilled, semi- skilled, and unskilled groups, clothing expenditures cor- respond to the occupational rank; however, the sequence of expenditures for the under 25 and 35—44 age classes can- not be judged to be significantly different. The absolute clothing expenditure hypothesis for each family age class cannot be accepted. 5. Family expenditures for house furnishing and equipment reach maximum levels in the primary age classes for each occupational category. The data in Table 2l indicate that all the profes- sionals, clerical, skilled, semi-skilled, and unskilled 206 categories reach maximum house furnishings and equipment expenditures in the primary age classes. Therefore, the hypothesis is accepted for these occupational categories. The self—employed category does not reach maximum expendi— tures for house furnishings and equipment expenditures un— til the 55-64 age class. The hypothesis cannot be accepted for the self—employed group. 6. Family expenditures for house furnishings and equipment do not follow the socio-economic occupational ranking scale in any age class. The data in Table 21 reveal that several irregulari— ties in expenditure patterns occur in the various age classes. In the under 25 age class, for example, the unskilled fami— lies demonstrate the highest expenditures of any occupation. And in the 25—34 age class, the professionals are followed in maximum expenditures by the skilled families. Expendi— tures for house furnishings and equipment in the 35—44 age class also do not follow the ranking scale. In this age class, the unskilled, semi-skilled, and skilled workers spend more on house furnishings and equipment than those in the clerical category. In the 45-54 age class, skilled and semi—skilled workers expend more on house furnishings than the clerical workers. Irregularities in expenditure patterns are also noted in the 55—64 and 65 and over age classes. Therefore, the hypothesis is validated for each age class. 207 7. Family automobile expenditures do not corres— pond to the socio-economic ranking scale in any age class. The hypothesis is accepted on the basis of the data in Table 22. In each age class, a lower ranking occupational category demonstrates higher expenditures than a higher ranking occupation. As indicated by the data in the under 25 age class, skilled workers have the highest absolute expenditures. Further, in the 25-34 and 35-44 age classes, the skilled workers rank second in average automobile ex— penditures. Similarly, the semi-skilled group spends ab— solutely more on automobile expenditures than either the self-employed or clerical in all the primary age classes. In the secondary age classes, other “irregularities” in expenditures are noted. The skilled group expends more than the self-employed in all the secondary age classes. 8. Maximum automobile expenditures for each of the manual occupational categories are con- centrated in the primary age classes while maximum automobile expenditures for each of the non—manual occupational categories are concentrated in the secondary_age classes. The data on the manual occupational categories pre— sented in Table 22 reveal that only the semi—skilled work- ers reach maximum automobile expenditures in the primary age classes. Therefore, the hypothesis is valid for the 208 semi—skilled occupational group. However, the skilled and unskilled occupational groups reach maximum automobile ex— penditures in the secondary age classes. As a result, the hypothesis for the skilled and unskilled groups cannot be accepted. The automobile expenditure data presented in Table 22 reveal that the professional and clerical groups reach maximum automobile expenditures in the secondary age classes. As a result, the hypothesis relating the non—manual expen— ditures for automobiles is accepted for the professional and clerical groups. But the hypothesis cannot be accepted for the self—employed category since maximum automobile expenditures are reached in the primary age classes. Socio-Economic Occupational Rank and Relative Income-Expenditure Relationships 1. In each family age class the average_propen— sity to consume for total family expenditures is inversely_related to occupational socio— economic rank. The above hypothesis states that the higher the occupational rank, the lower the average propensity to con— sume for total family expenditures. The percentage of dis— posable income expended for total expenditures is presented in Table 23. On the basis of the data for the average pro— pensities to consume, the hypothesis cannot be accepted for any of the family age classes. Generally, the average 209 propensity to consume increases as the occupational rank decreases; however, there are several significant varia— tions. In the under 25 age class, for example, the self- employed have the highest average propensity to consume. But in the 25—34 age class the self-employed have the low- est average propensity to consume. Irregularities are also noted in the 35-44 age class where the self—employed have a higher average propensity to consume than the professional group. Further, the skilled group demonstrates a lower average propensity to consume than the clerical or self— employed group. In the 45-54 age class, the self-employed have the lowest average propensity to consume. Further, the clerical group has the highest average propensity to consume of any age class. Similar irregularities are noted in the 55—64 age class. In the 65 and over age class, the self-employed group has the lowest average propensity to consume. 2. In each family age class the averagggpropen- sity to consume for the total of expenditure accounts classified as “necessities," is inverseiy related to occupational socio— economic ranking, On the basis of the data in Table 24, the hypoth- esis cannot be validated. Although the average propensity to consume for necessities more closely corresponds to the socio—economic ranking scale in each age class than do total 210 expenditures, there are still considerable variations among the occupational categories. Such variations are specific- ally noted in the primary age classes. The self—employed, for example, have a higher average propensity to consume than any of the other occupational groups in the under 25 age class. And following a similar relationship as for total expenditures, the self-employed have the lowest aver— age propensity to consume. Also, it is significant to note that the skilled and semi-skilled families have lower aver- age propensities to consume than do either the skilled or self—employed families in the 35-44 age class. In the sec- ondary age classes, the average propensity to consume for necessities does not follow the occupational ranking scale in any of the age classes. 3. In each successive occupational age class the ppoportion of income devoted to total food expenditures increases in the primary age classes and decreases in the secondary gge classes. On the basis of the data presented in Table 25, the hypothesis is not accepted for the occupational cate— gories in the primary age classes. The proportion of dis- posable income devoted to total food generally remains con— stant in the age classes of each occupation. In none of the successive age classes does the proportion of income devoted to total food expenditures continually increase. 211 In the secondary age classes, only the clerical and skilled workers exhibit a decreasing average propensity to consume. Therefore, the hypothesis can be accepted for clerical and skilled occupational categories. The other occupations generally experience gradually declining aver— age propensities to consume in the secondary age classes. 4. In each family age class of all occupations the average propensipy to consume for total food expenditures is inversely related to the socio-economic rank of the occupation. On the basis of Table 25, the hypothesis cannot be accepted. Although a general pattern emerges within the various occupational categories, that is, the lower ranking occupations having higher average propensities to consume than some of the higher ranking occupations. Never- theless, no strict inversed pattern is maintained through— out all the occupational groups in any age class. 5. In each successive occupational age class the ppoportion of income devoted to clothing ex- pgnditures increases in the primary age classes and decreases with family gge in the secondary age classes. On the examination of the data in Table 26, the hypothesis cannot be accepted for all of the occupational groups in the primary age classes. The hypothesis can be accepted for the professional, semi—skilled, and unskilled 212 occupational groups since the average propensity to consume increases in each successive age class. The self—employed decrease, then increase the proportion of income devoted to clothing expenditures in the primary age classes. The clerical group, by contrast, maintains the same ratio in the under 25 and 25—34 age classes, then increases the pro— portion of income. And the skilled workers initially in— crease, then maintain the same income—expenditure ratio in the 25-34 and 35-44 age classes. The hypothesis can be accepted for the professional, self—employed, clerical, skilled, and unskilled groups since the average propensity to consume for clothing expenditures decreases with each successive age class. Successive de— creases in the average propensity to consume are not found in the secondary age classes for self—employed workers. 6. In each family age class of all occupations the prgportion of income devoted to house furnishings and equipment is inverseiy_related to socio—economic occupational rank. The above hypothesis cannot be accepted for any of the occupational age classes. The data presented in Table 27 reveal that the average propensity to consume for house furnishings and equipment generally increases in the lower ranked occupations especially in the under 25 age class. The average propensity to consume also generally increases in the 35-44 age class. 213 In the secondary age classes, the average propensi- ties to consume for each occupation are similar in the 45- 54 and 55-64 age classes. In the 65 year and over age class, the semi-skilled have the highest average propensity to consume. Occupational Rank and Family Digposable Income Concentration l. Disposable income levels for all non—manual occupational families reach maximum levels in the secondary age classes. The above hypothesis is accepted on the basis of the data presented in Table 28 where higher disposable in— comes are found in the secondary age classes for all the non-manual occupational categories. The clerical families reach maximum disposable income levels in the 45-54 age class. By contrast, the self-employed and professional families reach highest disposable income levels in the 55- 64 and over age classes, respectively. 2. Digposable income levels for all manual work— ing families reach maximum levels in the primary family age classes. The above hypothesis cannot be accepted on the basis of the data in Table 28. In fact, all the manual occupational groups exhibit higher disposable income levels in the secondary age classes. The skilled and unskilled families reach their highest disposable income levels in 214 the 45—54 age class. The semi-skilled families, however, do not reach maximum income levels until the 55-64 age class. Absolute Savings Levels and Family_Age Class 1. In each family age class of the non—manual occupational categories family savings levels do not follow the socio-economic ranking scale. On the basis of the data presented in Table 29, the hypothesis is accepted for all the occupational-age classes except the under 25 age class. In the 25-34, 45— 54, 55—64, and 65 and over age classes, the self-employed families have higher savings than those in any of the other occupational categories. Another significant discrepancy in the occupational ranking scale and the family saving patterns occurred between the skilled and clerical families. Illustratively, the skilled families have higher total sav— ings in the 35-44, 45—54, and 55-64 age classes. 2. In each family age class of the manual occu— pgtional categories family saving levels follow the socio-economic ranking scale. On the basis of the data presented in Table 29, the hypothesis can be accepted for the under 25, 25-34, 35-44, and 45-54 age class. The hypothesis cannot be ac- cepted for the 55-64 and 65 and over age class. In the 55-64 age class, unskilled workers demonstrate higher sav— ing levels than do the semi-skilled workers. And in the 215 65 and over age class, the semi-skilled group has higher savings levels than does the skilled occupational group. Socio-Economic Occupgtional Rank and Income—Expenditure Elasticities l. The magnitude of income elasticity coefficients for total expenditures within each age class is inversely related to the socio—economic occupational rank of the occupations. From the data presented in Table 30, the hypothesis cannot be accepted for any of the occupational—age classes. Income elasticities are not completely inversely related to the occupational ranking system in any age class. How- ever, in most of the age classes the lower ranking occupa— tions do have higher income elasticities than the higher ranked occupations. Illustratively, in the under 25 age class, the elasticity coefficients are inversely related to occupational rank for the professional, clerical, and skilled families. In the 25—34 age class, elasticities again increase from the self—employed to the skilled group. Except for the unskilled group, the income elasticities are inversely related to occupational rank in the 35-44 age class. However, no uniform pattern for the elasticity coefficients appears in the 45-54 age class. In the 55-64 age class, the elasticities are generally lower for the higher ranked occupations but no clear pattern emerges. Finally, in the 65 and over age class, the elasticity 216 coefficients generally increase with the lower ranked oc— cupations. Except for the professional and unskilled groups the expenditures follow the ranking scale. Therefore, the hypothesis cannot be accepted. I 2. The magnitude of income elasticity coefficients for total food expenditures within each agg class is inversely related to the socio—economic occupational rank of the occupations. On the basis of Table 31, the hypothesis related to each occupational—age class cannot be accepted. In each of the age classes, families in the higher ranked occupa— tional categories do not consistently demonstrate lower income—expenditure elasticities for total food. In the under 25 age class, the food elasticities increase with decreases in occupational rank except for the semi-skilled and unskilled families. No general pattern emerges for the families in the 25-34 age class. However, in the 35- 44 age class the total food elasticities increase with de- creases in occupational rank between the professional and clerical groups. For the skilled, semi-skilled, and un— skilled families the total food elasticities decrease, in- crease, and decrease again. In the 45-54 age class, the food elasticities increase between the families in the self- employed and clerical categories. The food elasticities also increase between skilled and unskilled group. The total food elasticities in the 55-64 and 65 and over age 217 class do not reveal any identifiable pattern. The hypoth- esis cannot be accepted for any of the family age classes. 3. The magnitude of income elasticity coefficients for clothing within each age class is inversely related to the socio—economic occupational rank of the occupations. The data presented in Table 32 indicate that none of the elasticity coefficients for clothing expenditures in the age classes completely follow the occupational rank- ing scale. However, the elasticity coefficients for cloth- ing expenditures follow the sequence of the occupational ranking scale more closely than for total family expendi— tures and total food expenditures. The clothing elasticity coefficients in the under 25 age class increase with de- creases in occupational rank between the professional and skilled groups. This pattern, however, is not maintained in the semi—skilled and unskilled occupational groups. In the 25—34 age class the elasticity pattern also is maintained between the self—employed and semi-skilled groups. No iden— tifiable elasticity pattern is found in the 35~44 age class. The sequence of elasticity coefficients in the 45—54 age class is maintained except for the families in the semi- skilled category. In the 55—64 age class, the elasticity coefficients increase between the self—employed and skilled groups. However, the sequence is not maintained for the professional, semi~skilled, and unskilled groups. In the 218 65 and over age class the clothing elasticities generally increase between the professional and semi-skilled groups except for the divergence of elasticities in the self-em- ployed and skilled group. Although the general sequence of clothing elastici- ties is partially maintained in each age class, the hypoth— esis cannot be completely accepted for any of the age classes. 4. The magnitude of income elasticity coefficients for total housing_exp§nditures within each agg class is inversely related to the socio—economic occupational rank of the occupations. The data in Table 33 reveal that several irregulari- ties in the elasticity patterns occur in the various age classes. In the under 25 age class, the clerical families demonstrate the lowest elasticity coefficients for total housing expenditures. For the remaining occupational cate— gories in the under 25 age class no identifiable elasticity pattern emerges. In the 25—34 age class the highest elas- ticity coefficients generally appear in the clerical, skilled, and semi—skilled groups. The elasticity coefficients for total housing expenditures decrease between the professional and clerical categories. The unskilled group, however, demonstrate the second highest elasticity coefficients. The professionals in the 45—54 age class have the highest elasticity coefficients. In this age class, no identifiable pattern emerges. The elasticity coefficients increase with 219 decreases in occupational rank in the 55-64 age class be- tween the professional and clerical groups. By contrast, the elasticity coefficients decrease with the lower rank— ing categories. Finally, in the 65 and over age class, the elasticity coefficients increase with decreases in oc- cupational rank between the professional and clerical cate— gories. However, the remaining occupations do not reveal any specific elasticity patterns. In summary, the hypoth- esis cannot be accepted for any occupational—age class. Marginal Income—Expenditure Relationships 1. In each family age classy the marginal propen- sity to consume for total expenditures is inversely related to the socio-economic occu— pational rank of the occupations. On the basis of the data in Table 34, the hypoth- esis related to each family age class cannot be accepted. However, in the under 25 age class the marginal propensity to consume coefficients increase with decreases in occupa— tional rank for all the occupations except the semi—skilled. However, in the 25—34 age classification no identifiable pattern for the marginal propensities to consume emerges. A general pattern does emerge in the 35-44 age class except for the skilled and unskilled groups. In the 45—54 age class the highest marginal propensity to consume coeffici- ent is found in the professional category. The self-employed, 220 by contrast, have the lowest marginal propensity to consume for total expenditures. In the 55-64 age class, a general pattern emerges between the professional and skilled occu— pational categories. However, this pattern is not main— tained for the semi—skilled and unskilled groups. Finally, in the 65 and over age class a general pattern of increas- ing elasticities emerges between the professional and semi— skilled groups. The unskilled families have the second lowest marginal propensity to consume. Conclusions The conclusions are divided into two sections. The first section summarizes the findings presented in Chapter IV. The second section presents and interprets the family economic profiles for the various occupations. The hypotheses set forth to guide the research were formulated on the basis of occupational socio-eco— nomic rank. Two primary correlates of the socio-economic ranking scale are income and education. It appeared both reasonable and logical that generally the higher ranking occupations would spend absolutely more on most of the broad categories of goods and services. By contrast, it was expected that the higher ranking occupations would spend proportionately less of their incomes on the same goods and services. In many cases, the above relationship holds for the occupations. But when the broad occupational 221 categories are disaggregated into chronological age classes, considerable variations are noted in both absolute and rela- tive expenditure patterns. In other words, what holds for the aggregated occupational category (where all the age classes are combined) does not necessarily hold for the disaggregated family economic life stages. The summary of the findings reveals that each age class of an occupa- tion exhibits unique income, expenditure, and savings pat- terns. The summary of the findings reveals the following relationships. Summary of findings 1. Family total expenditures do not follow the socio-economic ranking in every age class. In the 35-44 and 45-54 age classes, however, total expenditures do fol- low the ranking scale. However, in these age classes the differences in total expenditures cannot always be judged to be significantly different. 2. The correspondence between occupational socio- economic rank and total food expenditures holds only for the 45—54 age class; however, the differences in expendi- tures in that age class cannot be judged to be significantly different. 3. The correspondence between occupational socio- economic rank and total housing expenditures holds complete— ly only for the 25-34 and 45—54 age classes. Nevertheless, 222 the absolute differences according to the significance cri- terion cannot be judged to be valid. 4. The relationship between family clothing expen— ditures and occupational socio—economic rank does not hold completely in any age class. Considerable variations occur in both the non-manual and manual occupational categories. 5. The relationship between house furnishings and equipment expenditures and family maximum outlays indicates that all the occupations have the highest expenditures in the primary age classes except for the self—employed group which reaches maximum expenditures in the secondary age classes. It was also revealed that house furnishings and equipment expenditures do not follow the socio-economic ranking scale in any age class. 6. The data related that automobile expenditures do not follow the socio—economic ranking scale. In every age class, a lower ranking occupational category expended absolutely more toward the purchase of automobiles. In the primary age classes, the skilled and unskilled workers expended more on automobiles than either the clerical or self—employed category. In the secondary age classes, the skilled group had higher outlays than did the self—employed. The data also reveal that in each age class in the second— ary life stages, clerical workers spend more than do the self—employed families. It was also found that maximum automobile outlays 223 for automobile expenditures in all occupations were gener- ally found in the secondary age classes. The semi-skilled and self—employed reach maximum expenditures in the primary age classes. 7. The average propensity to consume for total family expenditures does not correspond to the socio-eco— nomic ranking scale in any of the family age classes. The lower average propensity to consume for the self-employed and skilled workers in the 35-44, 45-54, 55—64, and 65 and over age classes especially upsets the ranking scale. The data for the average propensity to consume (all age classes combined) reveal that the self-employed have the lowest income—expenditure ratio. It is also noted that the skilled workers have a lower average propensity to consume than do the clerical families. The data also reveal that for all the occupations, the average propensity to consume for total expenditures decreases from the average of the primary age classes to the average of the secondary age classes. 8. The average propensities to consume for the total of expenditure accounts classified as necessities does not correspond to the socio-economic ranking scale. Significant exceptions to the "necessity“ income-expendi— ture ratios are noted between the skilled and clerical categories in the 35—44, 45-54, 55—64, and 65 and over age classes. In each of these age classes, the skilled workers have lower average propensities to consume than 224 do the clerical workers. 9. Generally the average propensity to consume for total food increases with each successive age class in the primary age classes. And the average propensity to consume generally decreases with each successive age class in the secondary age classes. Exceptions to this pattern are noted for the self-employed, semi-skilled, and unskilled in the primary age classes. In the secondary age classes, exceptions are noted in self-employed, semi- skilled, and unskilled groups. 10. Although the average propensity to consume for food generally increases in the lower ranked occupa- tions in both the primary and secondary age classes, there are several exceptions in many of the occupational-age classes. 11. The proportion of income for clothing expendi- tures increases in the primary age classes for all the oc- cupations except the self—employed, which decreases, then increases, the proportion of income devoted to clothing. In the secondary age classes, the average propensity to consume for clothing generally declines in all age classes. 12. The proportion of income devoted to house fur- nishings and equipment slightly increases with the lower ranking occupations. However, the proportion of income devoted to house furnishings varies only slightly among most of the age classes. 225 13. Disposable income levels reach maximum levels for the professionals, self-employed, and clerical in the secondary age classes. However, disposable income levels for the skilled, semi-skilled, and unskilled workers reach maximum levels in the secondary rather than the primary age classes. 14. Family savings levels for the professional, self-employed, and clerical workers generally do not fol— low the socio—economic ranking scale. Only in the under 25 age class does the relationship between socio-economic rank and family saving levels hold. In the remaining age classes the self—employed have the highest saving levels. 15. In all but the 55-64 age class, the saving levels for the skilled, semi—skilled, and unskilled fami- lies follow the occupational ranking scale. 16. The income—total expenditure elasticities do not follow the socio—economic ranking scale completely in any family age class. However, in many of the age classes the elasticity coefficients increase with decreases in oc— cupational rank. 17. There appears to be little relation between occupational rank and the income elasticities for total food expenditures. In the under 25 age class for profes- sionals, the elasticity coefficients indicate that total food expenditures are an inferior good. 18. The elasticity coefficients for family clothing 226 expenditures become more elastic, then less elastic, with decreases in occupational rank. The highest elasticity coefficients generally were found for the clerical and skilled families. 19. The elasticity coefficients for total housing expenditures indicate that for the professional families housing expenditures are sensitive to changes in income in the 25—34, 35-44, and 45-54 age classes. Housing ex- penditures by the self—employed families are generally in— come inelastic. The clerical families have elastic expen- ditures for total housing expenditures in the 55-64 age class. By contrast, the skilled workers experience elas— tic total housing expenditures in the under 25, 25-34, and 55-64 age class. The semi—skilled families do not experi- ence elastic total housing expenditures in any age class. The unskilled families, however, do have elastic expendi— tures in the under 25 age class. 20. The highest marginal propensities to consume for total expenditures generally are found in primary age classes for each occupational group. Furtheru the highest marginal propensities to consume generally are found in clerical, skilled, and semi—skilled groups. In the following section several possible determin— ants of occupational expenditure patterns are discussed. By examining some of the influences on expenditure patterns a basis for interpreting the findings and the economic 227 profiles can be provided. The interpretation also may aid in revealing some of the most significant variables respon— sible for the unique expenditure patterns found in the dif— ferent occupational groups. Interpxetation and Analysis of Selected Occupational Economic Profiles Probable Determinants of Occupational Expenditure—Saving Patterns In the following sections interpretive analyses of each occupation's economic profiles are presented. In- cluded in the analyses will be a comparison of specific expenditures by each occupation; a comparison of occupa- tional saving profiles; and an analysis of each occupation's disposable income and specific expenditures. Several ex— planations, hypotheses, and conjectures can be given for the differences in the economic profiles of each occupa- tion. Possible reasons for differences in the economic profiles are explained below under the following categories: 1. Income expectations by various occupational groups. 2. Differences in perception of income security. 3. Differences in family formation. 4. Occupational social and status requirements. 5. Differences in perception of savings and wealth. Income expectations by various occupational groups. Most families, regardless of their occupational group, 228 usually have some perception of what their lifetime earn— ings will be. The perception may be based on other fami- lies' earnings in that occupation, historical and projected trends, or some other prediction base. Illustratively, pro— fessional and self-employed families usually expect rela- tively high lifetime income paths. But, the clerical, skilled, semi—skilled, and unskilled groups expect a more uniform lifetime income path. That is, the lower ranked occupations can identify various constraints on their in— come patterns. Differences in perception of income security. In addition to the occupations having some expectations of their income patterns, families in different occupational group also have ideas regarding the security of their in— comes. In many cases, professionals perceive that their incomes are relatively secure. Security of the profession- al's income takes two forms: (1) the availability of al- ternative employment opportunities, and (2) the income re— wards as reflected in the alternative opportunities offered by their careers. Regarding the latter point, the profes— sional usually has more opportunities to "control" earnings than some of the lower ranked occupations. The self—em— ployed probably perceives the security of his income much differently than the other occupational groups. The self— employed may perceive his income path as being less secure than other occupations regardless whether he has higher or 229 lower income levels than other families. His earnings usu- ally are directly related to various environmental condi— tions and constraints as well as his own initiative. Fur- ther, the self-employed frequently experiences erratic in— come patterns; however, the expenditure patterns usually remain relatively stable. The stability of the self-em— ployed's expenditures may be explained by the Permanent Income Hypothesis. That is, families spend their incomes on the basis of their perception of what their permanent income will be. The clerical and skilled workers likely view their incomes as being relatively secure. In the case of the clerical and skilled workers, employment, union, and other factors influence the perception of income secur— ity, such as guaranteed annual wages, contractual arrange— ments, and possibly minimum wage laws. The semi-skilled and unskilled workers are likely to view their income paths as being less secure than other families in other occupa- tional groups. Differences in family formation. Another variable affecting the economic profiles of the different occupa- tional families are the differences in family formations. The professionals, for example, generally start their fam— ilies later than families in other occupations. The family formation can be delayed by attaining the educational re- quirements and training necessary to enter a specific pro- fession. By contrast, the lower ranked families normally 230 marry at a younger age. Consequently, for the lower rank- ing occupations, the family formation period begins at a younger family life stage. Occupational social and status requirements of the occupation. Expenditure patterns also are influenced by the pressures on the family for status. In the status case, certain expenditures frequently are perceived as giving the family status (automobiles, household appliances, clothing, housing, and house furnishings). Professional families may also believe that there are certain status requirements for their occupational level. The status re- quirements usually are in the form of durable goods. Perception of savings and wealth. Each occupational group's perception of savings and wealth also affects the expenditure patterns of the family. For some occupational groups savings are perceived as being a "buffer" between income and consumption. The "buffer concept" appears to hold for the lower ranked occupations and the younger fam— ilies in most occupational categories. Savings may also be perceived as being the basis for further additions to income. The latter case would normally be true in the case of investments and real property. Generally, the family age variable also affects the saving patterns of the family. In the higher ranked occupations, such as the professionals, the young family 231 may not place as much emphasis on savings as some of the lower ranking occupations since they expect higher future earnings. In the middle family age classes, there may also be a decrease in family savings due to other commitments, such as college education for the children or a new home. Finally, family savings may increase in the older age clas- ses as the family plans and prepares for retirement. Mention should also be made regarding the percep- tion of savings by the self-employed families. Usually in an examination of the saving profiles of self-employed families several erratic patterns exist. For example, if successful in business, the self—employed's financial status can change from an "entrepreneurial" base to an "investment- entrepreneurial" base. The self—employed family, for ex— ample, in the younger age classes may derive the majority of his earnings in the form of income. However, as the self-employed continues to develop the business he not only draws income from the business, but he may also be invest— ing some of the income into his enterprise. The investing of income into the business can change the "entrepreneurial base" into the "investment—entrepreneurial" base. And as the self—employed family progresses over the life cycle income may increase as well as the self—employed's invest— ment in the enterprise. In examining the economic profiles of each occupa- tional group in the following section, several relationships 232 are explored. First, the disposable income and selected expenditure relationships of each occupation will be com— pared. In the comparison, the relationship between the disposable income levels in each age class and total fam_ ily expenditures, total food expenditures, automobile ex- penditures, and total housing expenditures will be analyzed. Second, the family savings profiles for the manual and non— manual occupation categories will be presented. Following the presentation of the family saving patterns, a graphical comparison relating each occupation's expenditures for spe— cific expenditure accounts classified by age class will be presented. Interpretive Analysis of Selected Income—Expenditure Profiles Dispgsable income profiles. In Figure 7, the fam— ily disposable income profiles for the various occupational categories present several diverse patterns. The most di- verse profiles appear in the professional and self-employed categories. The professional profile shows a steady in— crease in disposable income until the 45—54 age class. Disposable income levels decline in the 55-64 age class and again increase in the 65 and over age class. From the disposable profiles it appears that the aggregated group of professionals in the under 25 age class do not discount the present value of their incomes for two reasons: first, families discounting present income for 233 $12,500 F" 12,000— - /\ 10,0008- // 9,000 g . / U C H ~— (1) g 8,000_; O 8 L... 0/..— Q, U) .H o Q‘7,ooo_. :: / /’ AfA—\; // f? / "\P 5,000,. ‘\\\' Professionals Self—employed o — Clerical 4!000[- ——-o. Skilled A, A -— Semi-skilled a Unskilled I 1 I l g L J Under 25 25-34 35-44 445—54 55-64 65 and Over Age Class Fig. 7. Family disposable income profiles by occupational category and family age class. 234 potentially higher incomes must make less income than could normally be expected. In other words, in discounting pres- ent income, the household head presently is foregoing alter- natively higher incomes for future earning capacities. The aggregate data does not confirm this since the average incomes in the other occupational categories are substan— tially lower than in the professional category in the under 25 age classes. However, it is reasonable to assume that families in specific professional groups do discount their present incomes. Second, the data indicate that on the average, professionals do not dissave. The apparent de- crease in incomes from the 55-64 age class for professionals may be explained in part by households retiring. Probably the increase in incomes from the 55—64 to the 65 and over age class is the result of highly "skilled" professionals who continue in the labor market. This may be the result of personal or financial incentives. As a result, the aver- age income figures are higher than for the 55—64 age class. It should also be noted that all the other occupational groups experience declines in incomes from the 55-64 to the 65 and over age class. The self~employed families experience sharp increases in disposable income between the under 25 and 25-34 age class. However, their income decreases between the 25—34 and 35—44 age class. Income increases steadily until the 65 and over age class where a decline in income occurs. 235 The other occupational groups, clerical, skilled, semi—skilled, and unskilled, generally exhibit parabolic income profiles. All of these occupations reach maximum expenditures in the 45—54 age class. Further, the range of disposable income does not vary as much as in the self— employed and professional categories. Total family expenditure profiles. As noted in Figure 8, family total expenditures generally follow the socio-economic rank of the occupational category. All the occupations reach maximum total expenditures in the 45—54 age class, except for the semi-skilled families. The semi- skilled reach maximum expenditures in the 35—44 age class. In the under 25 age class, the professionals have lower total expenditures than the self-employed or clerical groups. The lower total expenditures for the professionals in the under 25 age class can be explained in fact by the profes- sionals' preparation for their respective careers, such as law school, medical school, and graduate school. The profiles also reveal that the self—employed decrease expenditures sharply between the 55-64 and 65 and over age classes. The decrease in total expenditures may be the result of a working capital problem in the business endeavors of the self—employed group. In other words, fam- ily income may be employed in maintaining the enterprise of the self-employed group. The professionals, by contrast, increase total expenditures slightly between the 55—64 and $ Total Expenditures 11,000» 10,000»- 9,000.. 8,000 r 7,000”- 6.0008 5,000 236 Professionals _ __ _. Self-employed —° Clerical Skilled A _ Semi—skilled Unskilled i. l . Under 25 Fig. 8. 25-34 35-44 45-54 55-64 Age Class Total family expenditure profiles by occupational category and family age class. '65 and Over 237 65 and over age classes. However, it is noted that the disposable income of professionals also increases. Total food expenditure_profiles. The profiles for total family food expenditures in Figure 9 reveal several diverse patterns. In the under 25 age class, the self- employed have the highest expenditures and are followed by the semi—skilled, professional, and clerical groups. In the 25-34 age class, the self—employed group again have the highest expenditures. The self-employed are followed by the professionals and skilled occupational category. The expenditure patterns follow the socio-economic ranking scale in the 35—44 age class, except for the skilled cate- gory which has higher expenditures than the clerical group. In the 45—54 age class, expenditures follow the ranking scale. Two diverse patterns are noted in the 55—64 age class: the self-employed spend more than the professionals and the skilled group spends more than the clerical group. In the 65 and over age class, the semi—skilled and unskilled families spend more than the clerical or skilled group. Some of the variation in total food expenditures can be interpreted by examining selected occupational socio- economic characteristics. In the under 25 and 25-34 age classes, the self-employed families have larger families than the professionals. Further, the self-employed fami- lies have a higher average number of children under 18 years of age than do the professional group. The skilled families 238 Professionals $3, 000 r' .... ...... .... Self-employed .____ Clerical _ 0 Skilled A— Semi-skilled n Unskilled m 2 , 500 F" m w s 4.) a p s E352, ooo .. m m o o m 7'. ,J l , 500 '- o 8 1 , 000 - my _ 1 i 1 1 L 1. Under 25 25-34 35-44 45—54 55-64 65 and Over Age Class Fig. 9. Total family food expenditure profile by occupational category and family age class. 239 also have larger families in the 35-44 age class than the clerical category. In the 55-64 age class, the self-em— ployed have larger family sizes than do the professionals. For the most part, it appears that expenditures for food are more dependent on family size and the number of child- ren under 18 than on income. The effect of the decreasing family size in the secondary age classes is especially noted in the clerical, skilled, semi-skilled, and unskilled groups. Housing expenditure profiles. The profiles for total housing expenditures in Figure 10 reveal that the professional group increases total housing expenditures between the under 25 and 25-34 age classes. However, in the remaining age classes, expenditures are relatively con— stant. The self-employed group increases expenditures be— tween the under 25 and 25-34 age classes and decreases ex— penditures between the 25-34 and 35-44 age classes. In the remaining age classes, total housing expenditures re- main relatively constant. The clerical expenditures for housing remain relatively constant until the 65 and over age class, then they decrease. Housing expenditures for the skilled group increase between the under 25 and 25—34 age class, then gradually decrease in the remaining life cycle stages. The semi—skilled, by contrast, increase, then decrease expenditures in the family age classes. The unskilled decrease, increase, then decrease housing $2,500 2,000 Total Housing Expenditures 1,000- \\ (C 240 Professionals \\\\' Self—employed Clerical Skilled A _ Semi—skilled ——-n Unskilled l l I g I i Under 25 25-34 35—44 45-54 55—64 65 and Over Age Class b _—— _o Fig. 10. Total family housing expenditure profiles by occupational category and family age class. 241 expenditures. For the most part, it appears that once the initial housing expenses have been made, the clerical, skilled, semi—skilled, and unskilled groups maintain a stable level for total housing expenditures. Clothing expenditure profiles. As noted in Figure 11, the total clothing expenditures for all occupations, except the semi-skilled workers, reach maximum levels in the 45—54 age class. The semi-skilled workers reach max- imum expenditures in the 35—44 age class. Interesting vari- ations in the clothing profiles occur between the self-em— ployed and professional categories. The self-employed have higher expenditures in every age class except in the 65 and over age class. The other occupations generally fol- low the socio-economic ranking scale. Part of the variation in clothing expenditures be- tween the self-employed and professionals can be explained by differences in family size. In every age class, except the 65 and over class, the self-employed have larger fam— ilies. In the 65 and over age class average family size is the same for both the self—employed and professionals. Similarly, the self—employed families have more children under 18 in each age class. Therefore, it appears that average family clothing expenditures are fundamentally de- pendent upon family size, the number of children under 18, and the clothing requirements of each occupational group. 242 r $1,200— 1,100— 1,000— m - 8 s 900— 4..) a I'D — i x 800_ m m C, h E 6 700_ v-l U 600_ o 500- / Professionals " _. — — Self—employed n 400 ._ Clerical 7 _0 Skilled A— Semi-skilled t u Unskilled F '1’ 1 I 1 l 1 l Under 25 25—34 35-44 45—54 55—64 65 and Over Age Class Fig. ll. Family clothing expenditure profiles by occupational category and family age class. 243 House furnishings and equipment expenditurexprofiles. The expenditure profiles for house furnishings and equip- ment in Figure 12 reveal considerable variation among the occupational groups. Generally, higher expenditures are found in the lower ranking occupations. In the under 25 age class, the clerical and unskilled workers spend the most. Furthermore, families in the lower ranking occupa- tions usually spend a higher proportion of their incomes on house furnishings and equipment in the primary age c1as— ses. The professionals initially increase their expendi- tures, decrease, increase, then gradually decrease expen- ditures. The self—employed group reaches maximum expendi— tures in the 55-64 age class. The skilled and clerical groups again maintain high expenditure levels in the 55— 64 age class. All the occupations decrease expenditures in the 65 and over age class. In part, the variations in house furnishings and equipment expenditures can be explained by the hypothesis that the lower ranking occupations generally involve house furnishings and equipment to a greater extent than other expenditures. As has been pointed out in prior research, house furnishings and equipment are perceived as indicators of status and mobility. Automobile expenditure profiles. In many of the age classes in Figure 13, families in the lower ranking occupations spend absolutely and relatively more on House Furnishings and Equipment Expenditures 244 Professionals _ _ __ Self—employed $600(. . Clerical __ 0 Skilled 550... A— Semi-Skilled . u Unskilled 500 - 450 ... a 400'; ‘ /// 350,. 1 300 .— ' n .A 250,; b 200%” ‘ a 150.. 100 _. j/ 50 _ 1 I l l 1 1 Under 25 25-34 35-44 45-54 55-64 65 and Over Age Class Fig. 12. Family house furnishings and equipment expenditure profiles by occupational category and family age class. 245 $1,550- 1. 500..- 1,400 1,300— 1,200"’ 1,100_ 1,000 900* 800- 700.. Automobile Expenditures 600.. SOOr- 400,. Professionals - _ ...... — Self-employed 300 ———-. Clerical Skilled ' A -— Semi-skilled -—-o n Unskilled 200 ) CIT . l 1 l I. L l Under 25 25—34 35-44 45—54 55—64 65 and Over Age Class Fig. 13. Family automobile expenditure profile by occupational category and family age class. 246 automobiles than those in the higher ranking occupational categories. The professional, clerical, skilled, and unm skilled groups reach maximum expenditure levels in the 45_ 54 age class. The self-employed reach maximum expenditures in the 25-34 age class. And the semi—skilled workers reach maximum expenditures in the 35-44 age class. Further, fam- ilies in all the occupations except the professionals de- crease automobile expenditures after the 45—54 age class. The professionals, by contrast, decrease, then increase automobile expenditures. The high relative and absolute expenditures for the lower ranking families also can be explained in terms of symbols of status and mobility. Summary of Income-—Expenditure Profiles In summary, the most diverse disposable income pat- terns are found in the professional and self—employed oc- cupational categories. The erratic income patterns were explained, in part, by the training required to enter the professional category. Further, the possible working cap- ital problems are faced by families in the self-employed occupations in the older age classes. Families in the other occupations tended to exhibit parabolic disposable income patterns. Regarding the total expenditure profiles, fami- lies in most of the occupations reached maximum expenditures in the 45—54 age class. Generally, a decline in total ex— penditures was noted in the 55-64 and 65 and over age class. Variations in family total food expenditures were explained 247 by differences in the family size of the occupational cate- gories. In effect, for most of the occupational groups, it appeared that family size was a more important determin- ant of food expenditures than was income. Total housing expenditures appeared relatively stable in the 25-34, 35— 44, 45—54, and 55—64 age classes for all the occupations. Family expenditures for clothing appear to be similarly influenced by the expenditures for total food. That is, clothing expenditures in the clerical, skilled, semi—skilled and unskilled groups are fundamentally dependent upon fam— ily size and the number of children under 18. Several di— verse profiles were noted in the expenditures for house furnishings and equipment. Families in the lower ranking occupations generally spend a higher proportion of income on house furnishings. Further, in the lower age classes they also spend absolutely more. These variations were explained in terms of the value placed on house furnishings as symbols of status. Similarly, the diverse patterns for automobile expenditures were explained in terms of the value of the automobile to families in the lower ranking occupa- tions as symbols of status and mobility. Analysis and Interpretation of Occupational Saving Profiles Non-manual occupgtional saving profiles. Figure 14 reveals the unique saving profiles (as percentages of disposable income) for the professional, selfwemployed, and 248 wm>o cam mm .amsOammmmoum mnu How mmaamoum mca>mm .va mmmao mm4 emlmm fimlmfi ewtmm .mEOUCa sandmommam mo ommpswoamm we mmaaaEmm amoawmau mam .ommoamEmlmamm d amuaumao mmhoamEmlmamm.l amsoammomowm .mam enlmm mm news: 00. a _ /\ JOH- /A low. lom. [0V . emoouI etqesodsrq go abequeoied 249 clerical categories. The saving profile for the profes— sional category reveals that savings decrease from the un- der 25 to the 25-34 age class, then remain relatively sta- ble to the 55-64 age class, then decrease in the 65 and over age class. The self—employed saving profile increases sharply between the under 25 and 25-34 age classes, then decreases sharply between the 25-34 and 35—44 age classes. A sharp increase is noted between the 35-44 and 45-54 age classes. Finally, a slight decrease occurs between the 45-54 and 55—64 age classes and an increase between the 55—64 and over 65 age classes. The clerical saving profile reveals an initial increase in savings, a relatively sta- ble savings level, and an increase in savings in the 65 and over age class. The family saving patterns for professionals may _be explained by two hypotheses: first, the need for sav- ings has diminished since the number of dependents has de- creased. Second, since average income increases the need for future savings decreases. Families in most of the other occupational groups experience declines in income, making saving more important. The regularity of the saving through- out the professional family's life stages may be explained in terms of the professional's expectations of his lifetime earning path. As previously mentioned, savings can act as a "buf— fer" between income and consumption. For many occupations, 250 incomes vary more than consumption patterns. Therefore, when incomes fall, the families may decrease savings rather than consumption patterns. However, this pattern is not noted in the self—employed or clerical category. In both these occupational groups, disposable income decreases but family savings increase. The increase in family savings in the 65 years and over age class for the self—employed can be explained by the failure to place family income into the enterprise of the self-employed. Also, the demands for many household expenditures have decreased in the 65 and over age class. This explanation may also be true for the clerical group. Manual occupational saving profiles. The saving profiles for the manual occupational categories in Figure 15 reveal several significant relationships. The only man— ual occupational category which shows consistent positive saving levels is the skilled group. The saving profile for skilled workers reveals a sharp increase in savings until the 35—44 age class. Between the 35-44 age class and 55—64 age class, savings remain somewhat stable. Sav- ings decrease slightly between the 55—64 and 65 and over age classes. The semi—skilled saving profile reveals dis- saving in the under 25 age class. Between the 25-34 and 45—54 age classes, saving levels increase. Saving levels then decrease and increase again in the 55—64 and 65 and over age classes. Unskilled families reveal dissaving in .0Eooda 0anmommaU mo 00mua0ou0m >9 m0aaaEmm o0aaaxmss mam .U0aaaxmlafi0m .U0aaaxm 0gp How 001.30..a 0sa>mm -.ma .0am mmMaU 004 n0>o seumm smums ssnmm smumm mm “mean one mm a a fi 4 q a a T.. S S \ e D m. \ a. (o oo. \ \ nu _Q \\ 4\\ o\ D\\ . 1 mo. 4 . D d 4 _ U 0 II . o.lll(\\\ oa . I mafi o emaaaxmde IIIIDII emaaaxmuasmm d U0aaaxm O ON. emoouI etqesodsrq go abeiueoled 252 both the under 25 and 25—34 age classes. Savings increase between the 35—44 and 55-64 age classes. Between the 55- 64 and 65 and over age classes, the saving profile reveals that the percentage of disposable income devoted to savings decreases. In the manual occupational groups it appears that fundamentally savings are a function of income and size of the family. Summary of occupational savingxprofiles. In sum- mary, the saving profiles for the non-manual and manual occupational categories demonstrate several diverse pat- terns. For the professional and self-employed categories there appears to be a low correlation between income and saving levels. Rather, saving levels are influenced by the specific occupation of the family and the perception of the family of what their lifetime earnings will be. As previously explained, the stable pattern of the saving profiles of the professionals reflect the professionals' ex— pectations of their lifetime earning ability. Although the income of the professionals increases in the primary age classes, the percentage of income devoted to savings does not increase. By contrast, the savings levels of the manual oc— cupational categories generally follow their income profiles. When the income of an occupational category is low, savings also are relatively low. 253 Analysis of the Economic Profiles of Each Occupational Category Professional profiles. Figure 16 presents the dis- posable income and selected consumption profiles for pro— fessional families. Income increases sharply between the under 25 and 45-54 age classes. Then a decrease occurs in the 55—64 age class and an increase in the 65 and over age class. Expenditures for total food reach a peak in the 35-44 age class, then remain relatively stable in the 55— 64 and 65 and over age classes. By contrast, professional expenditures for total housing, clothing, and automobiles generally increase and reach a peak in the 45-54 age class. It is significant to note that expenditures for clothing and automobiles increase slightly in the 65 and over age class. Self-employed profiles. The disposable income- expenditures profiles for the self-employed families re- veal a different pattern than for the professional fami— lies. In Figure 17, the disposable income profiles initi— ally increase, decrease, increase, reach a peak, then de- crease. Unlike the professional families, expenditures for total food and total housing reach a peak in the 55— 64 age class. However, clothing expenditures reach a peak in the 45-54 age class. Expenditures for automobiles, on the other hand, reach a peak in the 25-34 age class. In 254 $13,000— 12,000 '" 11,000» 10,000” 9,000" 8,000+- 7 , 000 — Dis posable Income ——-——— ———- Total Food Expenditures . Total Housing Expenditures 6,000” 0 Clothing Expenditures A Automobile Expenditures 5,000" 4,000-— 3,000.. _,,_‘g‘~ /_ ' fl,— . ' N— o. ———4 *- r”’. .a—" *"— ___‘—_—“——-4r—-——— *"* 2,000 A v . /————_— o A 1,000 - A ./ °V ° E,. l L, 1 L 1 1 Under 25 25-34 35-44 45-54 55—64 65 and Over Age Class Fig. 16. Family disposable income and selected consumption profiles, professional families. 255 $13,000,. 12,000" 11,000r- 10,000»- 9,000;— 8,000.. 7,000- Disposable Income __. ___ Total Food Expenditures 6,000-* .___-0 Total Housing Expenditures . 0 Clothing Expenditures 5:000" ‘-—'A~ Automobile Expenditures 4,000 r 3,000 r . , ___..___. 0/ .w/ ——.\ 2,000— . //‘~-—- 4” \ .... _____./ ”...—_o \. o -———«“‘ 1,000”- .——:‘:°// A °\ ° AWA” 5‘ —-\ ;::2§ .A~._“ I I 1 l I J Under 25 25-34 35-44 45—54 55-64 65 and Over Age Class Fig. 17. Family disposable-income and selected consumption profiles, self-employed families. 256 comparison to the professional workers, the self-employed spend much less on automobile expenditures. However, a similarity of expenditures exists for total food, total housing, and clothing. Clericalxprofiles. The income—expenditure profiles in Figure 18 reveal that disposable income for the clerical families increase until the 45—54 age class, then decrease in the 55-64 and the 65 and over age classes. Expenditures for total food, clothing, and automobiles reach a peak in the 45—54 age class. Total housing expenditures, by con— trast, reach a peak in the 25-34 age class. However, in all the age classes, total housing expenditures remain rel- atively constant, except in the 65 and over age class. Skilled_profiles. The disposable income profile presented in Figure 19 illustrates that maximum income levels for skilled families are not reached until the 45- 54 age class. The profile also reveals that the incomes of skilled families do not vary absolutely or relatively as much as for families in the professional or self-employed categories. The profiles also reveal that expenditures for total food, total housing, clothing, and automobiles also remain relatively stable between the 25—34 and 55-64 age classes. Unlike those in the professional and self- employed categories, the skilled families reach maximum total housing expenditures in the 25—34 age class. This $9,000 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 257 r— _ Disposable Income ——-——— ———Total Food Expenditures . Total Housing Expenditures 0 Clothing Expenditures A» Automobile Expenditures H- .——-—’—'-—"\o ._____——-..—O ‘ X/A\ . ° ,1. .____.——-O-—-—-0 ...—02‘ WA‘HAM o O\o\ 1 1* i, 1 1 1 Under 25 25—34 35-44 45-54 55-64 65 and Over Age Class Fig. 18. Family disposable income and selected consumption profiles, clerical families. 258 $8,000- 7,000— 6,000— 5,000_ Disposable Income -—-——— ———-Total Food Expenditures .____.. Total Housing Expenditures 0 Clothing Expenditures A"""Automobile Expenditures 4,000- 3.000 2,000— /\ ' ~\‘>~‘::;?::.::222477-'——. / / f, / \\ ",,/ ‘::: 1,000- #A’A‘A—fl/fiAA / A\. ./o O\. 11 1 1 1 I 11 Under 25 25—34 35-44 45-54 55—64 65 and Over Age Class Fig. 19. Family disposable income and selected consumption profiles, skilled families. 259 may indicate that the skilled workers purchase a house and remain in the original house. Further, it may also indi— cate that maximum outlays for house furnishings and equip- ment are made in the 25—34 age class. Semi—skilled profiles. The profiles presented in Figure 20 indicate that maximum disposable income levels for semi-skilled families are not reached until the 55-64 age class. And as found in the skilled income profiles, there is a relatively small variation in income levels be— .‘flnfl’JB'DW'A" '. ': tween the 25-34 and 55-64 age classes. Maximum total food In.“ - ‘H expenditures are found in the 45—54 age class. But maxi— ‘ V mum total housing, clothing, and automobile expenditures are found in the 35-44 age class. And like the relative income range, the expenditures for total food, total hous— ing, clothing, and automobile expenditures are found between the 25-34 and 55—64 age classes. Unskilled profiles. In Figure 21 the disposable income profiles for the unskilled families illustrate that the income levels do not vary as much as some of the other occupational groups. Specifically, disposable income lev- els increase until the 45-54 age class, then gradually de— crease in the other occupational categories. Expenditures for total housing reach maximum levels in the 35-44 age class. But expenditures for total food, clothing, and automobiles do not reach maximum expenditures until the $7,000 6,000 5,000 4,000 3,000 2,000 1,000 1— Disposable Income __ _______ Total Food Expenditures . Total Housing Expenditures * 0 Clothing Expenditures a, Automobile Expenditures *— ‘W ’.\\ __/" 0 / \. \‘x ad A,__.___—o———A—¢\ /o/‘o_~9—§_\ ....r”” oér——‘ \. 1 L, 1 1 1 Under 25 25—34 35-44 45-54 55—64 65 and Over Age Class Family disposable income and selected consumption profiles, semi-skilled families. $7,000— 6,000- 5,0000 _ Dis osabl Income 4,000 Totgl FOO Expenditures Total Housing Expenditures Clothing Expenditures O A. Automobile Expenditures 3,000— 2,000— .\ ' ./.\. V *- \K. .,:0’ ”III 7:3><85T—_9—4 ./ / \0 1,000; :fie/ NO N A\. 1 1 1 I l 1 Under 25 25—34 - 35-44 45-54 55-64 65 and Over Age Class Fig. 21. Family disposable income and selected consumption profiles, unskilled families. 262 45-54 age class. Again it is significant to note the rela— tively stable level of expenditures between the 25-34 and 55—64 age classes. Summary of specific occupational profiles. To sum— marize the specific economic profiles for each occupation, it appears that total expenditures, total housing, and clothing generally follow the income profiles. That is, 1 families in each occupation relatively spend a similar 0 amount for these items. Expenditures for automobiles, by contrast, do not follow the same relative relationship to ‘IPMA\. .. I _ I the income profiles. In the automobile expenditure case, families in the manual occupational categories spend rela— tively more than the non—manual groups. Finally, it is important to note the erratic disposable income patterns, but relatively stable consumption patterns, of the self- employed families. The amount of income devoted to total food, housing, clothing, and automobiles does not vary greatly over the life stages of the self-employed group. This stable pattern tends to follow the tenets of the Per— manent Income Hypothesis. Comparison of the Occupational—Age Approach with the Social Classy Family Life Cycle, and Disposable Income Prediction Methods This section examines and interprets the influence of age and occupation on family expenditure patterns. More— over, since prior research has indicated that the occupation 263 variable is correlated with social class and that chrono— logical age is correlated with the family life cycle, an examination of this congruency will be made. This section is divided into three parts. The first part examines the congruency of occupational expenditures with social class expenditure patterns. The congruency of life cycle anal- ysis and chronological age classes are examined in the sec— ond section. The third part presents an evaluation of us- ing disposable income as a predictor of specific expendi— tures. Congruency of Occupation and Social Class Expenditure Patterns Due to the general lack of intensive empirical re— search on social class expenditure patterns it is somewhat difficult to make strict comparisons between the influence of social class and occupational category on consumption patterns. Nevertheless, a few observations can be made regarding the congruency of social class and occupational expenditure patterns. To make the comparisons, however, it should be noted that only the aggregated occupational category can be employed since research is lacking on social class expenditure patterns disaggregated by age class. Most of the research on social class expenditure patterns has illustrated differences in durable goods ex- penditure patterns, such as houses, household appliances, house furnishings, and automobiles. According to the 264 findings of Martineau, Warner, and Rainwater, notable dif— ferences occur in the expenditure patterns for these goods. Specifically, the findings indicate that the lower social classes usually spend more proportionately and sometimes more in absolute dollars than the upper social classes. However, empirical research on social class expenditure patterns for other categories of goods is lacking. Warner 11-"- ly,: found, nevertheless, that the proportion of expenditures for food, clothing and shelter expenditures increases from —. haw-1k the upper to the lower social classes. By comparing the ...—.. mug...- ". s a.“ prior research findings of expenditures by social class and the expenditures of the aggregated occupational cate- goriesl used in this study, several similarities in con- sumption patterns are noted. The higher ranking occupational categories and social classes have similarities in socio-economic char— acteristics. Illustratively, the upper social classes and higher ranking occupational categories have higher educa- tional attainments, higher incomes, and greater prestige than the lower ranking social classes and occupational categories.2 Moreover, the higher ranking social classes lAggregated occupational category implies that all the age classes are combined so that an average for the occupational group can be determined. 2Although occupational category usually is a pri- mary determinant of social class placement, occupational category and social class are not synonymous. 265 and the higher ranking occupational categories both demon— strate higher total expenditures than the lower ranked so- cial classes and occupational categories. But since both social class placement and occupational category are re- lated to the income of the family head, it appears that in both social class and occupational expenditure patterns total family expenditures are closely correlated with in- come. Research on social class expenditures and the find— . F ings of this study on occupational expenditure patterns indicates that considerable variation does occur in the s. : 5 .. .14— “_ dollars expended on specific expenditure accounts. As in— dicated above, this variation has usually been most pro- nounced in expenditures for durable goods. A safe premise, then, is that other socio-economic and socio—psychological factors also affect expenditure patterns. Consequently, income alone as a predictor of expenditure patterns may not be very reliable as a predictor of many family expendi— tures. Although expenditures for durables generally do not follow the social class ranking, expenditures for “necessities“ generally correspond to the social class and the occupational ranking scale. In the case of nec- essities, income also appears to be generally accurate in predicting expenditure patterns. In many respects, the hierarchy of social class and occupational classification appears quite similar; how— ever, there is considerable expenditure variation within 266 each occupational class. That is, when an occupational category is disaggregated into chronological age classes, considerable expenditure variations also are noted in the various life stages. These variations may cause in— accuracies in predicting expenditures. Consider the fol- lowing example illustrating the amount of variation within an occupational category. In the example, skilled workers ‘9‘- p.-~._.~tu. spend $347 for house furnishings and equipment when all the age classes are combined.1 Table 35 illustrates the i __l _.P __.-a’ -— variability in expenditures by age classes from the mean expenditures for the aggregated occupation (all age classes ‘1 combined).2 TABLE 35 VARIABILITY IN HOUSE FURNISHINGS AND EQUIPMENT EXPENDITURES BY AGE CLASS, SKILLED WORKERS Dollar Variation Percent Variation Occupational from Mean from Mean Age Classes Expenditures Expenditures Under 25 —$ 39 -11% 25—34 + 107 +31 35-44 - l7 - 5 45—54 — 37 —11 55—64 + 11 + 3 65 and over - 155 -45 Source: Table 21. 1Table 21. 2The variability in expenditures in Table 35 may also be due to statistical and data handling errors. 267 As illustrated in Table 35, there are significant variations in family expenditures within an occupational category. Similar variations also are likely to prevail in a social class stratum if the stratum is not disaggre- gated into age classes. However, such research on the within—stratum expenditure variation has been limited. In summary, both aggregate social class and occu- pational category demonstrate similarities in both socio- economic characteristics and expenditures. Research on expenditures by social class and occupational category also has indicated that income is not always the primary deter— minant of expenditures. Furthermore, the expenditure pat- terns of the disaggregated occupational classes demonstrate considerable variability. Congruency of Chronological Age Anaiysis to Life Cycle Analysis In recent years much research also has been con— ducted on family life cycle expenditure patterns and from this research several generalizations can be stated: 1. Income generally increases in each of the early life cycle stages; then, after reaching a peak, income de- clines. Frequently, the income peaks are reached before the household head reaches forty-five years of age. 2. The lowest incomes generally are found in the youngest and oldest life cycle stages. 3. Indications of general increases in family size 268 are given in most family life cycle analyses. The pattern is for family size to increase, then decrease. 4. The influence of all family earners generally is considered in estimating total family income. The num— ber of family earners generally decreases with increases in the life cycle stage of the family. 5. Expenditures for durable goods by stage in the family life cycle generally increase sharply between the first and second stage, then gradually decrease over the remainder of the life cycle. 6. Family asset—liability patterns also increase until a peak is reached (usually the 45-54 or 55-64 age class), then gradually decrease with increases in family age. In most cases the above generalizations also apply to chronological age analysis. In chronological analysis, the age of the family head is the independent variable. And when the analysis is employed, as in the case of the Bureau of Labor Statistics' research, changes in several socio-economic variables are reflected by increases in the age of the family head. For example, income generally in- creases and after reaching a peak, declines. As in life cycle research, the income peaks are usually found in the "middle age classes." In most instances, expenditures for durables initially increase, level off, then decrease. Family size and the number of family earners also increase, .i___. _... .——-.-—-—-—-—’.u‘..0npu \. . ——. 269 then decrease approximately the same as found in family life cycle analysis. Finally, a similar family asset-lia— bility pattern to that found in prior research on family life cycles prevails. An over-aggregation problem, however, frequently exists in family life cycle research. Most of the research on family life cycles has dealt only with an aggregation of consumers, such as all occupational categories combined. In the aggregation process the variability of the different i- _ #2.“.-———-nnru;u—< ... 0 ~ .‘ and unique socio—economic groups usually has been ignored. Variability in aggregated data can be illustrated by employ— y ing expenditure data from the present research. In Table 36, the average expenditures for house furnishings and equipment for the aggregate of all occupations will be compared to the average house furnishings and equipment expenditures for skilled families. In the analysis, both the absolute and percentage variability in the expenditures are illustrated. TABLE 36 VARIABILITY IN HOUSE FURNISHINGS AND EQUIPMENT EXPENDITURES BETWEEN THE AVERAGE FOR ALL OCCUPATIONS AND SKILLED FAMILIES Average Expen- Average Expen— Dollar Percent Age ditures All ditures Skilled Varia- Varia- Classes Occupations Families bility bility, Under 25 $406 $308 $-98 -24% 25—34 387 454 +67 +17 35-44 373 330 -43 —12 45-54 331 310 -21 — 6 55-64 333 358 +25 + 8 65+ 211 192 -19 — 9 lUnweighted averages. Source: Table 21. 270 As indicated by the data, considerable variations occur among the aggregated averages of all occupations and for the specific occupational category. In many cases var— iations in specific life cycle stages or age classes should be considered a weakness in employing only aggregated life cycle data. A similar weakness can also apply when the researcher uses chronological age classes which have not been disaggregated by "homogeneous" socio-economic charac- teristics. Evaluation of Occupational Income as a 1C Predictor of §pecific Expenditures y Although the findings of this research reveal that each aggregated occupation (all age classes combined) ap- pears to be a reliable indicator of total family expendi— tures, nevertheless, considerable variability in specific expenditure accounts occurs when each occupational group is disaggregated into specific chronological age classes. Since the variability of expenditures is present, certain limitations are placed on using occupational income to es- timate expenditures for specific family expenditures within the various family life stages. In other words, the ex- penditure patterns for the aggregated occupation (all age classes combined) do not necessarily hold for the specific occupational age classes. Specific age class variability can be illustrated by examining the income and expenditure patterns of all 271 occupational categories within a particular age class. Table 37 illustrates the variability of family expenditures. TABLE 37 ABSOLUTE RANKING OF DISPOSABLE INCOME AND SELECTED EXPENDITURES FOR ALL OCCUPATIONS, 25-34 AGE CLASS Expenditure Profes- Self- Cler- Semi- Un— Accounts sionals Employed ical Skilled Skilled Skilled Disposable A income 2 l 3 4 5 6 Total food 2 l 5 3 4 6 ‘ Food—away 2 l 3 6 4 5 1 Alcohol 2 4 6 l 5 3 ‘ Tobacco 6 4 5 2 l 3 7 House ;1 furnishings l 4 3 2 5 6 3 Clothing 2 l 3 4 5 6 Recreation 2 l 4 3 5 6 Automobile l 4 6 2 3 5 Personal care 2 l 3 4 5 6 Source: Ranking for disposable income data from Table 28; total food, Table 18; food—away, Appendix C; alcohol, Appendix C; tobacco, Appendix C; house furnishings, Table 21; clothing, Table 20; recreation, Appendix C; auto- mobile, Table 22; and personal care expenditures, Appendix C. In Table 37, the data reveal that in the 25-34 age class only two categories of goods, clothing, and personal care completely correspond to the ranking of disposable income in all the occupational groups. However, the cor— respondence of the occupations' income and specific expen- ditures is not as uncorrelated as the data indicate. The correspondence of occupational—income rank and specific expenditures can be examined more closely in Table 38. In the table, occupational rank and expenditure ranking 272 are compared to determine the differences in the ranking of expenditures by the various occupational categories. Table 38 relates that there is some correlation between occupational—income rank and expenditures; however, complete correlation is absent. Therefore, it can be stated that income data when used alone must be interpreted with some reservations. In Table 38, the number of times each occupational category ranks first, second, third, fourth, fifth, or sixth in expenditures is indicated. In effect, 1 Table 38 is an extension and summary of Table 37. ‘4 .‘~ ‘.‘. ‘5‘- .. TABLE 38 SUMMARY OF DISPOSABLE INCOME-SELECTED EXPENDITURE RANKINGS BY OCCUPATIONAL RANK, 25-34 AGE CLASS Occupations Ranking of selected expenditure categories ranked accord- Rank Rank Rank Rank Rank Rank ing to dispos— l 2 3 4 5 6 able income Self-employed 5 0 O 4 0 0 Professional 2 6 0 O 0 1 Clerical 0 0 4 l 2 2 Skilled l 3 2 2 0 l Semi-skilled l 0 l 2 5 0 Unskilled 0 O 2 O 2 5 Source: Table 37. In Table 38, for example, the self—employed cate- gory ranks first in amount of disposable income. Although the self—employed have the highest disposable incomes, they do not always demonstrate the highest expenditures for each expenditure account. Of the nine expenditure categories, 273 the self-employed have the highest outlays in five of the accounts and they rank fourth in the remaining expenditures. Thus, disposable income rankings may not be an adequate predictor of individual expenditure accounts. Implications of Findings for Market Delineation There are two primary methods by which the occupa- tion-age approach can be utilized in market delineation, purposes are: (l) to aid in locating prime market targets «_A‘Tt-r‘r -—.~ D for the purposes of concentrating the marketing effort; and (2) for studying market dynamics. Market Index Applications The first method utilizes a percentage index and can be employed in two primary ways. First, it can aid in locating average family expenditure concentrations by the occupation—age matrix. Computing the expenditure con- centrations is accomplished by converting the average fam— ily expenditures to index numbers and then delineating the desired markets. A modification of the first method is to introduce the number of families in a given market cell. The mean family expenditures are then multiplied by the number of families in the occupation-age cell. The result— ing product estimates the potential absolute purchasing power for the occupation-age cell. If the researcher de- sires to compare the aggregate potential purchasing power of each occupational-age cell, he can then convert each 274 cell into an index of purchasing power. If the necessary data are available, both methods can be computed for a city, Standard Metropolitan Statistical Area, region, or for the national market. However, in computing the market cell estimates for a Standard Metropolitan Statistical Area, it must be made clear that the projections represent only potential expenditures. When the purchasing power poten- tials are computed for the desired occupational-age classes the results may be "validated" with sales estimates from the Department of Commerce. It is likely, however, that the market potential expenditures will be lower than the estimates of expenditures by the Department of Commerce. The disparity in estimates can be explained in two ways. First, most Standard Metropolitan Statistical Areas exert "drawing power" over consumers living outside the Standard Metropolitan Area. Hence, the "actual" estimates of the Department of Commerce are likely to be larger. Second, only about ninety to ninety—five percent of the population can be declared to be in a specific occupational category. Nevertheless, these "unclassified" consumers spend incomes for goods and services.1 The aggregated potential method can be employed usefully as a benchmark for the marketing promotional effort, 1The problem of the unclassified consumers can be reconciled partially by establishing one or more "unclassi- fied“ consumer groups. In many cases, it is possible to further disaggregate these consumers into retired and un- employed groups. 275 especially advertising, for a given prime market target. For example, if the computations for household durable goods reveal that expenditures are concentrated in the skilled and semi-skilled families where the household head is under 35 years of age, then the appropriate media can be directed toward the desired market concentration. The first approach of converting family expenditures into the cross—sectional indexes is illustrated in Table 39. The ~._._’-__.—__.+.__.;_ method has the advantage of allowing the researcher to com— -.15-w pare expenditures on a relative basis. However, it also 1 provides for the simultaneous comparison of expenditures 1! by several different approaches. In Table 39 three differ— ent cross-sectional indexes are computed and compared for house furnishings and equipment expenditures. The cross-sectional expenditure indexes facilitate the comparison of average expenditure differences among family age classes of an occupation, among identical family age classes of different occupations, and for expenditures for all age groups in all the occupations. The first step in constructing the index is determining average expendi- tures for each occupational category. The average expendi— tures for all age classes within an occupational category are given a base index value of 100. From the base index an index for each age class of each occupation is constructed for the specific expenditure. This is accomplished by di- viding the specific expenditures in each age class by the 276 TABLE 39 PERCENTAGE EXPENDITURE INDEXES FOR HOUSE FURNISHINGS AND EQUIPMENT OUTLAYS Under 65 and 25 25—34 35-44 45—54 55-64 over Professionals Acrossa 97 118 99 101 81 76 Withinb 111 142 121 142 113 167 Allc 126 153 126 131 105 99 Self-employed ., Across 19 91 108 92 113 88 3 Within 17 89 109 104 127 156 F All 20 96 114 96 119 91 Clerical Across 173 122 95 82 97 34 Within 144 107 87 84 99 54 All 164 116 90 78 91 32 1, Skilled {fl Across 95 131 95 89 103 55 ’ Within 76 117 89 94 108 91 All 86 127 92 87 100 54 Semi—skilled Across 144 98 120 94 74 68 Within 112 80 101 90 71 102 All 127 86 106 83 66 61 Unskilled Across 188 83 116 94 92 21 Within 140 65 93 85 83 30 All 158 71 97 79 77 18 Source: Table 21. a . . . Percentage index of expenditures for an occupation based on all age classes of the specific occupation (un- weighted). Index A = expenditures for an occupational age class average expenditures (all age classes) for b the occupation Percentage index of expenditures within a specific age class based on a specific age class for all occupations (unweighted). Index B = expenditures for an occupational age class average expenditures (all age classesIIfor c a specific age class Percentage index of expenditures comparing each oc— cupational age class with average expenditures of all age classes (unweighted). Index C = expenditures for an occupational age class average expenditures for all occupations (all occupations and age classes combined (unweighted) 277 base expenditure level. The comparison of expenditures among the identical age groups of different occupations requires a similar computation, as does the calculation of an overall index of consumption for all the occupational- age classes. In Table 39, the index percentages for house fur— nishings and equipment expenditures are presented by each Intrane- -.-.- . '1 occupational-age class. The percentage expenditure index data in Table 39 . ..-“ .‘J~..l-.A.r_" reveal that in the professional group the highest relative .... ‘2; expenditures are reached in the 25—34 age class. Highest relative expenditures for the self—employed group reach maximum levels in the 55-64 age class. The clerical fam— ilies, by contrast, reach maximum expenditure levels in the under 25 age class. The skilled group, like the pro- fessionals, reach maximum expenditures in the 25—34 age class. The semi-skilled and unskilled groups reach maxi— mum relative expenditures in the under 25 age class. Relative expenditures for house furnishings and equipment within specific age classes of the occupations reveal a diverse pattern. In the under 25 age class the clerical group have the highest relative expenditures. The clerical category is followed by the unskilled, semi- skilled, professionals, skilled, and self-employed, respec- tively. By contrast, in the 25-34 age class, highest rela— tive expenditures are found in the professional category. 278 Following the professional category are the skilled, cler— ical, self—employed, semi-skilled and unskilled. The rank— ing of expenditures in the 35—44 age class is highest for the professionals. Following the professionals are the self—employed, semi-skilled, unskilled, skilled, and cler— ical. In the 45—54 age class, the professionals have the highest relative expenditures. Following the profession- als are the self—employed, skilled, semi-skilled, unskilled, and clerical. The self-employed, in the 55—64 age class, have the highest relative expenditures and are followed by the professionals, skilled, clerical, unskilled, and semi—skilled. In the 65 and over age class, the profession— als have the highest expenditures and are followed by the self—employed, semi—skilled, skilled, clerical, and unskilled. The percentage index of expenditures comparing each occupational-age class to average house furnishings and equipment expenditures reveals that the highest relative expenditures for all occupational-age classes are found in the clerical category in the under 25 age class. The second highest index is found in the unskilled group under 25 years. The third ranking relative expenditure index is found in the professional category in the 25—34 age class. Professionals in the 45-54 age class have the fourth highest relative expenditures. The fifth ranking relative expenditures are found in skilled and semi—skilled age clas— ses in the 25-34 and under 25 age class, respectively. d—m 279 Market Sector Dynamics Another potential application of the occupation— age market delineation approach is to use the method in monitoring socio-economic changes in the market. The so- cial class and occupational structure is changing contin— ually. And as a consequence, the expenditure patterns of various socio—economic groups change. It is important for market researchers to be aware of such changes in direct— ing their marketing programs. By observing the changes occurring in the market over a period of time an estimate of the future market composition can be made. The occupation-age approach offers several poten- tially important advantages when used to monitor changes in markets. First, it allows the market researcher to note fundamental changes in the basic socio-economic groups which comprise the market. Second, the approach allows the mar- ket researcher to be cognizant of socio-economic changes within these consuming groups. Third, the approach aids the market researcher in predicting future changes in con— sumers and market patterns. Depending upon the nature of the data available to the market researcher, the approach can be implemented relatively easily. To operationalize the approach, the researcher needs to know the number of families in each occupational—age class for two or more time periods. The researcher also needs an estimate of the mean family‘ 280 expenditures for each occupational-age class in each time period for the particular expenditure being researched. For example, the problem could be to determine what changes, if any, have occurred in house furnishings and equipment expenditures by skilled families between 1950 and 1960-61. The 1950 market potential base could be estimated by using the U.S. Census of Population (1950) and the expenditure estimates from a source, such as the Wharton—Bureau of Labor Statistics' Study of Consumer Expenditures,_Incomes, and Savings. The 1950 market potential base could then be es- timated by multiplying the number of families in each de— sired age class (estimates from the Census of Population) by the estimates of family expenditures for the age classes (estimates from Study of Consumer Expenditures, Incomes, and Savings). Similarly, estimates for the 1960-61 time period can be computed from estimates of families in the 1960 Census of Population and from estimates of expendi— tures in the Bureau of Labor Statistics' 1960—61 Survey of Consumer Expenditures. Depending upon the available data, the potential expenditure bases for 1950 and 1960-61 can be estimated for cities, Standard Metropolitan Statis- tical Area, regions, or nationally. After computing the market potential base, the desired time periods, then, the absolute and percentage change in expenditures can be de- termined. The market researcher also may want to determine 281 whether or not changes in expenditure patterns have occurred within several age classes over time. For example, the market researcher may desire to know whether the percentage of the purchases for house furnishings and equipment expen- ditures by families under 45 years has changed in a time period. To determine this, it would be necessary to sum the percentages of total house furnishings and equipment expenditures made by families in the age classes under 45 years. The resulting sum when compared to the 1950 base would be used to estimate these changes which have occurred in the expenditures. Finally, the researcher may want to compare the average change in expenditures in a family age class to the average change in expenditures for the aggre- gated occupation. In this case the average change for the occupation would be given an index of 100. Then the per- centage change in the particular age class could be compared on a relative basis with the change in expenditures in the aggregate occupation. Suggested Areas for Further Research This study has attempted to evaluate one approach that can be used to examine consumer expenditure patterns and that potentially can aid in delineating consumer goods markets. As in all of the currently used market delinea— tion methods, several improvements and refinements of this basic approach likely can be made. Several suggestions for additional research are suggested below: 282 First, research should be undertaken to determine as closely as possible the optimum occupational disaggre— gation necessary for purposes of market delineation. A criticism of employing only six occupational categories is that a possible overaggregation problem may occur. By determining as nearly as possible the occupations having both similar socio—demographic and family economic charac— teristics, the delineation process may be improved. In sum, a balance based on the economic feasibility of the method between over and under occupational aggregation should be sought. Second, research should be undertaken to evaluate the use of only four family age classes instead of the six employed in this study. The findings of this study indi- cate that the most erratic expenditure patterns occur in the under 25 and 65 and over age class for many consumer expenditures. By contrast, expenditures follow more closely to the ranking of the occupations in the 25-34, 35—44, 45— 54, and 55—64 age classes. Research employing only four age classes may desire to consider only the under 25, 25- 44, 45-64, and 65 and over age classes. For some research purposes it may also be desirable to ignore the senior age classes for each occupation. Third, a logical extension of the present research is to examine how specific products and brands change in importance to consumers in the different family life stages. 283 Such research could also include the study of brand loyalty by occupational category and age class. Fourth, studies are suggested for determining the possible incongruity of the Permanent Income Hypothesis and aggregated occupational expenditure patterns. Such research would determine whether or not the hypothesis ap— plies to all the specific occupational categories. The findings of this study indicate that the occupational cate- gories do not always have expenditure patterns which cor— respond to the earning power of the families. Fifth, the final area suggested for further inquiry is to investigate the effect of increasing “security” on occupational expenditure patterns. Increasing occupational security may lessen future family savings and, as a conse- quence, certain classes of products may increase in prior- ity in the family budget. .nl'v..1llhl.ll.l11lh1.l.l.l ). 1? APPENDICES APPENDIX A Socio—Economic Characteristics of the Occupational Categories 286 Socio-Economic Characteristics of the Occupational Categories Considerable differences exist among the socio- economic characteristics of the families in the various occupational categories. In this section the findings re— lating to the following family socio-economic variables are discusse 1. 2. 3. 4. 5. 6. 7. 8. d: Average Average Average Average Average Percent family disposable income family size education of family head number of children under 18 number of full-time earners of nonwhite families in the occupa— tional category Percent Percent home owners auto owners The above family socio—economic characteristics are presented according to occupational rank and family age class. TABLE 40 FAMILY SOCIO—ECONOMIC CHARACTERISTICS, PROFESSIONALS SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER- Under 65 and ISTICS Average 25 25-34 35-44 45—54 55-64 Over Average Disposable Income $10124 $6699 $8561 $9880 $11840 $10760 $12359 Family Size 3.6 2.8 3.7 4.1 3.8 2.6 2.6 Education 14.1 13.2 15.2 14.2 13.6 13.2 13.5 Children under 18 1.5 1.1 1.9 2.0 1.2 0.3 0.1 Full-time Earners 1.1 1.0 1.1 1.1 1.2 1.2 1.1 Percent Nonwhite families 3.1 7.7 2.7 2.8 3.6 2.9 0.0 Home owners 68.6 15 4 54.3 79.4 76.6 68.8 58.3 Auto owners 92.2 84 6 99.1 91.7 93.6 82.2 83.4 1““... all .. _ -n—‘O‘ vh- .. 287 Professional Families In Table 40 the family socio—economic characteris— tics of the professional occupational category reveal that the average disposable income level is not reached until the secondary age classes (45—54). Average family size reaches a peak in the 35—44 age class, then decreases through the secondary age classes. It is also noted that average educational attainment remains relatively constant through- out the age classes. The number of children under 18, like average family size, increases in the primary age classes, then decreases in the secondary age classes. A somewhat surprising finding, however, is that the number of full time earners increases slightly in the secondary age clas- ses. The percentage of nonwhite families remains relatively constant in all the age classes except the under 25 cate- gory where a higher than average percentage of nonwhite families appears. The percentage of home owning families greatly increases in the primary age classes and slowly decreases in the secondary age classes. The percentage of auto owners in the professional categories increases substantially between the under 25 and 25-34 age class and remains relatively constant. Self-empioyed Families The self-employed category has somewhat different socio-economic characteristics than the professional cate- gory. The findings of the self—employed occupational 288 category are presented in Table 41. In the self—employed occupational category, maximum average family disposable income is reached in the same age class (45-54) as in the professional category. However, average income declines considerably between the 55-64 and 65 and over age class which contrasts with the professional category. Regarding family size, the self-employed begin their families at a younger life stage than the professionals. It should also be noted that average family size for all the age classes is higher in the self-employed group than in the professional group. Educational attainment is considerably lower in all the age classes. The self-employed family also tends to have more children under 18 in the primary age classes than the professional group. The average number of full time earners in each age class remains relatively constant. The percentage of nonwhite families in the various age classes differs considerably from the professional category. In the professional group, for example, there is a higher per— centage of nonwhite workers in the primary age classes. By contrast, in the self-employed category there is a higher percentage of nonwhite families in the secondary age classes. The self-employed category also prefers to own homes more than the professional category; however, the professional group has a slightly higher percentage of families owning automobiles. ,‘ .in-..- ...-r -._—_ .11— -.L. u— ‘. 289 TABLE 41 FAMILY SOCIO-ECONOMIC CHARACTERISTICS, SELF-EMPLOYED SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER- Under 65 and ISTICS Average 25 25-34 35-44 45-54 55—64 Over Average Disposable income $10232 $4444 $9612 $7892 $11539 $12139 $9724 Family size 3.7 4.0 4.1 4.4 4.0 2.8 2.6 Education 11.4 10.0 12.9 11.6 12.1 10.2 8.8 Children under 18 1.5 2.0 2.2 2.4 1.4 0.4 0.1 Full-time earners 1.0 1.0 1.0 0.9 1.1 1.0 1.0 Percent Nonwhite families 4.1 0.0 0.0 4.7 6.8 5.1 0.0 Home owners 72.6 100.0 58.6 63.1 81.9 82.1 73.3 Auto owners 90.1 100.0 93.0 88.5 93.3 89.8 80.0 Clerical Families The data for the clerical families in Table 42 re— veal that average income for the category is not reached until the secondary age classes. Average income increases to the 45-54 year age class, then decreases in the succeed— ing age classes. The clerical category tends to have smaller families than self-employed or professional categories. Educational attainment of the clerical head is lower than the professional category but higher than the self—employed category. And the average number of full-time earners re— mains constant in the primary age classes but increases in the 45—54 age class. The percentage of nonwhite fami- lies in the clerical category contrasts with the self-em- ployed category. A higher percentage of nonwhite families 290 appears in the primary age classes of the clerical category as contrasted to the self-employed primary age classes. Considerable differences exist in the clerical category regarding home ownership and auto ownership. In the cler- ical category (all age classes combined) only about one— half of the families are home owners. Further, only about three-fourths of the clerical category own automobiles. TABLE 42 FAMILY SOCIO-ECONOMIC CHARACTERISTICS, CLERICAL SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER— Under 65 and ISTICS Average 25 25-34 35-44 45-54 55-64 Over Average Disposable income $7105 $5951 $6729 $6850 $8167 $7288 $6220 Family size 3.3 2.6 3.6 3.9 3.3 2.6 2.0 Education 12.2 12.8 13.7 12.3 11.7 10.8 10.7 Children under 18 1.2 0.6 1.7 2.0 0.9 0.2 0.0 Full—time earners 1.1 1.0 1.0 1.0 1.4 1.2 1.2 Percent Nonwhite families 7.1 11.8 7.4 10.0 5.8 3.9 0.0 Home owners 53.0 11.7 48.3 60.4 58.8 55.8 40.1 Auto owners 74.4 64.7 73.7 72.6 80.9 71.2 79.9 Skilled Families In the manual working occupational categories dif- ferent socio-economic variable relationships were found which differ significantly from the nonmanual occupational categories. Table 43 presents the data for the socio-eco- nomic characteristics of the skilled families. 291 TABLE 43 FAMILY SOCIO-ECONOMIC CHARACTERISTICS, SKILLED SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER- Under 65 and ISTICS Average 25 25-34 35-44 45—54 55-64 Over Average Disposable income $7104 $5358 $6379 $7167 $7688 $7628 $6227 Family size 3.8 3.1 4.5 4.3 3.7 2.9 2.2 Education 10.1 11.5 10.6 10.6 9.8 9.2 7.4 Children , under 18 1.6 1.3 2.5 2.1 1.2 0.4 0.6 . Full-time earners 1.0 0.9 0.9 0.9 1.1 1.1 0.8 Percent Nonwhite 1 families 4.5 4.8 8.2 2.8 2.4 6.0 4.8 ~ Home owners 64.8 14.2 48.8 73.5 71.2 71.4 66.8 Auto owners 87.0 95.3 92.9 90.9 81.7 84.6 66.8 For the skilled category average disposable income (all age classes combined) was approximately the same as for the clerical category. But the skilled workers do not surpass the clerical workers in disposable income until the 35—44 year old age class. The average family size (all age classes combined) of the skilled workers surpasses the clerical, self-employed, and professional groups. Further, the skilled working families reach maximum size at younger life stages than the clerical, self-employed, or profession— al categories. Average educational attainment is consider— ably lower than in the higher ranking occupational groups. Regarding the percentage of nonwhite families, the skilled worker category has fewer than the clerical group but more than either the self—employed or professional group. Home 292 and auto ownership are considerably higher in the skilled group than in the clerical category. As for family home ownership, both the skilled and clerical groups reach the highest percentage of families owning homes in the 35-44 age class. However, the highest percentage of skilled fam- ilies owning automobiles is reached in the under 25 age class, which is higher than most of the other occupational groups. Semi-skilled Families The semi-skilled socio-economic characteristics are presented in Table 44. TABLE 44 FAMILY SOCIO-ECONOMIC CHARACTERISTICS, SEMI-SKILLED SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER- Under 65 and ISTICS Average 25 25-34 35-44 45-54 55-64 Over Average Disposable income $6405 $4543 $5758 $6586 $6818 $6861 $6021 Family size 3.9 3.1 4.2 4.4 3.8 3.0 2.5 Education 9.3 10.7 10.5 9.9 8.7 8.1 _ 5.3 Children under 18 1.6 1.5 2.3 2.2 1.3 0.5 0.3 Full-time earners 0.9 0.3 0.8 0.8 1.1 1.1 1.0 Percent Nonwhite families 18.3 6.7 25.2 19.1 17.6 13.1 7.1 Home owners 48.1 6.7 33.8 54.1 51.0 64.0 49.9 Auto owners 75.9 73.3 72.7 78.4 80.8 73.7 50.0 The disposable income of the semi-skilled families is considerably less than the disposable income of the 293 skilled occupational category. Similar to the skilled cate— gory, the semi-skilled workers reach the disposable income average in the 35-44 year age class, which is similar to the maximum income levels of the nonmanual occupational categories. Of all the occupational categories, the semi— skilled workers have the largest average family size. How- ever, the educational attainment of the semi-skilled workers is lower than that of the other higher ranking occupational groups. The semi-skilled workers also have a much smaller average number of full-time earners in the primary age clas— ses than do the other occupations. The findings also reveal a much higher distribution of nonwhite workers in the semi- skilled category than in any of the other higher ranking categories. Also, the percentage of families owning homes or autos is similar to the clerical category. Approximately one—half of the families own homes and about three-fourths of the families own automobiles. Unskilled Families The findings of the socio-economic characteristics of the unskilled families are presented in Table 45. Average disposable income for all age classes in the unskilled occupational category is considerably less than for the other occupational groups. As in the skilled and semi-skilled occupational categories, the average dis- posable income level is not reached until the 45-54 age class. Similar average family size relationships, as found ._‘. ”... . -.sa. m.---Wr—_-ug >1 ,. .- ‘N-ig" . 1. 294 in the skilled and semi—skilled groups, are also found in the unskilled categories. Average educational attainment for the unskilled workers is slightly less than for semi— skilled workers and less than for the skilled category by one year. The average number of children under 18 is less than in the semi-skilled and skilled category. However, the unskilled group has the highest percentage of nonwhites of any occupational category. As expected, the unskilled workers ranked lowest of all the occupations in the percent of home owners category. Moreover, the lowest percentage of families owning automobiles is found in the unskilled group. TABLE 45 FAMILY SOCIO-ECONOMIC CHARACTERISTICS, UNSKILLED SOCIO- AGE CLASSES ECONOMIC PRIMARY SECONDARY CHARACTER- Under 65 and ISTICS Average 25 25-34 35-44 45-54 55-64 Over Average Disposable income $5513 $4303 $4671 $5682 $6495 $5519 $4798 Family size 3.6 3.2 3.9 4.1 3.7 2.8 2.3 Education 9.0 10.9 10.1 9.8 8.6 6.9 8.4 Children under 18 1.4 1.2 1.9 2.1 1.4 0.4 0.1 Full-time earners 0.9 0.8 0.8 0.9 1.0 1.0 0.9 Percent Nonwhite families 31.2 29.5 34.9 38.9 21.8 27.6 30.0 Home owners 40.9 5.9 27.2 39.0 55.1 49.5 50.0 Auto owners 62.1 52.9 66.7 65.6 69.7 49.5 50.0 APPENDIX B Selected Historical Consumer Budget Studies 296 .mm .o .Ammma .moauuo oeaueaud nemesum>oo .m.o ".u.e .sou0sar003vywms0ru mbao0m 0ca>3m C0Uau0E4 30$ .uon0a mo uC0Ebu0Q0Q .m.D «Eoum .aes .smmuamm .do .>a sued .amewa .coumomo am .02 pdmssuoo uaansm .mnummssummmmz no spasms ICOEEOU .woo0a mo muaumap0bm 050 so unoo0m a0sco4 rpxam 0gp Eouw 00>au0o «mumbom .00usom 0:0 ca 0u0£30ma0 ozonm 0morp 60am mw0aaOp 30m 0 an u0wmap p0caneou m0mm0au 0Eouca aa0 we a0sm ud0UX0 00aa000p00 aa0 mom m0usuaps0dx0 000u0>4 .0oaossou mo 0050009 ma0000 on 000 boo >08 mE0bH "maoz .m0uonm >u0uou0 ca 0000:0Hso madoeeou msoama>oad 0C0 .aao 0s0mou0x .poom m0ooaocam .000000 0008 on 0>au0a0u n0ar 0u03 roan: m0eousa p0>a0u0u mu0xuoz a0uau0a0 0:0 mu0CH00 0003 0u0r3 m0u0um 0:0 0COE0 003 mppmmsnu0mm0z .poaw0d wasp 04a 0.0a o.m o.m o.m o.m o.m m0mc0ox0 hupcom o.m o.m o.m o.m o.m o.m a0sm o.ma o.ma 0.0a m.ma o.om o.ma ps0m o.ma o.ma 0.0a m.oa o.m 0.0a 0ca£uoau o.am o.mm 0.0m o.mm 0.0m o.mm 00s0pmamnsm 0.00a 0.00a 0.00a 0.00a 0.00a 0.00a COau Edmsou uc0ausu wow m0uopaoc0ox0 m0 us0uu0m mmm.am ammm mmmm 00mm mamm mmmm a0800da M0COE a0scc0 000u0>< o.m m.m 0.0 m.m o.m a.m 0nam maae0m 000H0>4 m mma msa mm m mmm m0aaaE0m mo w0QEsz u0>o ou0 oomsam ommm oomm Omem m0mm0au mE0pH 0MSanC0me oom.am M0os3 w0pco w0pc: 000C: 0EOUCH pc0 UHQQ0H0OE0Q pc0 ommw 0C0 comm 0C0 omvm 0:0 comm aa4 mm0aU 0EOUCH , mm4AU MZOUZH Mm m2308 Q24 WMH8HU ma 2H mmmzm4m mw43 Q24 mmHAH24h m0 mzmm884m 20H8QZDmZOU Mm>mbm mmD8HQ2mme M88mmbmu4mm42 mmlvmma m0 mqm48 297 .- 1.....ITWQ. ‘1HV‘I. 1. \nA: a... ard, 11.n- . .sOI .11.. .ama .d .Aooma .euammo mnannanm uc0Edu0>oo u.U.D .QOp0sanm03m mmma Op m0Ea8 a0acoaou .m0b0um U0MHCD 0:0 mo moan Imap0um a0uauoumam .moms0u 0hr mo 300asm .m.D mmm .ammma .0uammo 0saucaum us0ecw0>ow .m.D u.U.Q .cou0canm03v 0mm0ru mpao0m 0sammm C0uau0e4 30$ .HOQ0A mo #C0Euu0d0o .m.D Eoum .aaa sued .Ammma .amma .nononaneeso .une .aeeum .nona "noanunoonm no 0000 .nonea mo u0soammaEEou 0gp mo monon0m a05sc4 nud0>0m pc0 zuxam Eoum U0>au0o ”mumbom .0saosoou wo 0030U0Q ma0uou Op 000 pos >05 mE0uH "maoz a.os m.om N.mm o.ma o.ma o.ma N.sa e.mm meanonnm m.ma m.ea o.ma o.oa m.sa o.sa o.ma s.ma onannoau o.m m.s o.m o.m m.o N.m m.e o.m enoaa one amen o.ma s.sa m.ma o.ma m.ma o.sa o.ma o.ma onamnon o.ma o.mm m.om m.os o.ss m.ms o.ms o.as ooon o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa noandsnenou #CmHHSU MOW. mmuaaau Iaoc0ox0 mo uc0uw0m ems.ao soo.ae memo seem oose mmmo omao meme enouna sends a0SCG0 000w0>4 m.s m.s a.s o.m o.m m.m s.m o.m euae saanem eoeue>< on on. oom mos moa.a oom, sm mom.m eeaaasem mo nensnz ne>o one oomsao ooo.an ooea oooa ooso oomo memeeao esena ennaaonedxm oom.aw u0ocs pc0 M0osz u0oco u0oc3 w0oss M0UCD 0EOUGH 0:0 UH£Q0M0OE0Q ooo.ao, one oooo one ooom one ooso one oome aan mm0aU 0EOUCH mm4dU QSOUZH Mm .mZOmmmm mmoz mo N m0 mmmmeZ A4UHmmAU Q24 Q043 N8HU ho mmHAHZ4m ho mmD8HQ2mQXm 20H8QZQm20U mme8HQ2mQXm mmZQWZOU m0 Nm>mDm amlmmma m0 mam48 298 .os .d .Aemma .euanno 0causaum uc0eeu0>ow .m.D u.U.Q .sou0ca5003v 00C0cu 0pan0m 0GMN5m C0Uau0E4 30m .Hoo0a mo oneneneeeo .0.o "noun .mmm .mmm .aem .os .amooa .nononaneezo uonea no nenoaeeannoo one mo uaoo0m a0sss4 ruc00pr0am .0oom mo 000aum aa000m 0C0 0ca>aa mo #000 "mumbom .0Ca0ssou mo 00:0U0n 0a0uou on 000 no: w0e 0&0pH "maoz .0aQ0aa0>0 poz .00au0so0 0C0 .0ca300au .0oom .0saun0aa .a05m .uc0u woa 00Houm0s0dx0 00: 000 ”us0>u00 no .u000Oa .u00u0oo .pC00c0Q00 0c .0a 000 u0>o 0:00 000 C0w0aaru m 0030 0uoe no: .0ma3 0 .xuoz 00 0209055 0 00: :MaaE0m a0euoe: 0 .>0>ao0 0:» wow 00caa00 04 a o.mm a.mm m.mm m.mm o.mm m.am o.ma m.ma m.ma a.ma m.ma m.ma a.om 00au0som m.ma m.sa a.ma o.sa m.ma m.ma m.ma o.ma o.aa o.oa m.m m.m o.ma 0na£uoau m.a a.a N.a a.a a.a a.a a.a a.a a.a a.a a.a m.a a.a us0aq m.m m.m m.m m.m m.m a.o m.s a.m m.m o.m a.m m.m m.o a03m o.ma m.ma m.ma m.ma a.ma m.ma m.ma o.ma m.ma m.ma o.ma m.ma a.ma us0m o.mm m.mm. m.mm 0.0m s.ao o.ao m.mo m.mo m.mo a.mo m.mo m.om a.mo 000m o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa soauoES0sou pa0uusu mom 00uoua000ox0 mo uC0UM0m Ame “mo Ame Amv “NV Ame ANS “mm “mm Ame Ame Ame ammm msouna >0coe 000u0>4 m.m 0.0 m.o a.o N.o 0.0 m.m m.m m.m m.m o.m m.m 0.0 0na0 >aaE0a 000u0>4 mma mm com omm mom omom mmmm ommm mmma mom maa mm mma.aa 00aaaE0m wo u0nesz ne>o oomao ooaao oooao oomo oooo ooeo oooo oome ooso oomm oomo eeeeeao eneea ends UCM HGUCS HQUCS HQUCS H$UC3 HGUQS HQUCS HOUC3 .HNUCS “QUE: .HmUCaa HQUCD QEOUCH IHUCQQKQ UCM oomam 000 0C0 000 000 000 0C0 000 0:0 000 0:0 aa4 uaro0u0OE0Q ooaaa oooam comm 00mm 00mm comm comm 0000 00mm comm 000au 0EOUCH 00oo .0.o ..u.o .nononaneezo eoneno meanen ddaxmm c00au0e4 30m .uoo0a mo 0:0Euu000o .m.D "Eoum .Aomma .COp0ca0003V mmm sau0aasm 0uau0au0um uon0a mo 300asm .m.D .000000 00pasD 030 ca 0sa>aa m0 #000 «mumbom 299 .0ca0coou mo 0050U0Q 0a0pOD Op 000 poo >05 0&0pH «mBOz m. 0. m. m. m. m. m. m. 0:00c0aa000a2 m. 0. e. o. s. m. m. m. noaneonom e. 0. e. 0. 0. 0. e. 0. onaoeen N.s o.m a.m o.m o.m m.a o.a s.a noaneenoem o.a o.a o.a o.a a.a a.a a.a o.a eneu aenoenen s.s m.s s.s e.s s.s m.s m.s s.s eneu aeuaoez m.m m.m a.a o.m a.a 0.a s.a a.a noaoennoeenene nenno m.m m.m a.a m.a m. o. a. N.a 0aanoeous4 o.am o.ma m.ma s.sa a.oa m.ma m.ma s.sa onannoao m.s e.s e.s s.s s.s m.s m.m o.s onendanoe one eonaneannnn s.a a.m a.m m.m o.m s.a m.m s.a noaneneoo oaoneenon s.s 0.s m.m s.m o.m m.m a.a m.m noaneneoanwen one nnoaa .aenn a.aa o.ma o.sa m.sa o.sa o.sa m.ma m.ma enamnon a.a a.a a.a a.a a.a s.a m.a m.a oouenoe m. m. o. m. m. m. m. m. 0000u0>0o uaaonoua< 0.0m a.om m.em m.mm 0.as o.ss m.os a.os ooon o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa noandsnenou nnennnu wow 00usua0C0dx0 mo us00u0m see.mn aem.mn sam.ao mmo.ao 0mm.a0 oeo.ao oaee mom.ao eeoona xenon eeene>< s.o e.m a.m o.m s.s m.s m.s m.s enae maanea eoene>< mmm mos smm.a ome.m mmm.m mms.m mmm omo.ma eeaaanea no nennnz ne>o oom.mo ooa.mo oo0.a0 oom.ao oowsao oomo eeeeeau meena ennnaoneoxm 0C0 u00c: u00co u00c: u00co 000:5 M000: 0EOUCH 0C0 Ua£QOHOOEOQ oom.m0 one one one one one aan ooa.mm oom.am oom.am oom.am comm 000aU 0EOUCH mm4AU QEOUZH Mm .mmNHm dd4 m0 mQH8HU 2H .QAHEU mzo 8m4md 84 28H3 mmHAHZ4m .mmmxmo3 Q8H23 m0 8QD8W mmD8HQ2mme 20H8QZDWZOU malmama 00 mam48 .ss .0 .xemma .eoanno onannand unen IC90>00 .m.o ".U.Q .9090sa9003c 000090 09a90m 0sa>3m c00a90E4 30$ .90909 90 9908990d0o .m.D "noun .mm .ma .dd .Aasma .nononanmezo 0mm naoeaanm moaneauenm nonea no neennm .0.o .ennao> >90EE3m .mmlomma .0909903 a00a90au 0C0 090C90m 0003 mo 00008009390ao >090: "mumbom .0ca0c3o9 mo 0030009 0a0909 09 000 900 >08 0E09H "mmoz .>aaE0m 900 mm 000090>0 90a93 .>990oo9d a0so090d 0C0 0Eo0ca .aaOd 0903 00903000 00908 a a. o.a m. m. m. s. s. s. o. 0.0 m. enoeneaaeueaz a. s. 0. e. o. s. s. s. m. m. m. noaneunom o.s 0.0 0.0 m.m o.s m.m s.m o.m o.s s.s o.m ouoenon one 0Ca0009 .Coa900900m m.m m.m m.m a.a o.m o.m o.m o.m o.m o.m a.a eneu aenoened m.m m.s m.s m.m m.s a.s o.s m.m o.s m.m o.s eneu aeuaoez e.m o.m m.m m.m s.a m.m m.m s.a o.m s.a o.m noaoennonenenn nonno o.m o.s o.s o.s o.s m.m m.m s.a s.a s.a m.m eaanononnn m.ma o.ma s.ma s.ma s.aa o.aa a.oa s.a o.m s.s o.oa onannan 0 o.m m.m m.s m.s m.s s.s a.s o.m s.m o.m a.s onenoanoe e eonaneannnn 0 o.s o.s s.s s.s m.s o.s e.m m.m o.m a.m o.s noanenedo oaoneenon 3 o.s a.m s.m 0.0 0.0 a.a o.m s.a a.a o.oa s.m noaneneoann 109 0:0 990aa .a03m m.ma m.sa o.ma o.ma o.oa o.sa s.0a m.ma m.om a.om s.sa onaenon o.mm m.mm 0.0m o.mm a.mm m.sm s.mm 0.0m m.em s.mm s.sm aonouae one ooon o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa o.ooa coapdE30000 9909930 9om 00939H0C0QX0 mo 9000900 cmomm mmmmm momma aommm mmmam 0mmam momam mmcam mmma cmmm mam.am a00X09 a000090m 909w0 0EOUCa >0COE a03cs0 00090>4 o.s s.s m.s o.s o.m o.m m.m s.m m.m a.m o.m enae saanew eoene>n amm 00a amm oae 00a.m emm.m sss.m mmm.m mam.a oaa mos.sa neaaanew no nensnz ne>o ooomo ooemo oosmo ooamo oooam oomao oomao oomm oooo memeeao meena ennnaonemxm 000 90003 90093 90093 90093 90093 90093 90093 90093 90093 0EOUCH 0C0 0a9d090OE0o 000mm Dam Cram UCM @990 080 Gram UQM DEM Ucm HH4 comma ccomm ccamm comam comam ccmam comm comm comm 0008U 0Eouca mm4qu QZOUZH Mm amm>O Q24 000.0m mo mmH8HU 2H mmmmeB QQ>OAQEQ ho mmHQHZ4h m0 >QD8m mmD8HQ2mQXm 20H8QZDWZOU mmlvmma Om Q4Q48 -987 1.0.1.... 1 .. *1.11HJ\V .1933. l.\ In... .xemma .euanmo onannann unenn90>oo .0.3 “.0.0 .non 10999003c 009090 099902 099>mm 900990E4 302 .90909 90 990E990d0o .m.D "809m .0E3ao> 9000 na 0a eanen .x .99 .999 .aa .a eesnao> .Aomma .eane>asenned no spanne>an3 .omoa1.0.3 nennn .00a908 a00a90a909m “009a>0m 090 00EOU9H .009399090me 90930900 90 >039m "mumbom 301 .09a093o9 90 0030009 0a0909 09 000 909 >08 0909a ”maoz .x09 >990QO9Q a090090d 090 .x09 aaoo .X09 0Eo09a a00oa 090 09090 .a09000m 0903 00903000 00X08a o.m m.a s.a m.a a.a c.a m.a m.a 0.0 m.a 03o090aa000az m. m. m. m. m. o. m. m. a. 0. 909900309 0. 0. 0. 0. m. m. m. o. 0. m. mnaoeen m.m m.m m.m 0.s o.m m.o m.m o.m o.m m.o 9099009009 o.m m.m a.m m.m m.m o.m m.m m.m a.m m.m 0900 a090090m o.s m.o 0.0 m.o o.m o.m o.m m.m m.m a.m 0900 a00a002 o.m m.m m.a m.a m.a m.a m.a m.a m.a s.a noanenuodenenn nenno m.m m.oa o.oa o.oa m.ma m.aa m.m m.m o.m a.ma 0aa9OEO93< o.ma m.oa m.ma m.ma m.aa m.ca m.ca m.ca m.m m.aa 099990a0 m.m a.m m.m o.m m.m m.m m.m a.m a.m a.m 990EQa300 090 00999099939 m.m m.o o.s 0.0 m.m m.m m.m m.m m.o m.m 909909000 0a0900309 m.m m.m o.m m.m m.m o.s o.m m.m m.m m.o 9oa90900a9m09 0 9909a .a03m o.m m.m m.m m.m m.ca m.ca o.ma m.ma m.ma m.ca 0990303 m.a m.a m.a m.a o.m o.m o.m m.m m.a o.m 0000908 m.m m.m m.m m.a m.a m.a m.a m.a m. m.a 000090>09 09a0900a< m.mm m.mm m.mm m.mm c.cm m.am m.mm m.mm o.mm m.cm 0Com c.cca c.cca c.cca o.ooa c.cc c.cc o.ooa o.ooa o.ooa c.cca 90990530900 9909930 909 009399090QX0 90 9900900 mommam ommmm momma omomm smoom mmomm ommmm mmmam ammm mm0.mm a00X09 a090090d 90990 0E009a >09oE 00090>4 o.s m.s m.m e.m m.m s.a a.m m.m m.m s.m enam 9aane9 eoene>n as sea ems ass mesa ooam mmsa was so eoo.s meaaanen no senenz 90>o ccocam comma cccmm cccmm cocoa ccomm cccmm cccam 00000au 0509a 0939a090mxm 090 90093 90093 90093 90093 90093 90093 90093 90093 08009H 090 0a90090OE0o ccccam 090 090 090 090 090 090 090 aa4 comma cccmm cccmm cocoa cocmm cccmm cccam 000aU 0EOU9H mm4AU QEOUZH 8m .mm>0 Q24 oom.N mo MQH8HU 2H mmHAHE4h mmxm03I44UHmmdU Q24 mmzm4mlmw43 m0 m02H>4m Q24 mmZOUZH .mmmD8HQ2mmxm mmEQmZOU m0 >QQ8W Omma am mqm48 APPENDIX C Other Selected Family Expenditures Appendix C presents other findings of the research. Selected data from Tables 52 to 56 were used in Chapter V in Tables 37 and 38. m.‘ n. -..... .-~_ —. , ... .5- .A. ‘g 303 00a emm omm mmm amm mam mmm mmm mmm oeaaaxeno mea asm mom osm smm mmm mmm amm mmm oeaaaneuasem mma mmm mom omm mmm aam 0mm smm mmm oeaaasm mom mmm mms mmm omm oam mas asm mom aeuaneao mmm 000 was mom mom mmm omm som mss oe>Oadnennaem moss mmma aomm mmma mesa mmma moms mass mesa aenoaeeemond ne>o soumm smnms memmeao ssumm sm1mm mm eeeeeau meemeau xznn one mo eo< neon: 000 004 anzoaendoouo 9neonouem snesand aan 09202000 00090>4 00090>4 00090>4 IOHUOm >M009000m >90Ea9m 00000aU 00¢ mm4AU Q04 Q24 224m UHZOZOUQIOHUOm 44ZOH84QDUUO Mm QEOI 202m N434 mme8HQ2mQXQ QOOQ >AHE4m mm mqm48 304 ll 5,. mmmmmau $04 on mHH moa moa OOH mHH mw moa moa UmHHHMmCD mma moa Ana mwa mmH NHH mm mad mma UMHHHXmIHEmm mm mna mmH owa mfia mNH om vma ova Umaaflxm va vw mma mwa waa mm mm moa NNH Hmoflumau mfia Hoa boa OHH mma waa va nma baa Um>oHQEmImHmm vmam mmam vuaa mmam mmam omam mmm mwam mwam HmCOHmmmmoum um>o volmm vmlmfi mmmmmau vvlmm fimlmm mm mmmmmau mmmmmau xz< mmmum>< mmmum>< IOHUOm Numncouwm ‘NMmEHum mmo vmnmm «mumw mmmmmau wvumm fimnmm mm mmmmmHu mmmmmau xz< mmmum>< mmmum>< IOHUOm Mumvcoumm Mumsflum mmmmmau mm< mmo wonmm «mumw mmmmmfiu ¢¢nmm ¢mumm mm mmmmmau mmmmmau xz4 mmmum>< mmmum>< IOHUOm \Numvcoumm mumeflum mmmmmHU 004 mmo wmnmm @mumv mmmmmau «vumm «mumm mm mmmmmau mmmmmau mz< mmmum>< mmmum>< IOHUOm mumvcoumm [INHMEHHQ mmmmmau wm< mm