AN ANALYSIS OF THE DETERMENANTS‘ 0F ROUTINE PERSONAL SERVICES EMPLOYMENT 4 ‘ Thesis for the Degree Of Ph. D. MICHtGAN STATE UNNERStTY' ' DAVLD I. VERWAY 1968 iHhSIS 0-169 . C .1‘. LIBRA 7‘ "“" Michigan St» University 7 / 3 1293 10283 224 This is to certify that the thesis entitled An Analysis of the Determinants of Routine Personal Service Employment presented by David Verway has been accepted towards fulfillment of the requirements for Ph.D. Economies degree in n ajor professor Date May 17, 1968 ABSTRACT AN ANALYSIS OF THE DETERMINANTS OF ROUTINE PERSONAL SERVICES EMPLOYMENT by David I. Verway Among the disadvantaged occupations is a group associated with routine personal service industries:w those industries which produce services that are easy substitutes for certain processes in the household, or home production. These industries are private households; laundries, laundry services, and cleaning and dyeing plants; beauty shops; barber shops; shoe repair shops, shoeshine parlors, and hat cleaning shops; and pressing, alteration, and garment repair shops. Using two-stage least squares regression on areal cross section data from the 1950 and 1960 Population Censuses with states as units of observation, this research demon- strates that there is a high positive association between the supply of disadvantaged labor and employment in certain of the disadvantaged occupations associated with routine personal services industries. The regression results sug- gest that this association is high for both males and females employed in the private households occupations as a whole and for both males and females employed in the dis- advantaged occupations in the laundry and dry cleaning industries, or as laundry and dry cleaning operatives. David I. Verway Employment as male barbers and male cobblers is highly associated with the supply of male disadvantaged labor. For occupations designated as babysitters, beauticians, and live-in domestics, the results are less clear cut. Supply is only one of the forces considered. The opposite of the coin of the factors determining routine per- sonal services employment is, of course, demand. One very relevant factor in the demand for routine personal services is the opportunity cost of home production. This factor is formal recognition that certain members of the household may find that their labor is worth more in the market than in the home. In other words, the optimum family decision may involve the wife's becoming employed outside of the home, and the employment of a domestic or otherwise having the household chores to release the wife for market labor. The variables selected to represent the opportunity cost aspect in this study are, for females, average weekly income outside of the private households industry as a percentage of average weekly income for the particular routine per- sonal service occupation; and for males, average weekly income as a percentage of average weekly income for the particular routine personal service occupation. The results indicate that opportunity cost is probably a relevant factor in the demand for female live-out domestics and that income distribution is a relevant factor in the regressions particularly those pertaining to males employed David I. Verway in the private households industry. The lodgings and restaurant industries are important elements in the demand for employees in laundering, cleaning and dyeing occupa- tions and male white collar employment is a significant factor in the demand for barbers. White collar employment is also an element in the demand for cobblers. The results for beauticians and hairdressers were largely insignificant, making it impossible to render any conclu— sion for those occupations. AN ANALYSIS OF THE DETERMINANTS OF ROUTINE PERSONAL SERVICES EMPLOYMENT By David I. Verway A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Economics 1968 w J ,1". y: i a» ‘ ,- J. an r s' . l r N C 1 0-D I.» u .o ‘10 9 e 2 1" .ng5DATTYST ACKNOWLEDGMENTS This research has received generous support from both Michigan State University and The University of Tennessee. The support was in the form of release time from my formal duties as well as ample free time to complete the statisti- cal analysis on their computers. To my committee, especially its chairman, Professor John P. Henderson, my gratitude for their patience in read- ing through several preliminary bits and pieces of manu- script leading up to the final report, and for giving much needed advice. Professor Thomas G. Moore was particularly helpful in suggesting the appropriate statistical technique and interpretation of some of the results. My greatest debt is to one of my colleagues here in the Center for Business and Economic Research at-The Univer- sity of Tennessee, Miss Patricia Ann Price, who typed those many bits and pieces of manuscript and supervised that por- tion of the data preparation and computer work which was done at this university. She also performed the invaluable function of listening attentively and offering advice as I thought aloud about some of the theoretical and statistical problems that emerged. The final typing of the manuscript was done by Ann Brown's Printing and Typing Service in Okemos, Michigan. David I. Verway ii TABLE OF CONTENTS Page ACKNOWLEDGMENTS . 11 LIST OF TABLES. iv INTRODUCTION . 1 Chapter I. ROUTINE PERSONAL SERVICES DEFINED. . 3 II. PREVIOUS STUDIES OF ROUTINE PERSONAL SERVICES. 12 Private Households The Laundry, Laundry Service, and Cleaning and Dyeing Industry Barber and Beauty Shops III. GENERAL THEORETICAL CONSIDERATIONS . 31 Labor Force Participation and Routine Personal Services Occupational Choice IV. PRACTICAL PROBLEMS IN MEASURING UNEMPLOYMENT . 50 V. FORMULATION OF THE STATISTICAL MODEL. . 60 Supply Demand Statistical Considerations VI. RESULTS OF TESTS . . 78 VII. SUMMARY . . . 126 BIBLIOGRAPHY 128 iii Table 10. LIST OF TABLES Employment Status by Age, Color and Sex for the United States: April 1960. . . . . . Female Rate of Employment in Private Households Occupations, Living-in, and Private Households Occupations, Living—out, by State for the Conterminous United States and the District of Columbia April 1: 1950, 1960 and 1960 as a Percent of 1950. . . . . . . . . . . Abbreviations Employed to Designate Variables and Sources of Data . . . . . . . . . Two-stage Least Squares Regression Results for the Rate of Employment of Females in House— holds Occupations——Living-out: 1950. . . Two-stage Least Squares Regression Results for the Rate of Employment of Females in House- holds Occupations—-Living-out: 1960. . . Two-stage Least Squares Regression Results for the Rate of Employment of Females in House— holds Occupations--Living-out: 1960/1950 . Two-stage Least Squares Regression Results for the Rate of Employment of Females in House- holds Occupations--Living-in, Model I: 1950 . Two—stage Least Squares Regression Results for the Rate of Employment of Females in House— holds Occupations--Living-in, Model I: 1960 Two-stage Least Squares Regression Results for the Rate of Employment of Females in House- holds Occupations--Living- in, Model I: 1960/1950.. . . . . . . . . . Two-stage Least Squares Regression Results for the Rate of Employment of Females in House- holds Occupations--Living- -in, Model II: 1950 . . . . . . . . . iv Page 57 69 80 9O 91 92 93 9M 95 96 Table Page 11. Two-stage Least Squares Regression Results for the Rate of Employment of Females in House- holds Occupations—~Living-in, Model II: 1960 . . . . . . . . . . . . . . 97 12. Two-stage Least Squares Regression Results for the Rate of Employment of Females in House— holds Occupations--Living-in, Model II: 1960/1950. . . . . . . . . . . . . 98 13. Two-stage Least Squares Regression Results for the Rate of Employment of Females in House— holds Industry but not in Households Occupations: 1950 . . . . . . . . . 99 14. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Households Industry but not in Households Occupations: 1960. . . . . . . . . . 100 15. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Households Industry but not in Households Occupations: 1960/1950 . . . . . . . . 101 16. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Households Occupations: 1950. . . . . . . . . . 102 17. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Households Occupations: 1960. . . . . . . . . . 103 18. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Households Occupations: 1960/1950 . . . . . . . . 10M 19. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the House- holds Industry but not in Households Occupations: 1950. . . . . . . . . . 105 20. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the Households Industry but not in Households Occupations: 1960. . . . . . . . . . 106 21. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the Households Industry but not in Households Occupations: 1960/1950 . . . . . . . . 107 V Table Page 22. Two-stage Least Squares Regression Results for the Rate of Employment of Males and Females in the Laundry, Cleaning, and Dyeing Occupations: 1950. . . . . . . . . . 108 23. Two-stage Least Squares Regression Results for the Rate of Employment of Males and Females in the Laundry, Cleaning, and Dyeing Occupations: 1960. . . . . . . . 109 2A. Two-stage Least Squares Regression Results for the Rate of Employment of Males and Females in the Laundry, Cleaning, and Dyeing Occupations: 1960/1950 . . . . . . . . 110 25. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Laundry, Cleaning, and Dyeing Occupations: 1950 . . . . . . . . . . . . 111 26. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Laundry, Cleaning, and Dyeing Occupations: 1960 . . . . . . . . . . . . . 112 27. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Laundry, Cleaning, and Dyeing Occupations: 1960/1950. . . . . . . . . . . . . 113 28. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the Laundry, Cleaning, and Dyeing Occupations: 1950 . . . . . . . . . . . . . . 11A 29. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the Laundry, Cleaning, and Dyeing Occupations: 1960 . . . . . . . . . . . . . 115 30. Two-stage Least Squares Regression Results for the Rate of Employment of Males in the Laundry, Cleaning, and Dyeing Occupations: 1960/1950. . . . . . . . . . . . . 116 31. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Hairdressing and Cosmetology Occupations: 1950 . . . . . . . . . . . . . . 117 vi Table Page 32. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Hairdressing and Cosmetology Occupations: 1960 . . . . . . . . . . . . . . 118 33. Two-stage Least Squares Regression Results for the Rate of Employment of Females in the Hairdressing and Cosmetology Occupations: 1960/1950. . . . . . . . . . . . . 119 3“. Two- -stage Least Squares Regression Results for the Rate of Employment of Males in Barbering Occupations: 1950. . . . . . . . . 120 35. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Barbering Occupations: 1960. . . . . . . . . . 121 36. Two-stage Least Squares Regression Results for the-Rate of Employment of Males in Barbering Occupations: 1960/1950 . . . . . . . . 122 37. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Shoe Repair Occupations: 1950 . . . . . . . 123 38. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Shoe Repair Occupations: 1960 . . . . . . . 12A 39. Two-stage Least Squares Regression Results for the Rate of Employment of Males in Shoe Repair Occupations: 1960/1950 . . . . . 125 vii INTRODUCTION This thesis contains the results of statistical tests of cross-section models of the supply of and demand for labor for routine personal service occupations in the United States. The hypothesis is advanced that the supply of routine personal service labor is a function of certain underlying conditions in the labor market: the supply of this labor is a function of unemployment, income distribu- tion, wage structure, and the availability of employment alternatives. As for demand, it is hypothesized that wage structure and income distribution along with other factors such as female labor force participation have a bearing on the rates of employment in the various routine personal service occupations. The statistical models are cross-sectional, with states (U8) and the District of Columbia as units of observation. The data on employment in routine personal service occupa— tions, as well as the bulk of the other statistical materials assembled for the models, are adapted from the Decennial Population Censuses for 1950 and 1960. There are, in fact, two different methods for adapting these data for the purposes intended here. One, the more common, is to use the observations for points in time, either 1950 or 1960. The other involves the use of data pertaining to changes in l the various magnitudes for each state between 1950 and 1960. In the present study, these variables are calculated on the basis of 1960 as a percentage of 1950. The statistical methodology employed is two-stage least squares. Very briefly, the outline of this thesis is as follows. Routine personal services are defined in Chapter I. Chapter II is a review of the literature. In Chapter III, the theoretical underpinnings of the demand for and supply of routine personal services are set forth, while Chapter IV is a discussion of the limitations in some of the data and the means of adapting to them. The statistical model is formulated in Chapter V and the results are pre- sented in Chapter VI. Chapter VII is a summary. CHAPTER I ROUTINE PERSONAL SERVICES DEFINED The term routine personal services was one employed by Stigler to designate those industries which produce services of a routine nature which ". . . can be performed by individuals with little or no formal training, so that H l In many consumers perform these services for themselves. other words, members of the family can perform these ser— vices in the home. Stigler designated as routine personal service industries the following: domestic service; laundering, dyeing, and cleaning services; housekeeping or housing services (hotels and lodging services); and barber and beautician services. In this study the definition of routine personal services differs from Stigler's definition in these two respects: pressing and alteration services and shoe repair services are included in the definition, and hotel and lodging services are excluded. Pressing and alteration services are included because in most families someone can perform these services. A large investment in equipment is not required. These services are included also for the 1George J. Stigler, Trends in Employment in the Service Industries (Princeton: Princeton University Press, 1956), p. 90. pragmatic reason that in some of the data used in this study, the statistics for pressing and alteration service are inseparable from those for laundering, cleaning, and dyeing services. Shoe repair services are included in the present study in the interest of inclusiveness and completeness. For while it is true that professional shoe repair requires specialized machinery, some member of the average household can accomplish shoe repairing of a sort. Hotel and lodging services are excluded from the defi- nition of routine personal services because by their nature they are generally not easily available otherwise and' because their use involves being away from home: similar services are not provided in the home. Lodging is routine only in the sense that many of the employees providing lodg— ing services need little formal training. Lodging services are excluded also for the similar reason that pressing and alteration services are included: the data do not lend them- selves to the method of analysis employed. Hotels tend to be concentrated in large cities, and their services are used by people who, by and large, are not nearby residents; hence, a state-by-state study of factors leading to high or low consumption of lodging services would not be very meaningful. In this study, therefore, the following industries, listed with their Standard Industrial Classification (SIC) number, will be defined as "routine personal service industries": the private households industry, SIC 8811; the laundries, laundry services, and cleaning and dyeing plants subgroup, SIC 721; beauty shops, SIC 7231; barber shops, SIC 72Hl; shoe repair shops, shoe shine parlors, and hat cleaning shops, SIC 7251; and pressing, alteration and garment repair shops, SIC 7271. According to the pub- lished definitions, SIC 8811 . . includes private households which employ workers who serve on or about the premises in occupations usually considered as domestic service. Households classified in this major group may employ individuals, such as cooks, laundresses, maids, sitters, butlers, personal' secretaries, and managers of personal affairs; and outside workers, such as gardeners, caretakers, and other maintenance workers. The households of farming establishments are classified in Major Group 01 . . . .1 SIC 721 includes family and commercial power laundries, SIC 7211; hand laundries, SIC 7212; linen supply establishments, SIC 7213; diaper service establishments, SIC 7214; dry cleaning and dyeing plants, except rug cleaning, SIC 7216; rug cleaning and repairing plants, SIC 7217; and industrial launderers, SIC 7218. SIC 7231 includes "establishments primarily engaged in furnishing beauty services. This industry also includes combination beauty and barber shops."2 1U. 8. Bureau of the Budget, Standard Industrial Classification Manual, 1967 (Washington, D. 0.: Government Printing Office, 1967), p. 307. 21bid., p. 279. 6 SIC 7251 includes barber colleges as well as barber shops while SIC 7251 encompasses, as the name indicates, shoe repair shops, shoe shine parlors, and hat cleaning and blocking shops. Besides garment pressing, alteration and repair shops, as indicated by its name, SIC 7271 includes valet service, fur cleaning, repairing and storage, and cleaning and laundry pick-up stations not owned by launderers or dry cleaners. Though it might be advantageous for pur- poses of this research to exclude some of these groups from the definition of routine personal services, the data that have to be relied upon for the analysis are not available in sufficient detail to allow such exclusions. For example, hat cleaning and blocking is a service not easily performed in the home since blocking requires specialized equipment. But hat cleaning and blocking data cannot be separated from those for other establishments in SIC 7251. The industrial classification system employed by the Bureau of the Census for its decennial Census of Population (CEN) differs in several important respects from the SIC. Though it corresponds fairly closely with the SIC, it is more aggregative for most industries. The industry private households in CEN is the same as SIC 8811 in terms of defi— nition but bears the code designation industry K. For the laundry, laundry services, and cleaning and dyeing plants subgroup (SIC 721), there is no separate detail. In fact, SIC 721 is combined with SIC 7271 in the CEN to form GEN 828, known simply as laundering, cleaning, and dyeing services. GEN 838 is composed of barber shops (SIC 7241) and beauty shops (SIC 7231). For shoe repair shops, the CEN definition (GEN 836) and the SIC definition are the same. In addition to classifying employed persons according to the kind of establishment in which they are employed, the Census Bureau in its CEN classifies them according to the kind of work they do. The kinds of occupations that are relevant to this study, routine personal service occupations, are those that are of a routine nature in the routine per— sonal service industries. Excluded from this investigation of routine personal service occupations are personnel, such as managers, and other white collar workers, such as office employees in a large laundry firm, who may be in the industry but whose occupations do not involve performing routine per- sonal chores. Also excluded are employees, such as laundry Operatives in a captive laundry of a large hotel or hospital, who are performing jobs of a routine nature, similar to the routine jobs in routine personal service industries, but in industries outside of routine personal services. A characteristic of the private household group of occupations is that it is contained entirely within the private households industry. That is to say, any person who is employed in a private households occupation is also, by definition, employed in the private households industry. That is not to say that there are not occupations in the private households industry that are not private households occupations.l Included in the private households occupation groups are babysitters, private household, CEN 801; housekeepers, private household, CEN 802; laundresses, private household, CEN 803; and private household workers, not elsewhere classi— fied, CEN P. The list of individual occupations that were included in the private households occupations group in 1960 includes a host of occupations, most of them involving the performance of tasks familiar in the typical household. These include cook, nursemaid, babysitter, laundress, ironer and kitchen worker. Others, less common, are governess, companion, and steward on a private yacht. Occupations included in the private households industry but not in the private house— holds occupation include captain of a private yacht, chauffeur, gardener, and domestic nurse. Obviously, some of the occupations in the private households industry or occupation are far from routine. Being the captain of a yacht, for example, is not a service that is easily performed in the ordinary private household. Inasmuch as there is no way to alter the CEN data so as to 1Nor is it to say that there are not occupations out- side of the private households occupation group that are similar to ones in it. For example, the occupation butler may be in the private households occupation group as well as the lodgings industry. A hotel butler is not in the private households occupation nor industry, but a butler employed in a private household is. make them correspond exactly with a strictly logical defini— tion of routine personal services, as the term is defined here, there is an irreconcilable data limitation. It seems improbable, however, that this limitation is serious inas- much as these atypical routine personal service occupations are probably of little quantitative importance in the sta- tistics. CEN 674 identifies laundry and dry cleaning Operatives. These occupations involve the performance of tasks such as ironer, marker, folder, presser, spotter, and others that are required in the operation of a commercial laundry or dry. cleaning establishment. While it seems reasonable to assume that in the statistics themselves, most of the persons classi— fied in occupation group CEN 674 are also employed in industry CEN 828, it should be noted that some persons in CEN 674 may be employed in hospital laundries or captive laundries in other kinds of businesses. One small set of occupations in this group is related to another one of the routine personal services. This set of occupations is: hat blocker, hat cleaner, hat finisher, hat former, hat ironer, hat presser, hat renovator, hat sizer, hat steamer, and hatter. These are occupations found in CEN 836, the shoe repair, shoeshine, hat cleaning, and hat blocking industry. The designation for occupations falling under the head- ing barbers is CEN 814. Occupations found in this group and in the corresponding industry are: barber, manager--barber 10 college, manager-—barbershop, and trichologist. Occupations in hairdressers and cosmetologists, CEN 843, include beautician, cosmetologist, manager——beauty parlor, and manager—-beauty school. It may be worth observing that there are occupations, like electrologist, that are found in beauty shops but are not included as routine personal service occupations. Moreover, there are within both occu- pation groups CEN 814 and CEN 843 occupations such as teachers in barber colleges or beauty operators' schools that are not included in the corresponding industry CEN 838. The group designated as CEN 515 pertains to shoemakers_ and repairers, except factory. This group includes occupa- tions in shoe repair shops like cobbler, dyer, helper, and shoe repair shop proprietor. The question of when a service is routine, or what determines if it is routine, may be raised about a number of the personal services. For example, when a woman has her hair styled by a French hairdresser in a beauty salon, she may or may not consider the service a routine one. In a broad sense, however, hairdressing can be done in the aver- age home by nonprofessionals using equipment available at any drugstore. In the same way, Chinese laundrying is not likely to be done in the average home, but laundrying is. All the services listed as routine have the same basic characteristic of being able to be performed in the home, though not necessarily in the specific way, nor with the 11 identical results, that a trained professional might perform them. It is also obvious that some of the routine personal service industries designated above provide services that are not being utilized as substitutes for their production in the home by family members. Industrial launderers, for example, by definition supply services to other businesses such as supplying linen to restaurants and hotels. The necessity for including these industries stems from the necessity to utilize CEN occupational data that contain occupations included in all of these industries. CHAPTER II PREVIOUS STUDIES OF ROUTINE PERSONAL SERVICES This chapter is a review of some previous studies of the industries included in routine personal services. Those for the household industry reveal what is fairly obvious: that persons employed as domestics are generally at the lower end of the social and economic strata. The other routine personal service industries have not been the subject of much previous research. The material in this chapter is arranged according to industry. The household industry is discussed first; then the laundry, dry cleaning, pressing and related industries; and finally beauty shOps and barber shops. There apparently has been no previous research on the shoe repair industry. Private Households In 1897 Professor Lucy Maynard Salmon's first edition of Domestic Service1 appeared. This work is evidently the first major statistical analysis of the employment of domestic servants in this country. Professor Salmon's study was addressed to the question of lLucy Maynard Salmon, Domestic Service (8th ed.; New York: The Macmillan Company, 1911). l2 She l3 . whether household employments are justified in resenting any intrusion into their domain, whether the individual employer is right in con— sidering household service exclusively a personal affair. continues, An answer to the question may be of help in deciding whether the difficulties that are found in the present system of domestic service arise in every case necessarily from the personal relations which exist between employer and employee, or are largely due to economic conditions over which the individual employer has no control. Still further, the con— clusions reached must determine somewhat the nature of the forces to be set in motion to lessen these difficulties.1 Salmon sent questionnaires to ". . . all housekeepers and their employees who can be communicated with by the members of the Classes [Vassar] of '88 and '89 and the Department of History."2 Relying upon the 3,550 replies to these questionnaires, data from the Eleventh Census of the United States, and various other published materials, Salmon derived three sets of propositions relating to "economic phases of domestic service." The first group of propositions concerns national and racial origin of domestics. (l) A large proportion of the domestic employees in the United States are of foreign birth. . . . In nine states and territories the number of foreign born domestic employees exceeds the number of native born white employees, in sixteen about one half of the white domestic employees are of foreign birth, in twenty-four states and territories the number of native born white employees largely exceeds the foreign born, while in fifteen states colored employees are in excess. l 2 Ibid., pp. 5-6. Ibid., p. 305. 31bid., pp. 74-76. The 14 (2) The converse of the preceding proposition is also true--the concentration of women of foreign birth engaged in remunerative occupations is on domestic service. (3) The foreign born population as a class seek the large cities. (4) The foreign countries having the largest absolute representation in the largest cities are Ireland, Germany, Great Britain, Sweden, and Canada and Newfoundland.3 (5) The foreign countries having the largest absolute and relative representation in domestic service are, in order, Ireland, Germany, Sweden and Norway, Great Britain, and Canada and Newfoundland. conclusion reached from this set of propositions is that . with the exception of the sections employing colored servants, domestic service is as a rule performed by persons of foreign birth belonging to a few well-defined classes as regards nationality, who prefer city to country life.» A second set of propositions relates to the influences of urbanization, the general level of wealth, and the availability of alternative employments. (l) The number of domestic servants is absolutely and relatively small in agricultural and sparsely settled states.6 (2) The number of domestic servants is absolutely and relatively large in those states containing large urban populations.7 (3) The aggregate wealth of a state has little appreciable effect on the relative number of domestic servants employed.8 1 Ibid., p. 77. 21bid. 31bid., p. 78. ulbid. 51bid., p. 80. 61bid. 8 71bid. Ibid., p. 82. 15 (4) The per capita wealth of a state has, with the exception of the Southern states as a class, a somewhat important bearing on the relative number of servants employed.1 (5) Domestic employees are found in the largest numbers relatively and absolutely, in the large cities.§ (6) The proportion of persons engaged in domestic service varies with geographical location and prevailing industry.3 (7) Neither per capita wealth nor aggregate wealth has an appreciable influence in determining the number of servants in cities. (8) The prevailing industry of a city, rather than its population or wealth, determines the number of domestic employees. . . . several of the manufac— turing cities rank comparatively high in per capita wealth.5 The general inference made from these propositions is that In states containing a relatively high urban population it is possible for wealth to command the services Of a large proportion of persons for work in domestic service. But in cities where wealth comes into com— petition with manufacturing industries the proportion of domestic servants is small. Where such competition does not exist the proportion is large. In other words, persons are willing to enter domestic service for a consideration in cities where no other avenues of work are open to them with the qualifications they possess. They are unwilling to do so where such Openings do exist. The third and final set of propositions relating to the economic aSpects of domestic service is about wages and hours of work. lIbid. 21bid., p. 83. 31bid., p. 84. “Ibid., p. 86. 51818., p. 87. 61bid., p. 88. 16 (1) Wages in domestic service vary in different sections according to the economic conditions of the several localities.l (2) Skilled labor [within domestic service] commands higher wages than unskilled labor.2 (3) The foreign born in the domestic service receive higher wages than the native born. . . . An explana- tion is found in three facts: (1) the preference of the foreign born for the large cities, where wages in domestic service are higher than in the country; (2) the large proportion of [N]egroes among the native born; (3) the relatively better class of foreign born than Of native born women who enter domestic service.3 (4) The wages of men engaged in domestic service are higher than the wages of women. (5) A tendency is found towards an [historical] increase in wages . . . .5 (6) The wages received in domestic service are relatively and sometimes absolutely higher than the average wages received in other wage—earning occupations Open to women. [This seemingly peculiar statement means that net wages, after room and lodging, are higher for domestics than for women in other occupations. For some domestics, money wages alone are higher.]6 Salmon gives a corollary to this proposition: High wages alone are not sufficient to counterbalance the inducements offered in other occupations where wages are relatively or absolutely lower but whose special advantages are deemed more desirable.7 (7) The wages paid in domestic service are on the average high, but the occupation Offers few Oppor- tunities for advancement in this direction.8 lIbid. 21bid., p. 89. 31bid., pp. 91-92 “Ibid., p. 92. 51bid. 61bid., p. 93. 7 8 Ibid., p. 103. Ibid. l7 (8) The amount of time unemployed [the unemployment rate] is less in domestic service than in nearly every other occupation.l (9) High wages are maintained without the aid of strikes or combinations on the part of the employees.2 From this set of prOpositions and the underlying data Salmon deduces . . . the conformity of wages in domestic service to certain general economic laws, the fact that the wage factor alone does not determine the number of persons in the occupation, and the existence of a few conditions which affect, perhaps unconsciously, 3 the willingness of the women to engage in this work. On the face of them, except for number seven, these propositions would appear to indicate that domestic service. is a most desirable occupation from the employee's per- spective. At the end of the last century this appeared to be a better than average paying job for women and the element of job security was favorable. That the contrary is true, that domestic service was not deemed a desirable occupation emerges from answers to the questionnaires sent out. Out of 562 answers by employees who were asked the question "What reasons can you give why more women do not choose housework as a regular employment?" inferior social status was given as a reason by 157 employees.“ There was Obviously a strong social stigma attached to this type of employment. An almost equal number of replies pertained to confinement on evenings and Sundays along with other 2 llbid., p. 104. Ibid. 31bid., p._lo6. LlIbid., p. 140. 18 manifestations of lack of independence in domestic service. Another serious objection to this employment was irregularity in working hours. Shortly after the turn of the century, another study of domestic service was carried out by Gail Laughlin.l This study involved a questionnaire approach similar to that employed by Salmon but was less ambitious than its predecessor. The results are so remarkably similar that no good purpose will be served by reporting them here in their entirety. Three paragraphs of Laughlin's report are worth reproducing here because of their incisiveness with respect to the objections to employment in domestic service. In speaking of the reasons which prevent women from entering domestic service, Dean Marion F. Talbot, Of the University of Chicago, expressed the Opinion that the objections already referred to, viz, indefiniteness of hours, unfit sleeping accommoda— tions, the imposition of restrictions, etc., were causes, of which social position was the result. The reasoning is valid, but these conditions are‘ themselves results from an underlying cause. That underlying cause is the basic principle upon which the whole system of domestic service, as it exists to-day, rests; and that principle is that in domestic service it is the person who is hired and not, distinctively, the labor of the person. In all other occupations it is labor which is contracted for; in domestic service it is, in effect, at least, the laborer. In other occupations the contract is for the performance Of certain specified services; in domestic service the contract is, usually, for the entire time of the laborer, who is then expected to‘ perform, not only certain labor which has been specified, but, in addition to that, is expected to perform whatever services may be required; who is expected, in short, to be at all times subject to the call and direction of the employer. . . . .‘ lGail Laughlin, "Domestic Service," Report of the Industrial Commission (Washington: U. S. Government Print- ing Office, 19017T 19 The services demanded, in many cases, of domestic workers are in accordance with these views. Fre- quently, perhaps usually, the general servant is expected not only to cook, wait on table, and perform such other duties as may be included among legitimate household duties, but she is expected also to run on errands to any part of the house for any member of the family, and to perform various other personal services for any member. Household labor has not had applied to it the economic principles which have been applied to other occupations. It has not been put on a business basis. The relation of employer and employee is still regarded as largely a personal relation. The vast majority of household workers are wives, who give their services on a purely personal basis. This fact has had a considerable effect in making the relations of hired household employees with their employers more personal than economic. But a per- sonal relation between employer and employee inevitably becomes the relation of superior and inferior, rather than a contract between equals, and this is what has developed in domestic service. To remove the social stigma from domestic service, and thus to attract into that service a larger number of intelligent employees, household labor must be established on a business basis.1 A study of more specialized nature was Isabel Eaton's "Special Report on Negro Domestic Service in the Seventh Ward Philadelphia." Eaton like other students of the sub— ject noted that Negroes loomed in disproportionate numbers in domestic service and made this observation. The probable reason for this disproportion is not far to seek when we remember the unpopularity Of domestic service which keeps whites out, and reflect that the colored prejudice which is known to Operate against the Negro in nearly all departments of labor lIbid., pp. 759-760. It is interesting to note, par- enthetically, at this point in the discussion that the National Committee on Household Employment was formed in 1965 partly to act as a ". . . clearinghouse and coordinator for all organizations concerned with upgrading the status of private-household employment . . . ." United States Women's Bureau, 1965 Handbook on Women Workers, Bulletin No. 290 (Washington: U. S. Government Printing Office, 1966), p. 271. 20 except drudgery, actually works in his favor in the matter of domestic service, where the competence Of Negro waiters and the superior skill of Negro cooks is generally admitted. Hence, Negro labor, following the line of least resistance, flows in enlarged streams into the channel of domestic service.1 Noting that domestic service at the turn of the century attracted mainly young persons, Eaton commented: The fact that the highest point of excess of youth . . . is reached at twenty-three to twenty— five years is significant, and suggests the query why it is that domestic service so clearly attracts the young of both sexes and of all races. It is safe to say that one of the most prominent deter- mining causes is necessity of immediate income. Many young men and women are obliged by circumstances to undertake some form Of work which, while requiring no capital and no particular course of training, still yields an immediate return, which is certain to provide them at least their board and lodging, with a small amount for living expenses. This is the chief reason why the first employment of young men and women just beginning to support themselves is so Often "going out to service." Eaton also finds evidence of color discrimination even within the domestic service industry. Nonwhites evidently receive less pay for the same position as whites. In general the Eaton study corroborates the findings of other authors that employment as a domestic carries with it a social stigma which renders it an occupation that is not eagerly sought after by the typical employable individual. lIsabel Eaton, "Special Report on Negro Domestic Service in the Seventh Ward Philadelphia," chapter in The Philadelphia Negro (Philadelphia: University of Pennsyl- vania, 1899), p. 434. 21bid., p. 443. 21 In "America's Domestic Servant Shortage," Ethel M. Smith examines the effect of the then new and restrictive immigration law and concludes, It seems far more probable that it is the changing occupational status of women in Europe as well as in America that is primarily responsible for the con- tinuing problems of shortage in domestic service wherever it occurs.1 Another comment worth repeating here is in regard to status and pay. The social stigma, the low wages consequent on this and other things, the isolation of the job, its long hours and its complicated requirements under average conditions have not stood comparison with the regular hours, the better pay, the better social status and the companionship of factory, store, office or telephone exchange. The household and kitchen occupations are the least standardized, the least modernized, the most feudal of all the work in the modern world. Apparently, by the mid 1920's domestic service had lost its competitive edge with respect to the rate of pay. Fortune magazine examined "The Servant Problem" in the late 1930's and made more or less the same conclusions regarding the matter as were made in the previous studies. In 1938 Fortune asserted On the one hand there are people with money to spend for domestic service. And on the other there are 8,000,000 unemployed. It is an appalling situation. Of that 8,000,000 is it not likely that a large number are highly eligible for domestic service? Why is it not possible for at least 1,000,000 unemployed to find homes with 1,000,000 families in lEthel M Smith, "America's Domestic Servant Shortage," Current History, XXVI (May, 1927), p. 218. 21bid. (Italics supplied.) 22 which the wife is overworking herself? And if those 1,000,000 families could not afford full- time wages, would it not be possible for the government (which will help you to build a house, and which is supporting the unemployed anyway) to make up the difference? It is a tantalizing question, and there is an answer to it, and the answer is no. And the reason the answer is no lies with the women, who have not succeeded in solving the servant problem. And the reason they have not solved the servant problem is that they have not struck at the hidden root of it.1 George J. Stigler investigated the subject of employ— ment in the private households industry and reported his findings in a National Bureau of Economic Research monograph.2 His research was primarily concerned with the reasons for the decline in the servant population relative to the popu—8 lation as a whole between 1900 and 1940 in the United States. Stigler, like Salmon, noted that wealth has no Obvious effect upon the number of servants and hypothesized that eguality of the distribution Of income, rather than the amount, may be a factor of considerable impor- tance. A society with relatively many families at both ends of the income scale would provide goth a large supply of servants and a large demand. Stigler examined the racial and geographical characteristics of servants and found that 1"The Servant Problem," Fortune, March, 1938, p. 82. 2George J. Stigler, Domestic Servants in the United States 1900-1940, Occasional Paper NO. 24 (New York: National Bureau of Economic Research, 1946). 31bid., p. 6. 23 The low social status of domestic service, the absence of vocational or educational requirements, and the discrimination practices in other lines of employment seem adequate to explain the fact that immigrants and [N]egroes have constituted more 1 than half of female servants since 1900 . . . There are he finds, three levels of use of domestic service. In the South there is a servant for every 10 families, in the northeastern states one for every 14, and else- where one for every 20. Since [N]egroes and immigrants have supplied a majority Of servants, high levels in the South and along the eastern seaboard are to be eXpected.2 In one of his statistical analyses using data from the 1940 POpulation Census, Stigler found that average annual earnings of female servants varied positively with city size and negatively with percentage of the servant population classified as non-white. He also found that the distribution of earnings among female servants was relatively unequal compared with service workers, manufacturing Opera- tives and clerical workers. With respect to the length of the workweek, Stigler asserted that "both extremely short and long hours are common in domestic service."3 Among factors affecting employment of servants, Stigler discussed family size, female labor force partici- pation, urbanization, family income, household technology (adOption of vacuum cleaners, washing machines and other lIbid. 2Ibid., p. 9. 31bid., pp. 19-20. 24 appliances), manufacture of prepared foods, movement into apartments where upkeep is less, and the decline in boarding houses. The major results of the study are: l. x1 = 204.1 - 1.58 x2 + 4.38 x3 R: .731 (.32) (.81) where X : average 1939 earnings of full time servants outside of cities of popu- lation greater than 250,000. X : percentage of servants who are non- white outside of cities Of population greater than 250,000, and X3 : percentage of servants in cities of population greater than 10,000 but less than or equal to 250,000 The units of observation are the 48 states, and the data are from the 1940 Population Census. 2. An income elasticity of demand for servants of 2.0. 3. A price elasticity of demand of —2.3. Both two and three are based upon the 1935-36 Consumer Purchases Study by the Bureau of Labor Statistics. 4. x1 = 56.27 (100 — x2)‘°33SO R: .906 where Xl : percentage of female workers who are servants and X2 : percentage of female servants who are nonwhite. The units of Observation are 33 large cities, and the data are from the 1940 Population Census. 25 5. X = .307 + .367 X2 - .00136 X3 R: .701 (.081) (.00097) X1 = .349 - .293 X2 where Xl : ratio Of servants to service workers' wages, X2 : ratio of the number of servants to the number of service workers, and X3 : percentage of nonwhite servants. The units of Observation are large non-Southern cities and the data are from the 1940 Population Census. 6. United States (48) X1 = 30.34 - .278 X2 + .046 X3 R: .510 (.074) (.015) Southern States (14) x1 = 81.25 — .365 x2 + .055 x3 R: .589 (.156) (.023) Other states (34) X1 = 12.72 — .096 X2 + .028 X3 R: .507 (.051) (.011) where Xl : servants per 1,000 families, X2 : mean annual wage, and X3 : income per family. One of the important Observations made by Stigler in connection with "the servant problem" is that "if there is a servant problem it is primarily the problem of offer- ing enough to draw persons into domestic service."l lIbid., p. 36. 26 Previous studies had centered on the low social status of domestics as a cause of the "shortage" of servants. It is, of course, probably true as Stigler indicates that a rise in the servant wage rate would do much to eliminate the so-called shortage. There are several other published works on domestic service. Most of these are listed in one or another of the bibliographies published by the Women's Bureau. Nothing would be gained by examining these items here since interest centers primarily on the character of the industry and that seems to have been fairly well established from the sources cited above. It does seem worthwhile, however, to mention some bits and pieces of studies done in another connection that have a bearing on the central subject of this section. The first Of these is the chapter on routine personal ser- vices in Stigler's Trends in Employment in the Service Industries, mentioned previously. The bulk of the material in Trends' section on domestic service is based on Stigler's previous monograph on Domestic Servants. One important addition is his regression analysis which includes a measure of income inequality. X1 = 5.82 — 0.109 X2 - .00024 X3 + 0.511 X4 R:.94 (.0032) (.00059) (.096) where Xl : servants per 100 families in 1940 X2 : average annual wage of a servant in 1939, X : income payments per family in 1940, and 27 X4 : percentage of income received by upper one per cent Of income recipients in 1940. A comparison of this equation with the 48 states' equation reproduced above (Stigler, item 6) reveals that family income becomes insignificant in the regression when a measure of income distribution is introduced. This sub- stantiates Stigler's previous argument that income dis- tribution rather than level determines the relative magni- tude of household employment. The Laundry, Laundry Service, and Cleaning_anngyeing Industry An early study of the power laundry industry indicated that much Of the work itself in a power laundry was unpleas— ant, requiring constant standing and for some occupations, considerable muscular strain. These work patterns, along with a high level of noise, meant that these workers gen— erally suffered from fatigue by the end of the workday. The study also pointed out that many of the plants in this industry had a warm humid atmosphere.l There was apparently also a serious deficiency of many Of the amenities found in other industries, things like adequate and clean washroom and toilet facilities, cool drinking water, and lunch or rest rooms. All in all, the report of the Women's Bureau suggests that Operatives in power laundries suffered from loathsome working conditions. lEthel L. Best and Ethel Erickson, A Survey of Laun- dries and Their Women Workers in 23 Cities, Women's Bureau Bulletin NO. 78I(Washington, D. 0.: Government Printing Office, 1930), pp. 17-22. 28 The length of the workweek in power laundries varied according to the section of the country. The most characteristic week, by section, was as follows: Per Cent of the Women Western 48 hours and under 97.2 Eastern do 80.2 Middle Western 50 and under 54 hours 51.7 Southern 54 hours and over 48.41 Out of 19,180 women in the survey, 5,076 were Negroes.2 Data on wages are also given in this study, but there is no comparison with the wage rate in other industries. Within the power laundry industry itself there was a rather wide dispersion in wages depending on the particular job or occupation within the industry. Nonwhites earned sub- stantially less than white women irrespective Of occupation. A recent survey of problems and prospects in the laundry and dry cleaning industry reveals something about the nature of employment in this industry today. The Operations of launderers and cleaners call for several kinds Of labor. The biggest group, and the most costly in the aggregate, is in production. These are largely unskilled workers who receive training sufficient to perform tasks in the marking, lIbid., p. 43. 21bid., p. 61. 29 sorting, washing, finishing, and assembly operations. Finishing requires the greatest number Of workers, as hand labor is in some way involved with every item processed. Men are generally employed for the washing, extraction, and drying operations; women predominate in finishing and other production jobs, except in drycleaning plants. Other important job classifications include office workers, salesmen, and delivery, or route salesmen. Office skills are, of course, required of the first of these groups. Route salesmen should be able to handle relationships with customers or potential customers. Management has a serious problem in hiring and training workers whose productivity can be maintained or improved. Most owners are convinced that the cost of labor limits them to the unskilled labor market. Those recruited must be willing to work for low wages and be adaptable to the training necessary to perform at an acceptable level. A high rate of absenteeism is likely to be a serious problem with such employees. It is also observed that power laundries must compete with laundromats, and wash and wear fabrics. Dry cleaners also have been affected by competition to some extent, but this segment of the industry has continued to register growth. Drycleaners have not been influenced by the external competition as have launderers. And rising produc— tivity enables cleaners to resist the upward pres- sure On prices better than laundries. The cleaner is still in the enviable position of offering services2 for which many consumers feel he is their only choice. Barber and Beauty Shops Stigler examined some statistics for barber and beauty shops and noted that 1Business and Defense Services Administration, Th2 Laundry and Drygleaning Industry, A Study Of Problems and Prospects (Washington, D. 0.: Government Printing Office, 1965), p. 48. 21bid., p. 2. 30 the number of barbers has not grown as rapidly as the male population, while the number of workers in beauty parlors has increased many fold more than the female population.1 He cites the safety and electric razors along with the rising popularity Of being closely shaven as causes of the relative decline in barbering. Women's fashions and the invention of the permanent waving process are given as reasons why the beauty shop industry has expanded relatively. He notes that "both the barber and beauty parlor industries are organized in small shops, operated chiefly by single proprietors."2 Stigler also notes that state licensing requirements pose something of a barrier to entry into the barbering occupation, but the effect of these barriers on average wages for the industry is difficult to measure.3 lStigler, Trends in Employment in the Service Industries, p. 101. 21bid., p. 103. 31bid., p. 105. CHAPTER III GENERAL THEORETICAL CONSIDERATIONS In setting up the necessary hypotheses to be tested empirically, it is necessary to make a theoretical investi- gation as a means of uncovering the general principles that may be expected to govern employment in these occupa— tions. This chapter delves into such matters as the manner in which peOple make decisions about their labor force participation and the resultant implications for routine personal services, factors influencing occupational choice and the implications for routine personal services, and the effects of unemployment on routine personal service employ- ment. Labor Force Participation and Routine Personal Services That the household or family is the relevant decision making unit for studying consumption behavior has been recognized. In 1962, Jacob Mincer wrote: The analysis of market labor supply in terms of consumption theory carries a strong connotation about the appropriate decision—making unit. We take it as self—evident that in studying consump— tion behavior the family is the unit of analysis. Income is assumed to be pooled, and total family consumption among family members depends on tastes. It is equally important to recognize that the decisions about the production of goods and 31 32 services at home and about leisure are largely family decisions. The relevant income variable in the demand for home services and for leisure of any family member is total family income. A change in income of some family member will, in general, result in a changed consumption of leisure for the family as a whole. An increase in one individual's income may not result in a decrease in his hours of work, but in those of other family members. The total amount of work performed at home is, even more clearly, an out- come of family demand for home goods and for leisure, given the production function at home. However, unlike the general consumption case, the distribution of leisure, market work, and home work for each family member as well as among family members is determined not only by tastes and by biological or cultural specialization of functions, but by relative prices which are specific to individual members of the family. This is so, because earning powers in the market and marginal productivities in alternative pursuits differ among individual family members. Other things equal (including family income), an increase in the market wage rate for some family member makes both the consumption of leisure and the production of home services by that individual more costly to the family, and will as a matter Of rational family decision encourage greater market labor input by him (her).1 We may envisage human preferences as governing deci- sions to sell labor in the market place in the following manner. Each decision unit, be it family or single indi— vidual, must make purchasing, investment, and labor market participation decisions as a supplying unit. These two sets of decisions are inter—related and depend upon prefer- ences and the inventory or resources and abilities within the unit, and market or other constraints without. One of lJacob Mincer, "Labor Force Participation of Married Women," in Aspects of Labor Economics (Special Conference Series NO. 14) (New York: National Bureau of Economic Research, 1963), pp. 65—66. 33 the alternatives that is relevant to the present discussion is between labor and nonlabor. Whether or not the unit must indulge in labor depends upon its ability to supply its needs with recourse to labor: viz, its inventory of re- sources, both financial and nonfinancial. The greater this inventory, and the return that may be earned from it, the smaller is the need for the unit to indulge in labor activity. Another alternative is between market and nonmarket labor. Given that the unit chooses to utilize some of its labor resource, it must decide whether to engage in home produc- tion, participate in the labor market, or employ some com- bination of the two. The prospects of Obtaining a satis- factory return in the market place may be so minimal that the unit will utilize all of its labor for home production, engaging in, say, subsistance agriculture. Or one member of the family may go into the market with his labor in exchange for wages with which to purchase goods and ser- vices in the market place for the satisfaction of the family. The other members might provide the remainder of the family's wants through the use of their nonmarket labor. This is the typical arrangement in many homes with the husband selling his labor in the market place while the housewife provides nonmarket labor for the accomplishment of the hOusehold chores. And there may be various combinations in between these extremes and beyond them. The head may work full time and the wife part time. The family may be part 34 time subsistence farmers and part time laborers. Both might work full time, hiring a domestic to do the house- hold labor. In one sense, labor market participation for the family or single individual is a continuum or scale run- ning from zero to a maximum of 100 per cent. At the zero point on this continuum, the unit will either not engage in labor activity at all, because of a large bank account or other means which allow it to consume without currently producing, or it will engage in home production. There may be another kind of continuum here inasmuch as the unit may combine some production with a drawing down of an inventory of resources or utilizing the return from them to finance current consumption if it has such an inventory. It may, moreover, engage in home production and exchange some of its fruits for other items in the market place. At the other extreme, the unit engages in no home production and has no inventory, but exchanges all of its labor in the market place for money or income in kind with which to purchase other goods and services. Where will the unit locate on this continuum? Assum- ing that it will endeavor to maximize its satisfaction, and given its inventory, it will examine the situation in the market place and make comparisons of the labor requirements for supplying the needs of the unit. A unit with extremely high earning capacity in the market would probably sell all l'llulll'llll‘ 35 of its labor in the market place and with the wherewithal purchase all of its consumption items in the market place. It might have a housekeeper and a chauffeur and hire people to do some of the more specialized tasks around the home such as repairs and maintenance and occasional interior redecorating. The unit may be looked upon as a producer which purchases inputs such as maid and home repair service and produces output, the product of its labor in the market place. The inputs supply the foregone home-production that it may maximize its satisfaction by its greater earnings in the market place. Looked at in a slightly different way,. the market value of the labor of this unit exceeds its value in home production. The decision to engage in remu- nerative work makes the unit at the same time a supplier of labor in the market place and a demander of goods and services to replace home-production. This aspect Of labor force participation is probably best thought of as the opportunity cost aSpect. For example, in considering the demand for domestics, Mincer noted that "The wage rate of the domestic servant must be viewed in relation to the price of employing the wife at home, which is the opportunity cost of foregone earnings in the market."1 1Jacob Mincer, "Market Prices, Opportunity Costs, and Income Effects," in Measurement in Economics, Studies in Mathematical Economics and Econometrics in Memory of Yehuda Grunfeld (Stanford, California: Stanford University Press, 1963), p. 74. 36 Another consideration that needs to be touched upon here is one relating to the value of time used in consump- tion. Becker has noted that For example, the cost of a service like the theatre or a good like meat is generally simply said to equal their market prices, yet everyone would agree that the theatre and even dining take time, just as schooling does, time that often could have been used productively. If so, the full costs of these activities would equal the sum of the market prices and the foregone value of the time used up. In other words, indirect costs should be treated on the same footing when discussing all non—work uses of 1 time, as they are now in discussions of schooling. In other words, "Behind the division into direct and indi- rect costs is the allocation of time and goods between work—oriented and consumption-oriented activities."2 Therefore, the structure of the economy will be determined to some extent by this allocation Of time be- tween work—oriented and consumption-oriented activities. It seems worth mentioning that the division between work- oriented and consumption-oriented activities is not without some ambiguity. In order to bring out the difficulty of this division of concepts it is useful to introduce the idea Of consumption involving either the active or passive participation of the person or persons engaging in the consumption activity. This ambiguity may also have a slightly different manifestation in that remunerative 1Gary S. Becker, "A Theory of the Allocation of Time," The Economic Journal, September, 1965, p. 494. 2 Ibid., p. 499. 37 labor may have some consumption orientation, to the extent that it is enjoyable. With respect to the question of why women work, it has been noted that Financial remuneration is, however, not the sole reason that so many women are in the labor force. It is significant that the more education a woman acquires, the more likely she is to seek paid employment, irrespective of her financial status. The educated woman desires to contribute her skills and talents to the economy not only for the financial rewards, but even more to reap the psychic rewards that come from achievement and recognition and service to society.1 Active participation in consumption activities may be thought of as those consumption-oriented activities that involve the creation of something. Most hobbies, for example, result in the creation of some end product, say, a rose garden. On the other hand, passive participation in a consumption activity may be defined as that which causes no product to be forthcoming like, say, watching a play or a movie, or simply daydreaming. Viewed in this manner, the division between work-oriented and consumption— oriented activities is somewhat arbitrary. Work, either in the market place or in the home, may be enjoyable and have some consumption aspects. Leisure activity which involves the creation of something either tangible or intangible, has some of the aspects of production or work. The amount Of home production Of the unit is partly a 11965 Handbook on Women Workers, Women's Bureau Bulletin No. 290 (Washington, D. 0.: Government Printing Office, 1965), p. 5. it‘ll-l 38 matter of taste. One may or may not wish to have a rose garden. It seems appropriate at this juncture to examine in some detail the factors that may be weighted in the con- sumer unit's decision to purchase those services which have been defined as close substitutes for home production. Let us consider first those routine personal services which involve the release of time Of one or several family mem— bers. These are ones which involve the performance of some of the household chores such as cooking meals, cleaning, clothing and shoe care, caring for younger members of the family, and shopping for the day—to—day household needs. The alternative ways in which the family meals may be prepared include having a mamber of the family cook meals, hiring a domestic to cook or to cook and do other household chores, or eating out. If a family member or domestic pre- pares the food in the household, then there is a variety of methods of accomplishing that task. The latest equipment and prepared foods may be used in order to economize on time spent on food preparation or more basic methods may be employed. In addition, some combination of home and outside the home food preparation and consumption may be used as in the case of catering where the food is prepared away from home but consumed in the home. Another release-time case worth considering is the doing Of the family washing. Laundering may be accomplished 39 by the housewife using primitive or modern home laundry equipment, by a domestic using primitive or modern home laundry equipment, or by someone outside of the home, a washerwoman or commercial laundry, which specializes in laundering. It should also be noted parenthetically here that it may be done by a family member or domestic in a laundromat, but that for a family member to do it requires a time input on the part of the family member. An addi- tional consideration is the use of fabrics which minimize the effort required to produce a neat appearing garment, such as drip dry shirts. Some clothing care, such as dry cleaning, is not really too amenable to home production or accomplishment by a domestic. Often special techniques or chemicals are required which makes it rather unlikely that the chore will be performed in the average household. But then too, Spot remover and an iron are satisfactory substitutes for dry cleaning in many households. Shoe upkeep is much like dry cleaning in that it is really not too amenable to home production or accomplishment by a domestic. Commercial shoe repair probably should not be considered as a very close substitute for home produc- tion. It should also be noted that there is an alternative to shoe repair, and this alternative is the purchase of new shoes when the Old ones become sufficiently worn as to warrant either repair or replacement. 40 General cleaning and upkeep of the household must perforce be done on the premises. Accordingly, substitutes for home production of the house cleaning are pretty much confined to hiring a domestic. Here too, however, it should be noted that there are devices such as vacuum cleaners that may be used to diminish the burden of the chore or to shorten the length Of time required for its performance. Babysitting may be done either on the premises or in the home of another. Nurseries compete with babysitters in the private household. It may be noted that one or many persons can be hired to perform these chores. A single domestic may be required to babysit, clean and cook, or a domestic may be hired to assist the housewife in the domestic chores. Other routine personal services require participation of the consumer of the service. The occupation companion is one of these. Indeed a companion is hired for the express purpose of accompanying the employer. For the most part, the routine personal services associated with cosmetology also require the presence of the consumer of the service, as being barbered or manicured or having one's hair shampooed and set. In a sense, the purchase of those routine personal services having the characteristic that they release some family member for other prusuits may be thought of as a 41 purchase of time. The time may be utilized either for con- sumption or production. A babysitter may be employed in order to free the housewife for remunerative or non— remunerative employment outside of the home. Or the house- wife may simply idle away the released time. True some domestics may be hired primarily as a means of conspicuous consumption a la Veblen. But this situation too probably has the outward manifestation of time release since the domestic so hired, classified as babysitter, maid, or what— ever, does at least bear the title of one who performs release time activities. Routine personal services having to do with the appearance of a member of the unit may also be production or consumption motivated. To the extent that appearance is important to the occupations of the persons in the con- suming unit, barbers, beauticians, shoe repair and clothing care services become production inputs since they are important to the appearance of the individual. On the other hand, they are also partly consumption oriented to the extent that they are desired in themselves for the feeling of well being they give to the user. The purchase of routine personal services, then, may be motivated in part by a motivation to get release of time for engaging in some other form of production, as an input for production, or by a motivation to consume either the service itself, or time freeing for other consumption. 42 One additional consideration of the motivation for this consumption is that inasmuch as work may itself produce some satisfaction on the part Of the worker, the apparent production motivation for consuming routine personal ser— vices may contain an admixture of consumption motivation, if the consumer employs routine personal services so as to be able to engage in remunerative or non—remunerative work that he enjoys. It may be worth noting here also that there are other services as well as goods whose consumption is multifaceted in the above manner. Automobile tune-up is an example Of a service that frees the automobile owner from home production for a work-oriented or a consumption-oriented activity. These other substitutes for home production are not the object of this study, however, and there seems little to be gained by dwelling on this matter of the motivation behind their consumption at this juncture. They may, it should be noted, however, be competitive with some routine personal services. T.V. dinners, for example, compete with the services of a domestic cook. Routine personal services consumption, then, can require the consumption of time, or it can effect the free- ing of time of a family member. Whether it is time freeing or time consuming service, on the part of family members, it may be consumption or production motivated, or some com— bination Of both. Relevant considerations besides income and price are opportunity cost and the value of time. 43 Occupational Choice Up to now the discussion has proceeded on the basis of consideration of the decision unit, be it a single individual or family. For a single individual, one person both makes and implements the decision; but although a family unit may make the decision by some process, it is individuals within the unit which carry out the decision. Each member, discrete within the unit, either participates in the labor force or does not. Each member possesses desires and abilities that are unique. Given a family process of balancing preferences, the participation of each member will depend to some extent upon the contribution that he can make to the total family satisfaction in his various uses in home production or market production. For some family members, the unit may decide upon investment in human capital as the Optimum choice. Mores and social values may have a deciding influence on the decision of some units. It is still a value widely accepted that the proper role of the woman is the home. There seems to be no really detailed theory of occupa- tional choice which reveals why, for example, doctors in business today became doctors rather than, say, plumbers. Some very worthwhile research has been done in the area of occupational choice, however, and some of the general find— ings seem relevant here. Ginzberg, one of the pioneers in the area, from intensive interviewing of a sample Of students 44 in various stages of the formal educational process, drew several general conclusions about occupational choice. The outstanding conclusion from our findings is that occupational choice is a developmental process: it is not a single decision, but a series Of deci- sions made over a period Of years. Each step in the process has a meaningful relation to those which precede and follow it. From this primary finding, there follows a second important generalization: the process is largely irreversible. This is a result of the fact that each decision made during the process is dependent on the chronological age and development of the individual. Time cannot be relived; basic education and other eXposures can only be experienced once. Of course, the individual can shift even after he has tenta- tively committed himself to a particular choice. But the entire process of decision-making cannot be repeated and later decisions are limited by previous decisions. The primary finding that occupational choice is a process leads to a further generalization: the process ends in a compromise. Throughout the years of his development the individual has been trying to learn enough about his interests, capacities, and values and about the opportunities and limitations in the real world, to make an occupational choice that will yield him maximum satisfaction. If he could base his choice on but one element, such as his interests or capacities, without regard for the job market, the income structure, and the social prestige which attaches to different kinds of work, his choice should be simple and direct. However, a series of factors, both internal and external, affect his decision. He must renounce to some degree the satisfactions which he might derive if he based his choice exclusively on a strong interest, a marked capacity, or a realistic Opportunity. He must find a balance among the major elements. Hence, the compromise aspect of every occupational choice.1 Elsewhere it is noted that The differences in eXposure and stimulation in the environments of the upper and lower income groups contributed to differences in decision-making, for lEli Ginzberg et al., Occupational Choice, An Approach to a General Theory (New York: Columbia University Press, 1951), pp. 185-186. 45 occupational choice is greatly influenced by family, community, and school. These differences between the two groups indicate that the upper income group has a much wider range of choices and is in a much better position to Obtain whatever preparation is required for the realization of their final choice. For instance, the high school senior in the upper income group who was looking forward to studying medicine not only had no anxiety about financing the long period of preparation, but he already knew that his parents would assist him financially if he should marry before he completed his studies. It is interesting to note that there was only one boy in the lower income group who, during his fantasy period, had looked forward to being a doctor; and quite early in puberty he realized that this choice "did not suit him." The presumption is that he had become aware of the realistic difficulties that faced him and he therefore put the idea aside. The case material suggests that one of the major limitations facing the lower income group is their modest level of expectation with respect to their occupational choice. Certainly they would encounter increasing obstacles in seeking to realize vocational goals which require a long period of preparation and economic investment. However, many of them might be able to overcome these obstacles if they were deter- mined to do so; but frequently they do not even consider it.1 Another interesting observation is that It might appear that children from upper income families have almost complete freedom in making an occupational choice, while those from a lower income group are very restricted. However, society places a high evaluation on some occupations and a low evaluation on others, and these ratings exercise an important influence on the choices which individuals make. In this way, children from upper income families are actually limited. The son of a doctor will not maintain in late adolescence and young adulthood a desire to become a carpenter because of an early and strong interest in and a capacity for woodwork. He usually transforms this interest into a hobby while he seeks a career that promises greater income and prestige.2 llbid., pp. 152-155. 2Ibid., pp. 134—135. 46 During the decision making process, occupational choice is to some extent dependent on the constraints Of the market place. These constraints are more formidable for some than for others. Females and certain minority groups may find it exceedingly difficult to gain entry to certain occupations. The individual making the occupa- tional choice then must weigh his own tastes, values, and abilities against these constraints, mapping a course toward that occupation which seems most promising or satisfying. One of the very important constraints, alluded to above, is the need for means to finance long periods Of education. and training for some professions. Many persons, otherwise perhaps qualified to become professional workers are con- strained by the lack of the wherewithal from embarking upon such a career. Doubtless the constraining influence of the market place grows in intensity with the passage of time. By the time a male is fifteen it is too late in life for him to make the initial beginning on a baseball or football career. At thirty he is too Old to begin the initial training for a career as a surgeon. There is, on the one hand, the problem of age. The prime age of a baseball player is such that training must begin at an early age. For sur- geons, it is difficult to gain entry to a training program after a certain age. On the other hand, for the surgeon, which requires a sizable investment in time and human 47 capital in the person undergoing the training, the expected return on the investment is greatly diminished with the passage Of time because the productive time left after training is shorter than if the training had begun upon graduation from high school in the late teens. Then there are occupations which have virtually no entry constraints and require very little time and human capital investment in training: gas station attendant, hospital orderly and the others which have been designated in previous chapters as disadvantaged occupations as well as some like manufacturing operatives in low wage, scab industries, like apparel and textile manufacturing in the South. In summary, then, we have on the one hand the family unit making decisions about consumption of goods and ser— vices and the supply of market labor. On the other hand, talents and abilities of this unit pertain to individuals in the unit and not to the unit as an entity. Decisions may be made by the unit as a whole, but they are implemented by means of the acts of the individuals in the unit. Another element in the decision making of the unit takes account of the abilities and earning capacity of the individual members. Where the unit consists of the single individual, this dichotomy is unimportant. But in the case of the multiperson family, the skill mixture of the group does add this additional dimension. 48 The idea of an occupational hierarchy according to societal evaluation was mentioned above in the quote from Ginzberg. Most rankings of occupations according to their standing in this heirarchy would probably place profes- sionals and managers near the top. White collar workers such as office supervisors might be placed somewhere near the middle of the hierarchy while many manufacturing Operatives would be near the bottom. Also at or near the bottom would be most of the routine personal service occupations.l 1A search of the literature on status and educational requirements for occupations has revealed that the occupa- tions herein designated as routine personal services are .generally believed to possess below average status. Two Of the routine personal service occupation groups have been explicitly singled out by the Manpower Development and Training Administration as possessing low status. ". . . Workers are unwilling to enter or remain in jobs characterized by low wages, lack of occupational prestige, unpleasant working conditions, and limited opportunities for advancement. These factors have been chiefly responsible for the widespread shortages of such workers as hospital attendants, household employees, restaurant workers, and laundry workers." (Manpower Rgport of the President (Washington, D. C.: Government Printing Office, 1967), p. 150.) For more complete information on occupational status see Duncan's ranking of all Census of Population occupational categories in Albert Reiss, Jr. and others, Occupations and SOCial Status (New York: The Free Press Of Glencoe, Inc., 19617, pp° 263‘275- The training and educational requirements for most routine personal services are also below average. The major exception is for shoe repairmen where the training requirement is far above average, but the educational requirement is below average. Certain occupations within the broader groups, such as governess, may also have high educational or training requirements. These probably are of relatively minor importance in the occupational group 49 The eXpectation, then, is that the choicest occupa- tions are staffed by persons with outstanding ability or from a high income group with the wherewithal to finance a long training period. At the bottom of the hierarchy are those unable or unwilling for one reason or another to compete for the better occupations. In other words, it is to be expected that disadvantaged persons are employed in those disadvantaged occupations known here as routine per— sonal services. These are occupations where the barriers imposed by sex, age, color, appearance, personality, pre- vious background and others Of like kind are likely to be least imposing. These occupations may be thought Of as the last resort for some persons, or as the only resort. in which they are categorized. Information on educational and training requirements for occupations is published in Manpower Administration, Selected Characteristics of Occu— pations (Physical Demands, Working Conditions, Training Time) 1966, A Supplement to Dictionary of Occupational Titles, 1965 (Washington, D. C.: Government Printing Office, 1966). CHAPTER IV PRACTICAL PROBLEMS IN MEASURING UNEMPLOYMENT The general theoretical considerations given in the previous chapter are helpful for knowing what factors should be represented in a statistical model. But in the model utilized in the present case, the variables that have to be employed for representing unemployment are de- ficient in several respects. This chapter is a detailed examination of some of the more serious of these deficiencies. Assume that the oval shaped diagram represents the total population. We may divide the population into these NLF-HP ( NLF‘O ' MEM AO RPS Fig. 1. 50 51 groups. Those persons who are in the labor market but unable to get a job, UE; those persons who have left the labor force out of discouragement at being able to find a job, NLF—D; those employed in routine personal service occupations, RPS; those employed in other disadvantaged occupations, OD; those employed as manufacturing operatives, MEM; all other employed, A0; those not in the labor market but engaged in home production, NLF-HP; and those neither in the labor market nor engaged in home production, NLF-O. It is generally accepted that high unemployment indi- cates that relatively many Of the labor force cannot, because Of institutional rigidities that do not allow wage rates to fall to a level to absorb their services at the market wage rate, or else lack Of marketable skills on the part of the unemployed, find jobs. In either case in a closed economy it is to be anticipated that there will exist a comparatively large pool of labor that has been thrust toward the least attractive jobs when the unemploy- ment rate is high. If it is assumed that institutional factors are not so rigid as to preclude some functioning of the labor market mechanism, then normal economic forces should come into Operation, manifesting themselves in relatively high employment in routine personal services when the unemployment rate is high. More specifically, a high unemployment rate should, other things being equal, swell the supply of available labor for routine personal 52 services. If there is general unemployment, the general wage rate should be low; if unemployment is confined to disadvantaged labor, then the wage rate for disadvantaged labor should be low. In the latter case, high unemployment should cause high routine personal service employment because of its effect on the supply of labor for routine personal services. But when a general unemployment situa— tion exists, there is also a shortage Of demand for routine personal services; general unemployment may affect the supply function for routine personal service employment, but it affects the demand function as well. Whether routine personal service employment increases or decreases depends upon the net effect Of these shifts. Figure l pertains to a closed economy at a moment in time. Over time, there will be changes in the system caused by changes in birth rates, demand shifts, and all Of the other factors that give rise to economic change. If the system is an Open one, and time is variable, and it should be borne in mind that this study is based upon analysis of 49 Open economies in which there are no serious barriers to the transfer of resources and migration across state borders, high general unemployment seems likely to be accompanied by migration of those with marketable skills to economically more attractive areas. The unemployables will remain, exacerbating the generally poor labor market situation. For much the same reason that high unemployment is expected 53 to be accompanied by a high rate Of employment in routine personal services, through its effect on supply, an increase in unemployment seems likely to be accompanied by an increase in the rate of employment in routine personal ser- vices. A number of complications are encountered in inter- preting the real condition that is indicated by any observed UE among the population. These may be subsumed into three categories Of problems: the real meaning of labor force data on employment and unemployment, the problem of age, sex and race mix on any Observed labor force participation or unemployment rate, and the problem of seasonality in the data. One of the difficulties in the concept Of unemployment is that it takes no account of those persons who have de— parted altogether from the labor force because they have given up hope of finding employment, those designated in Figure l as NLF—D. To the extent that these people exist the real condition that is supposed to be measured by an unemployment statistic alone is underestimated. These per- sons might be in the labor force even though unemployed, if they actually believed that there were any hope Of obtaining gainful employment in the labor market. The debate about this measurement problem has been stated most succinctly by Kenneth Strand and Thomas Dernburg: 54 there are three main hypotheses that have vied for attention. The "discouraged worker" hypothesis holds that when economic activity declines, workers become discouraged and leave the labor force. The "additional worker" hypothesis maintains that labor force participation increases at low levels of economic activity when "secondary" workers enter the labor force under the pressure loss of work by the "primary" worker. The "offset" hypothesis maintains that any inflow of additional workers is offset by an outflow of discouraged workers so that, on balance, the over-all participation rate remains virtually constant, or that at least there is no clearly discernible cyclical relationship.1 The problem here, Of course, is doubt about the effect on labor force participation (LFP) Of entry of secondary workers into the labor market, workers that otherwise would be in NLF. From the aggregate data, there is no way of distinguishing a person in NLF-HP or NLF—O from one in NLF-D. Both are outside of the labor force and that is all that is known about them. Their reason for being outside of the labor force, the fact which would make possible the necessary classification, is not known. Some method of inference must be employed to discover the meaning of an observed LFP. Strand and Dernburg, for example, using time series regressed the employment ratio (per cent of civilian non—institutional population that is employed) and the ex- haustion ratio or unemployment compensation on the LFP ratio to validate both the discouraged worker and the additional worker hypothesis. They found that lKenneth Strand and Thomas Dernburg, "Cyclical Varia— tion in Civilian Labor Force Participation," The Review of Economics and Statistics, XLVI (November, 1964), 378. 55 an initial decline in employment from a cyclical peak results in large—scale discouragement and with- drawal from the labor force. Subsequent declines in employment are met by a smaller decline in labor force participation. As the period Of economic slack grows longer, pressure on additional workers to enter the labor force builds up and this tends partially to Offset the discouragement effect.1 These introductory remarks by Strand and Dernburg serve also to bring out the second type of complication men- tioned above: population composition. First it should be noted that the concept of LFP until March 1967, applied only to the 14 years Old and over population. (The age limit is now 16.) In other words, persons under 14 were, by definition, neither employed nor unemployed. Accordingly, the LFP rate for the entire population is functionally related to its definition since it can vary solely on the basis of the number of persons under 14. This matter of definition poses no problem since the LFP and UE data pub- lished by the federal government are designed to pertain to the population 14 years old and over. The meaning of Figure 1 above is, accordingly, modified slightly and pertains only to the population in the age group 14 and over. Eliminating this minor problem by definition may have solved one rather small problem of concept, but there remains a host of other complicating factors that stem from population composition. The sex and age make-up Of the population may have an important bearing on the overall LFP lIbid. 56 rate. Strand and Dernburg, for example, based upon some preliminary research results, found that, when the data are classified by age and sex we find, as expected, that the discouragement and additional worker effects are strongest among the female population and the very young and very old males. The older population is distinguished from the younger population in that while the discouragement effect is equally strong, the additional worker effect, as measured by the exhaustions ratio, is not. As anticipated, we find that labor force participation among males between the ages Of 25 and 54 is less elastic with respect to changes in aggregate employment than is participation in the other groups. It is not, however, true that labor force participation among these adult males is autonomous. All groups of all ages and of both sexes succumb to both the dis- couragement and additional worker effects.1 Besides this divergence of functional relationships between different groups of the population, there is a matter of characteristic differences for LFP and UE rates between different population groups. The numbers in Table 1 reveal substantial variation in both LFP and UE rates among the age groups, sexes, and color groups of the population. There is clear evidence that the female nonwhite LFP rate is above the female white LFP rate, while for males the white LFP rate is above the nonwhite LFP rate. It will be seen also that teenagers have relatively low LFP rates and high UE rates. The point here is that population composition can have considerable bearing on the overall labor force status lStrand and Dernburg, p. 391; but see Jacob Mincer, "Labor-Force Participation and Unemployment, A Review of Recent Evidence," in Robert Aaron Gordon and Margaret S. Gordon, eds. Prosperity and Unemployment (New York: John Wiley and Sons, Inc., 1966), p. 81 for a critique of their methodology. r the f‘ . ~ A K... Total TABLE l.—«Employment status by age, 57 H {\e—I «'11) (x) 0‘11. ‘1‘: mm (“J .2) ‘1 fr 1 K 1411“) fix.) . 4 ‘ d? C) \O (\lr-‘Lj ; («‘Vl V'j-lv— (. jfh ( .’ «“3 4—4 4.1: l ”J r-i'f“ - :T 1?. u “xxx; o l"\ \O o {‘x N o d .‘, . : U, . ”1., 4 . A Aufi A An“) " “n‘. ' “1“ T 5 J " ‘lf‘ " "l’ ‘ ' .1" L1,“ '3 L ' f\- 3') J "J (I? ("’1 I -1 (7‘ ,.4 4 L- t v v 7 n‘. , r. Ll\ Ll f ’ 1V ("1 ‘x \A") IU 5', r.‘ I ‘ .V ‘5‘ t'» (‘l 11\ . 4 (“xi (TA \1': v . 4 r 4 ”I: A A 0 A L1 \ _ {if .3 r j v1] u“, :74 (‘2) {[2 ("‘1 ("W ( ~. (v'~ ‘Vf‘ _‘_' 1;". f. , 4 5,‘\ (d, .31 ‘1') (‘~.! "Man 5"» .”~ H \«;‘v X 1? " I» r. .4 .74 :‘x.’ , i I . J‘s (V) ,—4 I r-‘Il I x- .' ' 1:»: fl 0 (h v{ ' (I l\ - ,A( U - K*~\\J 0 L1 _.,’ o A A V\ A Ag..- 0 RC") fl '_1 A A f A Af~ A An) A Au\ (, I (1) *1?) (1g I‘“- r 4 r 7“ r" f ‘1) . ‘_ (‘ r "(I r—4 O . , I: KL) r LIME) \f‘o 3;“ ( , ' I " f . r "(3 LIN 4 IV ‘0 u') 1 it" I“ ’1“ n" g 1’". 1 I i J O_) A A A A A A A y, a. A MC) -.' H r“ Ki: '1‘ r“ .' I v r—‘ (*9 H P4 C\ r t (\- (‘d 1 Lfl W“) (‘0 \0 1"": ' :1 :3 3' I“) 4.? WWW ' (‘r ""71 f“ ‘L‘ .1'\ ? l l r 1 (T y..- r-I i" .1 —.’ " Y‘) ('7‘: (Y1 (3‘. n' ‘1‘ - "I ' ’7 WA“ - '4'? - L.*( ["1 - V. "g .\ .—4f~~ <2) G Q'L) 0 A o O a | I fiL _ . ‘A ‘ r . :l .\ . :u‘ .g\ . \.7 ‘ I 3‘» (“J .—4 '7" f“ 1'15? I t L. A A 'W " '~' — A A-‘f A Ah. I; :4 ”4 W I "4 3 f‘ f I, .v‘, ,7 l " r r-4 (\3 Nut Hf wr -:~ :f‘: M“ :ILP. " " " " " " " (. '1 C~ l. l t\ i: C C H .—: . . : e4 . a .. E) r O i , i j ,_4 ,1 91-4 1) ~20 (J. .3 ~. 4 r) I.’ ’4' - ‘1 I) V t.‘ I‘ :7 -' U (1) 5| 1 3 _ . r4 :7 I. I.) f “ T *3 ‘17 . [‘ ‘-- . u\ x . I I ‘ ‘x E: u‘ _ 1: ('VW '» ’ . . J J 4 . _ ) w .3. . , - up ._‘\ - ‘0 r» . ._‘\ 1 - :-, x) T) - 1p )1) - .‘ - ("'I‘C‘ - 1. 0 “1 ‘ 0 “‘11 0‘ 0 V) 0‘ A (I ' A A V] A A (V A A|A\ A Ar)\ (r .74.: 74 .3.‘ I) (M J-V) 1‘1 7-.‘ .o * ‘. «r r-r "0 J :3. .' ~ 1 (T3 1“; . 1 1‘] "-1 f, 3‘ if, 7 -: r , r4 Ax m . ’1 0413. W (k. '3 1’") '5; r4 . .) f? 4 p, (D A A A A A A A 4‘ . a a f) \‘x '\ I.) o . . '* . a j ' 4 ('4 i. . a ' f") 'v .-. i4 ") LEW J ‘1 ll ~ '1 1‘ ' . H 4 57 "4.) Fl ‘1" l J ”V2"! 1 t J “A ."» '7" I) I) ,O \i) P " A" . if» M. 1r "1f 1‘) ,—4 r» ”'3 (1 I . ,7 (‘~. 1 y 'r '7‘ y u} [.113 o ‘L J‘. a l” (4 o ‘1 1. 0 l v . 3’ ‘ o Hfl o 5‘ 7'7, 0 y] 0 " ) a " ' G 0 :1 fl 0 - A A1 A A s) A A _') A Agf {-1) I) ' ’ -1 i .f ' I ‘ l I) ' 4' 4 01:“. )‘ ‘i’ 3' ‘ t: u 7 ’. l‘~ I .‘\ I “ "' V ' " ’- ' l ' I ‘f A A A A I- A L) m ‘A. 44‘. ' (v) [‘~ ,. I) ‘ ~ _ l ‘ ‘1 ‘ ‘1’” ' fl ‘ . g I: t 1, ' ‘ u ' 7 L . ' , _ t f _ ' - , - 7 - I a f - t - T _ 0 T . - * ‘ - _' . ' ’ . _,° J . A A:-.. A A , A A f A A ‘ VA A “ A A " A A 7 A A '- l.\wl J »' " .\. ‘ I . {‘~ ') l . A “A v , ') 1 1'1) ‘ ' .‘l ) ‘ it ’-’: ') T »'7 ,1 1 L_‘ 1' _ 1 1' T - . -\ 4 r’ I r A A A a UN (4 If I "‘J l“ ' \1 r4 ‘1 4 "f- _) ' 7. [A 7‘ - ' L, . D ’ — ‘ p) ’ - I .V [‘~"'C) “an“ 3-1 .1 , L‘I z"""." 7 X" "‘1 7‘ " r'l tl’i o ,—4 7) c in j . 3 lg, . p s . ~ . . 1» 4 . 3 fr, . 0 A-l‘) A A J A ANN A A. c A A j A A l A A j A A :3 \L) " I ~ 13\ r» " — g) "'1 3' —— 5? n; 4 " x ("j . r z‘ .T V) v‘ ( M) ""3 r t“- ‘ I, " " S) ; _” (,1 7‘ ‘. ‘~ Q a.". V '. ‘ .4 V. ‘ H 1‘ A A A A A A A A u a n .. “9W ‘1 1' 1' ‘ ~. ‘. 4 4 r T (J “4-: i,\ 74 : , t d) 0) (1) (1) 2c :0 :4 < “1 ') 1') ) ’e e "E‘ '3‘ 11 I Y) O n Y‘ 3 Force *4 "U I.) or-t "3 I «<1 ’1 if) H T) 2.. L4 '. L. u 1.21:. t.) 11.1. p '5. 21. a) :9 f: C) P‘ .1. ") m (U (I: J" '5 :3) p ’76 (I) D A) _Q 5) ‘2 r4 2. a) r-4 L H L. «a H :4 c1 4.) 7,1 3* .4 -3 1m?) with": 03:00 1'13an .1171 p.111 (9.171 337‘ Q. "-4 D. >. I.) H O. {a :3 P; Q. i: .D Q) -.—4 r1) 1.) ”p 3 (1, ¢'0 “~TA’ .C~?95fl m ~*’>fi “4’ C=33‘M O O are C::>n ~+::> L.C A F) C14 (1) b] J») 3 Q—I Q) »] b) 6: L14 Q) - l J») (:I '51-: 0) >—] A) p 111 (W (1) :{ L :7) '1) IL: v_" 'Vx 1) \/ "1 I) Q) A :5 al' C :>. (0 F: :5 :0 7’: >3 :8 H on :4 .3 c: -.4 c .3 5: mi . 4 u 7‘: ~4 w .1 ,C 7—1 '—+ :1: O .—1 HI .1: .—( r—l (I; »-l r4 .1: ,g; {1. c (‘3 —4 11. '4 '74 .‘l. 3*. 4 :1, 7'. is: as .7 (U .. n J m (1) rd: 33 <1) «3;: m 7.: H :‘E .1: :2: "—4 -3 '1 <1» .1. :‘z z‘: n.) «i i .‘12 1) r—i &) D. .LJr—4lJLL -—4.-.>r—41J.L r-«IJ.41)Q.. :. :5. Q) ,_.4 ;~. I 4:. ’1, a) O ”J .z. <1) '1) 0 Ex. :8 ‘ '3 ex. :0 O ’3 EL. 1) H L: (11 .1» H (I 1:} 3‘. "-4 :1 1:] " ~H : Ix} ,. 94 e: .1 :4 ~« E« .1 2: -4 -£‘ .4 z: is 94 ,3 1. -) ~n :1 .4 r) r» *5 s: -» :4 ~) 7 '~ -- ~w EU ”3 l) 0) 3', .11 C) U 7:: .L If... It. '1‘. ‘2: {14 fr. 58 of the population. A population with unusually many adult males will have, other things equal, a higher LFP than one with unusually many teenage females. Another serious defect in the unemployment statistics used in this study stems from the fact that the rate of un- employment in many states may be subject to seasonal varia- tion. The data on employment and unemployment published in the Population Census pertain to April 1950 or April 1960. Consequently the rates for some states may be unrepresenta— tive of the underlying unemployment situation. It seems reasonable to expect that the employment status of people. and consequently the industrial unemployment rate of a state would be much more transitory than a state's industrial structure or the share employed in a particular set of indus- tries or occupations such as routine personal services. Therefore while there are grounds for being reasonably con— fident that Census of Population employment rates in various occupations or industries are fairly accurate for depicting basic structural differences existing during the 1950's and again during the early 1960's, one is hesitant about placing the same degree of confidence in the unemployment rates as indicators of corresponding unemployment conditions. Inasmuch as unemployment data are compiled by state agencies in connection with their Bureau of Employment Security activities, it might be thought that their data, rather than the Census data should be used here. While it 59 is true that using annual averages published for states by these state agencies would ameliorate the seasonal problem, there are a number of reasons for avoiding these data. I. They are not available for 1950 for most states and for a few states in 1960. 2. Their reliability in 1960 is questionable.1 3. The state averages are for total employment only; there are no disaggregations. M. Corresponding labor force participation data are not compiled by state agencies. 5. They are inconsistent with corresponding employ-. ment data or employment data from other sources in that they are concocted.2 I See, for example, Joseph C. Ullman, "How Accurate are Estimates of State and Local Unemployment?" Industrial and Labor Relations Review, XVI (April, 1963), pp. D3D-US2; and John H. Lindauer, "The Accuracy of Area Unemployment Estimates Used to Identify Depressed Areas," Industrial and Labor Relations Review, XIX (April, 1966), pp. 377—389. 2For the concoction formulae see Bureau of Employment Security, Estimating Unemployment (Washington, D. C.: Bureau of Employment Security, March, 1960, reprinted April, 1961). CHAPTER V FORMULATION OF THE STATISTICAL MODEL It was observed in Chapter I that one of the attrac- tions for utilizing areal cross section data for testing the theory is that these services are likely to be consumed by the residents of states wherein they are produced. But this element is also a disadvantage in the statistical model in that many of the data utilized pertain to both consumers and suppliers of routine personal services. The median income figure for a state, for example, is based upon a distribution of persons or families which includes these suppliers as well as demanders and there is no method for unentangling the median income statistic for one group from that for the other. Another disadvantage of the state by state data is multicollinearity among some of the variables which the economic theory indicates should appear in the independent set of equations. Several earlier attempts to test the statistical models revealed, for example, that wage rates for different industries are highly correlated among states. The wage rate for females in domestic service is highly correlated with the wage rate for females outside of domestic service, and both of these wage rates are highly correlated with the other wage rates in routine personal 60 61 services. Thus, although economic theory would indicate that the wage rate for the particular routine personal service occupation being analyzed should appear in the demand relationship along with the general wage rate and other relevant variables, as a practical statistical matter this cannot be done using the data employed here because of the high correlation between these two variables. Notwithstanding that these problems preclude specifi- cation of an all encompassing statistical model for testing the theory, it seems worthwhile to set forth a theoretical model drawing on the ideas presented in previous chapters,, and then to modify that model so as to make it adaptable for statistical testing. Supply Given that the routine personal service occupations are among those which are ranked low in the occupational hierarchy in terms of status and pecuniary reward, it seems rather obvious that as a general rule the supply of persons for employment in these disadvantaged occupations will depend upon the size of the labor pool which is disadvantaged in some sense and cannot compete for the better employment Opportunities.l Who are these disadvantaged? First off there are those who are disadvantaged by virtue of their l"Disadvantaged occupation" is not a new term; in its Handbook for Leaders, the National Committee on Household Employment refers to the private household occupation as a disadvantaged occupation, p. 31. 62 ancestory, namely nonwhites and foreigners. Second there are those who are disadvantaged by virtue of their unem- ployability for other reasons such as lack of a marketable skill. Third there are those who are forced into the labor market because the primary breadwinner is unemployed; those are among the so—called secondary work force. These groups may be thought of as constituting the supply of disadvantaged labor for all disadvantaged occupations, those in routine personal services as well as those outside of them. It may be expected that the availability of suitable alternative employment opportunities will diminish the supply of disadvantaged labor for routine personal service occupa- tions. Other possible considerations in a generalized supply function are the level of affluence of these disadvantaged persons and the prospective wage rate in routine personal service occupations. A very general supply function which embraces these factors is: Srps = a1 + a2 NON + a3 FBW + an UE + a5 SEC + a6 ALT + a7 S-AFFL + a8 er8 (l) where S indicates that this is a supply function rps of labor for routine personal service occupations NON indicates the attribute nonwhite FBW indicates the attribute foreign born white 63 UE indicates the level of unemployment SEC indicates the attribute secondary worker ALT indicates alternative employment opportunity for disadvantaged labor S-AFFL indicates the level of affluence of the disadvantaged or potential suppliers of routine personal service labor w indicates the wage rate in routine rps personal service occupations. Inasmuch as the analysis in the next chapter utilized per- centages or rates for many of the variables, equation (1) should be interpreted similarly. Thus NON pertains to the percentage of the population that is characterized as non- white. FBW refers to the percentage of the population that is foreign born white. UE indicates the unemployment rate. ALT indicates the percentage of employed persons that are employed in suitable alternatives. In S-AFFL and wr may ps be taken at this Juncture as dollar variables rather than rates. S—AFFL might be interpreted as median income of the disadvantaged while wrps is the market wage rate for routine personal service occupations. SEC acquires operational significance through further specification of the model which involves disaggregation of the supply function into its male and female components. Hence 6A F—S = b + b NON + b rps l 2 FBW +bu F-UE + b M—UE 3 5 + b6 ALT + b S-AFFL + b8 w (2) 7 rps’ where "F-" or "M-" indicates female or male as the case may be, is a generalized supply function for females for routine personal service occupations. The vague SEC variable of equation (1) is replaced by the variable M—UE in equation (2). The idea is that a high male unemployment rate forces females into the labor market as secondary workers or breadwinners, contributing to the supply of female disadvantaged labor. The variable F-UE indicates the condition of the female labor market itself. As was observed in the preceding chapter, some unemploy- ment may fail to get reported as such in the unemployment statistics. PeOple may simply drop out of the labor force altogether after prolonged unemployment. Consequently some unemployment may be hidden in a low labor force participation rate, producing F—Srps = C1 + c2 NON + c3 FBW + CH F-UE + 05 F-LFP + c6 M-UE + c7 M-LFP + c8 ALT + c9 S-AFFL + clo wrps. (3) The expected signs of the coefficients are: c2, C3, C4’ C6, C10 greater than 0, 65 0 c8, c less than 0- 7’ 9 An additional reason, apart from the one just given, for having M-LFP in the expression is that a low male labor force participation rate, whether it represents male unem- ployment or some other cause, would, other things equal, make it more necessary for women to enter the labor market. For males, the supply function is simply: M—S = d + d NON + d rps l 2 3 FBW + du M-UE + d 5 M-LFP + d6 ALT + d 7 S-AFFL + d8 ers (A) It is not appropriate to consider the condition of the female labor market as a factor in the supply of males for routine personal services inasmuch as males are the primary bread- winners. This is not to say that the same factors that relate to the condition of the male labor market are irrele- vant to the condition of the female labor market. It is to say simply that the situation in the female labor market is not a determining factor on the male labor market condition in the same manner as the situation in the male labor market influences the female labor market. Demand Unlike the supply for labor for routine personal ser- vice occupations, which can be considered as two separate functions, one for males and the other for females, demand functions for routine personal services can be expected to 66 vary with the nature of the service. There is, first off, the consideration that outside of the private households industry, where the consumer of the service is also the employer of the labor, demand for routine personal service labor is derived. In barber shops; beauty shops; shoe repair shops; and laundering, pressing, cleaning, dyeing, and garment repair establishments, labor is combined with capital and other inputs to produce a service. Consequently, the value of the marginal product (VMP) schedule for the industry is the true demand function for routine personal service labor for these industries, where the firm or establishment may be thought of as an intermediary between the raw labor input and the final consumer. The data utilized in this study preclude estimation of the VMP schedule for these routine personal service industries, hence it is simply assumed that the demand for labor for routine personal service occupations is a direct function of the demand for the routine personal services themselves. Abstracting from the firm as intermediary, the following generalized demand function for routine personal service labor may be anticipated: D = e + e D-AFFL + e P + eu PRICES 2 3 rps + e 00 + e6 CHAR (5) where 67 D-AFFL indicates the level of affluence of the potential purchasers of routine personal services Prps indicates the price of routine personal services PRICES indicates the prices of other goods and services OC indicates the opportunity cost of employing a family member to do the housework that could alternatively be accomplished by means of purchasing routine personal services CHAR indicates special characteristics pertaining to particular routine personal services, to be explained shortly. Inasmuch as the firm is assumed away in the model, Prps may be considered as being identical with the corresponding wrps which appears in the supply function, equation 3 or A. The theory underlying the introduction of the variable OC, it may be recalled from Chapter III, was elucidated by Mincer. He defined the market wage rate that the housewife might earn as the opportunity cost of employing her in the home. Along with this wage rate, he had in his demand function for domestics the wage rate for domestics. Subsumed under the catchall CHAR is a variety of factors, none of them relevant for all of the routine 68 personal service occupations. Let us consider first the demand for females in households occupations. There are, in reality, two components to the occupation group, females in private households occupations. They are divided accord- ing to whether they are live-in domestics or live-out domestics. The numbers in Table 2 reveal the order of mag— nitude of the respective rates of live-in domestics (F-HHOin) and live-out domestics (F-HHOout). There it will be seen that in most states, F—HHOin is very small as compared with F-HHOout. Moreover, as a matter of fact, there is a high negative correlation between F—HHOin and F-HHOout in the data for both 1950 and 1960. In the particular demand func- tion for F—HHOin it is necessary to allow for this sub- stitutability for domestics living-out, consequently allow- ance must be made in the model for F—HHOin for the wage rate for live—out domestics as well as that for live-in domestics; and similarly for the model for F-HHO out allowance must be made for the wage rate for domestics, living-in as well as for domestics, living-out. Another consideration that would seem to be particu- larly relevant to demand for all private households labor is the labor force participation of females. Female labor force participation in and of itself is a source of poten- tial demand for both male and female domestic labor to replace the foregone home production of the employed female. Now it may be recalled from equation (3) that TABLE 2.--Female rate of employment in private households occupati' households occupations, of Columbia April 1: 69 living-out, by state for the contermincus 1950, 1960 and 1960 as a per United 5 cent 7‘ L‘ ’A-\) 7 «'113r- ALIA—L. TH’ -/'.A\/ of .s-in, cs firL/x $3,)Uo and private and the District F-tfifilout 1960 as C C) A—le State 1950 1960 per cent 19,“ 1960 per cent of 1950 of 1930 Mississippi .;3 .15 77 1?. 9 ??.?1 116 Alaoama .AC .30 C? lo #9 16.36 99 Georgia .“3 .32 75 L” ”L 15.91 )1 Louisiana .hd .5; 65 Lt.w: _%.71 10a South Carolina .H7 .33 71 ;t.'. 16.3” 102 District of Columbia 1.3“ .au 71 9.7% 9.55 101 Florida 1.10 .70 6A 1“ no il.f3 61 Arkansas .57 .Al 72 12.ofi 13.3 11” North Carolina .78 .ul 52 11.9% 11.57 99 Texas .bi .60 72 1:. r 10.?s 93 Tennessee .79 .5" *9 1; ‘fl ;;.?3 92 Virginia 1.U6 .7“ by 1;. 7 13..» P5 Oklahoma .71 .56 51 7.;5 5,7w 5% Vermont 3.1m 1.11 :2 7.7‘ Y )j 103 New Mexico .13 .0‘ i1 €.,; 9 ;1 1 Arizona 1.35 . 3 i: 7 t) '..J ”i Delaware 1.61 .:3 S; -?.'" 7.." Kentucky 1.ug .22 Cl v. 4 5.fl3 ‘- Aaryland l.‘; .K‘ 66 1'. * 7.fi5 4 Nevada .JC .34 a) . 3.20 1" Maine «.12 l.-i ;5 . :.91 - Washington 1.22 .t- :0 H.ni 6.f3 l~i California 1.:2 .g- to h “i 5.13 107 Kansas .91 .3h ct L o.”n 11% Colorado ;.;7 .2» ~o . : ' ;. 1'5 South oakota 1.'7 .cb t; -.'7 i j“ ‘ } West Virginia 1.t5 L.tj t; T. - .2; . Oregon 1.23 .oo ; r l; .. Wyoming .,3 .Ho ;j . E New York 2.;‘ -.j3 if 3. Idaho .70 .;L i: . 7.i; Montana .72 .‘u 53 w. ; i 1 Nebraska 1.02 .u~ 6? i r 1 Michigan 1.51 . i S7 .4' . L 5 Missouri 1.ufi .uo 31 P.19 I 2 Indiana .37 .53 =7 3.37 2.3L 1 a North Dakota 1.25 .33 66 ~.i: 7.2} 177 Iowa 1.22 .71 E3 9.1» 6.3} 1;? Ohio 1.2“ .71 5? 5..b 5.33 -3. Pennsylvania 1.36 .7? 53 ‘..l L.“” 33 New Hampshire 1.5M .59 5% 1.23 d. 3 11‘ Minnesota l.U5 .79 55 5.1: 5.15 i~ Connecticut 1.97 1.56 69 3.95 3.’t New Jersey 1.35 .85 61 9.- 3.. .Illinois 1.16 .63 5A 3. a 3. , ~ Massachusetts 1.16 .53 A5 2.fo 2.7“ 1 Utah .5‘ .25 ug J.K€ A.L Rhode Island 1.28 .67 53 2.43 2.6 i v Wisconsin 1.25 .65 54 3.”3 5.5. 1. Source: Derived from ~. Bureau of the Census, U.S. Census of “woulation: 1960. 7O F-LFP was introduced as companion of the variable F-UE to indicate the condition of the female labor market. It seems more appropriate, however to consider the female labor force as primarily secondary, and the condition of the female labor market a function of M—UE and M-LFP along with other variables indicated in equation (3) but excluding F-LFP. Still another important consideration in the demand for labor for direct employment in private households is the composition of the population with respect to single persons. It is expected that the demand for private houses holds labor will vary negatively with the percentage of the population that consists of unrelated individuals (URI). It seems reasonable to posit substitutability between live-in and live-out domestics. It does not seem reasonable to posit substitutability between males and females in pri- vate households occupations. Thus the demand for males in private households occupations is not hypothesized here as functionally related to the female wage rate for domestics. Notwithstanding this nonsubstitutability, it does not appear to be unreasonable to anticipate a positive relationship between F-LFP and the rate of employment of males in private households occupations. A family in which the wife is employed outside of the home is a likely source of potential demand for, say, gardeners and other occupations included in private households that are staffed primarily by males. 71 It is difficult to conceptualize a theory of demand for labor for those occupations that are in the private households industry but not among private households occu- pations (see pages 7 and 8 for this distinction). The diversity of occupations in this category is wide. Because of this difficulty, it is hypothesized that fOr males or females for employment in occupations in the households industry but not in those classified as households occupa- tions per se the item CHAR in equation (5) subsumes only URI and F-LFP. For the group of occupations included under the head— ing laundry, cleaning, and dyeing operatives (LCDO), there seems likely to be a high degree of substitutability between male and female labor. Accordingly, the wage rate for males and females both in these occupations should be reflected in the demand functions for both males and females for these occupations. Another kind of competition for LCDO, male or female, is specialized equipment in households which washes and dries washable fabrics with relatively little labor input on the part of the user. As a matter of fact, using data from the 1960 Census of Housing to calculate the percentage of occupied housing units with washer and dryer, it was found that this percentage is highly corre- lated, negatively, with the LCDO rate of employment. Unfortunately, similar data for 1950 are not available, pre- cluding the use of this variable in the model for LCDO. 72 Other variables that have been selected to represent special demand factors in the LCDO case are F-LFP, percent- age of total employment classified as white collar (T-WCOL) and percentage of employment that is in the lodgings and restaurant industries, (HANDE). T-WCOL represents a specialized demand for laundry and dry cleaning stemming from the general requirement for a neat appearance in these occupations. HANDE represents another specialized demand for laundering and dry cleaning, not only from waiters and waitresses in restaurants but also from the linen require- ments of restaurants and lodging places. For barbers, specialized demands may be expected to derive from male white collar workers (M-WCOL) and from male waiters (M-WAIT). For beauticians, these particular sources of demand are female white collar workers (F—WCOL) and waitresses (F-WAIT). For both white collar workers and waiters or waitresses in restaurants, appearance is impor- tant, so that in addition to specialized clothing care requirements, they also register specialized demands for the trades involving cosmetology. Similarly, the demand for the services of shoe repairers can be expected to be related to HANDE and T-WCOL variables. Obviously a number of factors that were discussed in Chapter IV have not been introduced for consideration in the models presented in this chapter. There has been no 73 consideration, for example, of the value of time per se and the effect of the tire input required for the purchase of routine personal services associated with cosmetology. There has also been very little consideration of the sub- stitution effect among routine personal services. Only in the case of substitution between F-HHOin and F-HHOout and between males and females for employment in LCDO was the matter of substitution considered. But it is obvious that domestics are substitutes to some extent for the services of laundries and dry cleaners, and that within the group of industries designated as laundry, dyeing, and cleaning there is competition among types of establishments. Laundro- mats, for example, compete with power laundries. The primary reason for ignoring these other relevant factors in the analysis is lack of good data. Statistical Considerations Another reason for not introducing additional factors, or marginal theoretical value compared with those already considered as most germane, is that the regressions simply became bogged down with too many variables. Not only are critical degrees of freedom lost with too many variables, other problems such as multicollinearity cause additional difficulties when the regressions become excessively loaded with variables. Indeed, the equations specified thus far suffer from this difficulty. 74 Before discussing multicollinearity, it seems worth- while to bring up the matter of a simple data limitation which affects the variables designated as indicating affluence. This problem was alluded to at the outset of this chapter, and derives from the fact that the state income data for families pertain to all of the families in the state: those who are suppliers of routine personal services as well as those who are demanders of them. For— tunately, a very nice method for avoiding this problem altogether exists. Stigler, it may be recalled (see page 22) observed that income distribution is‘a relevant factor. in both the demand for and the supply of domestics. Since Gini indexes for states may be calculated from available data it is appropriate to include Gini index (GINI) as a variable in the demand and supply functions, removing D-AFFL from the demand equations (equation 5) and S-AFFL from the supply equations (equations 3 and A). Introducing GINI creates other problems for NON is highly positively correlated with GINI and FBW is highly negatively correlated with it. Several preliminary attempts to get significant regression results for particular routine personal service occupation groups suggested that FBW is a relevant factor in the supply of female domestics, living—in while GINI is the relevant factor to include in the other regressions. Accordingly, for F-HHOin, two models seem worth testing: one with FBW as a supply factor and one with 75 GINI as a supply factor. For all other occupation groups, GINI is the appropriate variable in the supply equation, and NON, FEW, and S-AFFL need to be deleted altogether from the supply equation. Parenthetically, it seems worth point- ing out at this Juncture that the high correlation between GINI and NON suggests that the presence of nonwhites is a major determinant of income inequality. It may also be worth observing that FBW is fairly highly positively cor- related with median family income while GINI is fairly highly negatively correlated with median income. Another source of multicollinearity is found in the . data on wage rates. In equation 5, P is highly corre— rps lated with DC, the latter defined by Mincer simply as the going wage rate for females. An appropriate transformation involves calculating DC as a percentage of Pr which is, in ps this model the same as wr This new variable, call it pS' PCNR, is in fact a better representation of opportunity cost than the simple market wage rate that the wife may earn inasmuch as PCNR is an expression of the net value of the routine personal service that may be purchased to replace home production. Moreover, it may be considered as a relative price variable. This is so because the general level of prices (PRICES in equation 5) seems likely to be reflected in the general level of wages. Viewed in this manner, PCNR replaces Prps’ PRICES, and OC in equation (5). Now obviously, there is no way to separate the price 76 relationship (which is relevant in the consideration of the household as a consuming unit, abstracted from its role as a production unit) from the wage relationship (which is relevant to the consideration of the household as a produc- tion unit which substitutes purchased routine personal services to release a family member for remunerative labor). But in any case, data for PRICES are simply not available. How does this transformation affect the supply equa- tion? Unfortunately, PCNR is highly negatively correlated with Pr for all of the routine personal services. Con- ps sequently, PCNR must replace Pr in the supply equation if ps prices are to be considered as factors in both the demand and supply equations. One additional instance of relatively high correla- tion is that found to exist between M-UE and F—UE. Accord- ingly, it seems appropriate to drop F-UE from all regres— sions. These changes and deletions result in: Supply RPS f + f l 2 M-UE + f 3 M-LFP + f4 PCNR + f5 GINI + f6 ALT (6) Demand RPS g1 + g2 PCNR + g3 GINI + g“ CHAR (7) except, as noted above, for an alternative model of F—HHOin in which FBW replaces GINI in the supply function. CHAR, in equation (7) is a general term which may be interpreted as 77 referring to particular factors affecting each of the separate routine personal service occupation groups. The variables selected to reflect ALT are for females, the per- centage employed outside of the private households industry that are employed as manufacturing operatives (F-MOP), for males, the percentage employed outside of the private house- holds industry that are employed as operatives of all kinds (M-OPS), and for males and females taken together, the per— centage employed in manufacturing industries (F—MFG). CHAPTER VI RESULTS OF TESTS The reduced form equations for equations 6 and 7 are: PCNR = 31 + 32 GINI +33 M-UE + 34 M-LFP + 35 ALT + 36 CHAR (8) RPS = kl + k2 GINI + k3 M—UE + ku M—LFP + k5 ALT + k6 CHAR. (9)‘ The coefficients of these reduced forms expressed in terms of the original structural supply and demand equations (equations 6 and 7) are: J2 = _'5"":"'" i 3' 9 J3 = ' ' ' 3' f- : Ju = "" :3f 3 B2 A g2 A g2 A J = f6 3 = “Eu k = g3 fu ‘ g2 f5 5 S2 ' fu’ 6 E2 ‘ fu 2 fu ‘ g2 -s f -s f -s f k =.___§__E3 k = __£__i_, k =.__£__é_, 3 fu - S2 A f4 ' g2 5 f4 - g2 _fu5u u ‘ g2 78 79 Accordingly, the expected signs of the coefficients in the reduced forms are: 32 indeterminate; if f5 is greater than g3 the sign is positive; if g3 is greater than f5 the sign is negative; J3: k2: k3 J“, 35, k“, k5 negative; positive; 36 if the structural coefficient for the particular CHAR is positive, 36 is negative, and vice versa; k6 if the structural coefficient for the particu4 lar CHAR is positive, k6 is negative and vice versa. In other words, the coefficients for M-UE in both reduced form equations are expected to be positive while those for M-LFP and ALT are expected to be negative. For CHAR, the coefficient is expected to be opposite its expected struc— tural value in the PCNR reduced form, and the same as its expected structural value in the RPA reduced form. Table 3 has been prepared to assist the reader in locating the meanings of the abbreviations which appear in the analytical tables that follow and in much of the sub? sequent narrative material. Besides identifying the nota- tion, the table reveals the sources of the underlying data. 80 TABLE 3.--Abbreviations employed to designate variables, and sources o1 oat“. Abbreviation Variable Stlrce F-HHOout Females employed in households occup:tio ns, living-out 3.3. cev‘.s of Pepilltién' as a percentage of tot;1 employed females llht av. 1. . uer:,s f “WIIIit- ' 1QN‘ F-HHOin Females employed in households eccutiiions, living-in -ane as above as a percentage of total employed Iemales F-HHl-O Females employed in households in stry but he" in .are u; r ve households occupations M-HHO Males employed in households occupations as a per— Same as 3D 1e centage of total employed males H-HHI- Hales employed in households industry but not in -awe as at ve households occupations T-LCDO Males and females employed in laundry, cleaning, 3nd its is at“ e dyeing occupations as a percentage of all empl .ei persons ' F-LCDO Females employed in laundry, Clea ining, ani dyeing .are , ;* occupations as a percentage of t;t.al employed fem.les H-LCDO miles employed in laundry, cleaning, and dyeing oceu- 51 e :5 .' 'r pations as a percentage of total employed males F-BEA Females employed in hairdressing and cosmetology weou- Jame a; kbfije pations as a percentage of total employed females fl-BAR iales employed in barbering occupations as a per- Same as 1U It centage of total employed males A-SO Hales employed in shoe repair occupations as 1 lie 3 up 'w percentage of total employed males PCNR For females calculate} on 'he his is of average week-» r r R-Bu' r, :- .', income of femaLes ezrzpl::y'ei wutsL e :f the pri”u:e l-fnl—W, «H , — _., households industry as a {exoent Me of the average H-lif , P—. , j- , weekly income of females in the pirtio tlaz' .ccupzti»n calii-gze: f: m i',- iw group to which the equation pertains. Eor males t'e . . .v'_J;__' ;_;_fi - calculated on the basis of average week . income of ti_n' . l i ‘ .._~w_ 7 all employed males as a percentage of average weekly 3 ‘ id- r fjfv income of males in the part titular u?cu alien grrup to R—B A, J—- u, JLl '— —, which the equation pertains. Ancthe: exception for talt.;i Lin; i=v l‘, .; 7 males was made in the case of n-LCDO where PCNR ir'a 1 pint z uri;;, *r' 5 indicates the average weekly income of all employed snaps, r: amle rigilr n females outside of the private households industry Yr m the sub: ____:; “ :;_ as a percentage of average weekly income of males 11.5 art the ‘7:_;" H in laundry, cleaning, and dyeing occupations. In Bu.ine.s 147:. the case of T-iCDO, the variable PCNR represents average weekly income oi all employed Iemales out- side of the private households industry as a per- centage of average weekly income (weighted) of all persons employed in T-LCDO. For M-h o and H-Hnl-O, PCNR represents average weekly income of all employed males as a percentage of average weekly income of males employed in the pxi ate households industry. GINI Gini index for families and unrelated individuals Calculaied .313; duff r w the 'I.S. 7V?'*;;‘__;‘;_i____— ticr: gfl=J-ni the A Sea: Llw — :11“ 3‘ 1' ‘. 1' along wit] 1- xi-» -—L:Ii. Tax weturns flu57737__fl _‘ Trail}: iual tin: (xv .;K Fva .1 for 1353. F;r me‘lrfiwl I. see Lav id I. .erway, "A Rankitnzzai Stzures by r- equality Using Csnsus ind ' Tax Data” in The MC”lUJ ~L®mQO f“\ 4 / on Adv r» m: OHom mmqlm Hmb mozlm mquE Mali . HXHU mzom pcmpmcoo EmoH .omma "poo wCH>HHIImc0flpmdzooo . . moaocmmso: CH mmHMEmm do pcmsonoEo mo comp one pom moHSmmL COHmmmLmop mmpmsom pmmma mpmhhuozeuu.: mamae households in oyment of females b for the rate of emol ' n results occupations-l greSSlo TABLE 5.-Two stage least squares re g-out: 1980. lvin . PCIO F-LFP URI 0? .1. -.I A F 0.. De. I '1 1:1— "14‘ du :11- I A . L PC Constant Item U9 observations .128 (.uou) 7M6 (1.113) 3* D 09'?) 9.7 (V) i’\ KY“, 1) If (Kl ( V‘D , — 165.u N} PC (a) .020 + (j . - I (T) F-HHOout (b) Ln (Y) (\J :0 H (“J Supply (a) .015 /"\ oh (.133) 91 .020 (.OU3) + .007 (.100) (.m) .:9 + 9 mand m (d) ’ations V ~ V abser 37 A : Ln m (V) A C\ t» m (\J (V) O C I V I I /\ cum 2 “A" Lf\ r‘I no [\1 ;, o o o r J +9¢ + /\ or Q (’1 \‘O , (“V O O 0 T? (Y) \_/ I l 3K ” ‘\ 4! YR it“: Ch f ' D (Y) ‘ .) ”) O O O r‘ i \d’ I I (F) L.” \ ‘ 1'\ ‘7 ’T\ I C) 1; 1'1: O O O (\5 \_/ + + /‘ \ 2*: umw “ ‘1 .' C' 3 1'1. ’1 9., O O U ."v’W ' I \. + + :6: <:‘ ..‘\ D O O I ‘ ,H\ ' D J r—J. fiI \ I + + _\ ~T .‘.\ O I N 73 Q\ (“J H + I 11) 7S (4’) 9:. 7T“. can >‘ ‘ L) I I- . L: A A m M .44, \o.L . ‘J I‘LJ u. D. «v ._J I' v. ~ :4 I (I I \1 I O \_/ I AWN 1 '.~.\ I "VA (A: r 'I I 0 ‘~_ '. N 3;\ v I gr} I J I O _. ‘V I I I O ,l V W 1"} O ' I .' . "‘ ‘ .3 ti . O O I I I J 2K F.‘ l -—T 1 1 » "'J l "3 O I C J v .. K O ‘ J W? ‘4 kv H. 11) . H) ,C ¥j £92 .thm- use- and mx mgr mm mochwflnmcmflm mmpmowagm * o . J .ri.HV «7.4.HJ .2I-.4J \nrfl.uJ AKHO H\ ATJW .wr \ I- xr-. ,-\ I m.v _._ I I)\. I “Id\. I I.I+ ...\.......n + \.Ix,,. I d...-..fl...mnIu,,I_ C an.H o,u at: + - . r. r t I 1 . . )0 .\ .1 .J/ IJ .1 a; .. I OI I .7. ; A b... _.. A J v A fl 3 .r o . \_ d F. e \1 e I . p... o \J I, I I I I. \ J JI 4 IA/ Mfg“. n+u .Irmo m”... I I o I *xvl-..flo I + 0.0.1.. + vuggowm/ KII¥ \J «I \ J J I _I 9 .- 1 . . , pomh.v “1:0.1 Amm:.v A:.a.v 1.09. 1 g g... , l. - muwm. + *mmm.w+ *Mmm.al M I. I mmw.m «w». I oxm. I W. Va + no: vllk Adv . o 3.14.: ; \ WI. - / x I I /. ..Ia.\) I J . I o / \ q “I o I. AmN.K\_ V A.......\a FO¥ F,” WA. V p7. \. \IN.F \¢ \1 A a? J \. r... I_I . .I I; 0 XI. ) 0 . II I .. O .JvJI .u \/ \I KHZ... + $15....» + *7n,.:. I .c C I IVJQI. w h.. I II . I b v.1 + lifirh. Adv erme>meoo mm ,I I I....I/ II}. \I._\..10/ x)\.~\o/. AO~:.V Ania..- d m:._0 V “1,? mil: \ wad. I Con. + Omz. I .u?. + m.u.H + 0.00% I CC Emfl AUV ./\/l J_! .\ JI/ .:II. ,IJI o 41.]. 4. / Awm: v A-ow.a\ As-“ U A. . Ax.x \ one. + mmo.H .91. + .ms. + -:mfi.m + e.esm I sheasm Aev o . Awe *mwm (\J r—‘I (\J O\ (“J * /‘\ xx) LIN (‘4 + r-\ who.v A H20. I LIN m .C) (T ‘ CY.) r—I \_/ (\I A r I.‘\ eIa O \ / A [1 ‘) H \_,/ heoomzum Aha Afleo.v Aama.v Ammo.v Aheo.v nflem.v Ameo.v m-hm.v *moe. + *MHN. + mom. I :00. I *mmz.fl 3mm. I “mm. I m.emm + mzoi Amv meowpm>pmmoo m: 1) OHom mmqlm H O: mOSIm meIE whiz 2H0 xzom pCMpmcoo EmpH .omma\owma "pSOImcfi>fiHIImCOHpmasooo moaonmmzon CH mmHmEmm do ocoE>quem no mama mcp pom muddmmp coflmmmpwmg mmsmsdm pmmma mwmpmlozell.m mqmHggsm Amv A333.V Hmzo.v A3H3.v AmH3.v mamo.v Apm3.3 A333.mv 333. I mmc. I Hwfl. + *mmq. + T33. I Hmo. + Hn_.3 + 33H.H + cHamHIm va Ammm.v A3H3.HV Ammm.mv A3m3.3 AHmH.u3 “3mm.»3 A3.3qu *mHm.HI Hm .mI 333.HI mom. + 33:. I “33.nI 3.333 + H.33m + uzom Amv mcoHum>mepo mm Aozo.v Aomo.v “mom.v ANI.343 \H33.v 3H3. I 3H3. + HmH. I 3,.3H + :33. I 3H.HH + 3cmam3 A33 “H33.V AH33.V A7m3.3 A ;H.Iv «3n3_3.v HHO. + 3H3. I *DHH. + mmq.m I man. I 33H.m + qunm HOV Amoo.v A3H3.V AH33.V Amoo.v A333.V A 33.3 Am_fi.nv 3H3. I Hfio. + *qu. + *mfio. + 333. I mmm. + 331. I mHH.m + cHommIm ADV Ammm.v A3H3.3 Amm: mv Homm.v Ammo.mv AOIH.3V Am.mmmv *qu.HI mam. + *HHH HI mom. I :mo. + mmm.HI mm.mom + 3.5:m + mzmm Amv mCOHpm>mepo m: oHom meIm Hm: mosIm meIs sz: HzHo mzuu pcmpmcoo EmpH .ommH "H HmUoE CHI wcH>HHIImCOprQ3000 mUHocmmsos CH mmHmEmm mo ucmEonqem no mpmp mnp Lou mpHSmmp COHmmmpmmu mmpmsvm .pmmmH mmmumlozell.n mqmde 94 .memH pawn Lmu .a vgu um o:MngHcaHm meMOHficH .x. A3H3.v A333.V A333.V “333.33 A333.v x333. I 333. + HH3. + 333.3 333. HHH.3 + 333533 A33 A33..3 A33 .3 “333.3 Am3m.33 AHH3I.V . *333. + 3H3. + 333. + 333.3 333:. 3H3.m I 3H3333 Amv A333.V A333.V A333.V “333.3 A 33.3 A333.3 A3I3.33 333. I 333. + m33. + 333. + a... I 333. + I33.m :33. I :Hnsz. A33 A3H3.V A333.mv Am3a.xv A333.3 A333.33 ”303.3 A3.3-43 *mmm.mI ©33.H+ 3mm.:I 3mm.HI \33. + Hnm.m+ m.m.H m.m3m+ x23; A33 mcoHum>meQo 3m A333.v “HH3.V A3H3.3 A3I3.3v A3333.3 333. I 3H3. + 333. + 333. 3333. ma3. + 333933 A33 A333.3 Aum3.v A333.V A333.33 “HHJ3.3 333. + 3H3. + 333. + mum. 3333. row. I 3Hgazm A33 A333.V AmH3.V A3m3.v A333.3 AH33.3 A333.V AHHH.33 333. I 333. + *333. + 333. + 333. I 333. + 333. 333. + cHammIm A33 A333.V A3H3.Hv A333.mv “333.3 Aqu.mv A333.mv Am.HH33 *NH:.NI 005. I *mao.ml moo. + mwm.m+ mam.:+ *m.wmc mm.mm + mzom Adv mCOHpm>meno m: 3H33 gqum Hm3 mozIm mqus m3I: H2H3 3233 33333303 EmpH .ommH "H HmUoE nCHImcH>HHIImCOHpmq3000 meocmmzoc :H.mmHmEmm mo ucmEmoHQEm no mpmp mcp pom muHSmwp COHmmmpwmp mmLmSUm.ummmH mMMmeozeII.m mqmmeoo 3m A333.3 A333.3 “333.3 ”333.33 “333.3 *333. I 333. + 333. I , 333.3 + *HH3. I 333.3 + 323533 A33 A333.3 - A333.3 ”333.3 mug..3 A333.3 . . - 333. + H33. + 333. I 333. + I 333. + 333.3 + >33333 A33 Ammo.v ”moo.v AmHo.v AN:Q.V Amoo.v Ammo.v Acmn.3 A mm.mv . *nmo. + *oHo. I H00. + #:QH. + HHQ. + :30. I :ma. I 3mm.fl + rmo.m + CHQrEIm ADV A33m.v Amom.v Aoww.v A733.mv Asmm.v H333.Hv Harm.m3 an.3wHV *333.3I *333.HI 333.H+ 333.:I 33H. + 333. + 3.3. + I. 33+ H.3Hm + . 3233 A33 mCOHpm>meoo m: Emu OHom mmdlm H53 mOfilm mmdlw ‘-—< malfi HRHO mzom pCMpmcoo EmpH .ommH ”HH HmooE .CHIwCH>HHIImCOHpoq3ooo moHogmmsoz CH mmHmEou go pcoE30HoEo mo moms mco Lou mUHsmmL :onmmmeu monogam ommmH mwmumloseII.OH mqmmmmpo 3m 3333.3 3333.3 3333.3 3333.33 3333.3 *3m3. I 333. + 33.3 I $333.3 + 3333. I 333.3 + 3:35m3 333 AHm3.V . 3333.3 3333.3 3333.3 Am33.3 *333. + 333. I 333. I 333. + *333. + 333. + 333333 ADV 3333.3 3333.3 3333.3 3333.3 3333.3 A33Hmo.v 3333.3 3333.33 *333. + *333. I 333. + 333. + 3333. + 33333. + 333. + 333.3 + 333. I aflozmIm 333 .A3mm.av A33m.3 3333.3 3333.33 A33m.3 3333.33 3333.33 33.3333 *333.3I *333.3I 333. + 333.3I 333. + mam.m+ H33.3+ 3.33m+ H.333+ 3233 333 mCOHpm>L®mDO m: 333 3333 mqum 333 332I3 mquz m3Iz 3233 3233 pampmcoo 5333 .3333 "33 Hmnoe .sHIwcH>HH IImCOHpmasooo mUHonmmsog CH mmHMEmm mo pcmEmOHQEm mo mumg mnu you mpHSme COHmmmemL mmpmzvm pmmma mwmumIozsII.HH mqm 'W ( ‘Jpplby + 7'30‘1‘ (W J . 13%;: m L.’ \ r— 1 ma no I' g‘ r'\.' r”\ r»). 99 .Hm>mH pcoo Loo mm on» pm mocmOHuH: Hm mopmoHocHx Aooo.o Ammo.o AHHo.Ho Amoo.o moo. + ooo. + HoH.H + Hoo. u zoo. . ocoEmo Ago mooo. “moo.o AHHo.o AHmH.Ho Amoo.o ooo. + *Hoo. . Hoo. . mom.o + zoo. + omo.m + Hfiooom Amo AoHo.o Aomo o floooo.o Amoo.o AHHo.o Aamo .o oHo. + moo. + oooo. I *Hmo. . Hmo. . mom . + Hoo.o + ouHmmuo Aoo Amo:.mo Aooa.mo AmoH.o Ammo..o Aomo.mo Am.ooao mmo.o+ mmmd+ *wmoou moo.mn oom. mgommn 3.3T mzom Am; mCOHum>pmmoo mm Aooo.o homo.o Amom.ao Amoo.o mHo. I Hmo. + soo.H + ooo. . moo. + ocosmo Aoo Aooo.o Homo.o Afino.o Ammo.mo Aoao.o ooo. + moo. . moo. + moo.a I ooo. + Hmm.m + zoooom on Amoo.o AoHo v ”moo.o AmHo.o Asao.o Amoo.o ooo. . *mmo + moo. + *Hoo. . HHo. . Hom. . ooH.o + oqumuo Aoo Amoo.Ho Homo.mo Amo3.o Hamm.m Aooo.mo Ao.oHHo Hmo. u moo. + .oam. . moo.a+ moo.mu o.ooa+ oo.om a mzoo Aoo mCOHpm>mepo m: .omouo Hmo ooyuo moous moi: HzHo mzoo ocoomcoo EmoH .ommH .mQOHmeSOOO mUHonmmso: CH no: pun >mesocH mUHohmmso: opp CH mmHmEmm mo unmEmoHoEm mo mung on» now mszmmL COHmmmemL mmpmsvm. ammoH mpr.mI lozell. MH mgmoH moon Loo mo man pm mocMoHuchHm mmumOHocH * Aooo.o AHHo.o Homo.o ”Hoo.o *omo. I Hmo. + *me.I + Hoo. I mOH. I osmEmQ Hcv “moo.o Homo o A Ho.o HHn\.o HHmoo.o . Hoo. + :Ho I oHo. I *Hmm.m + mooo. + oom. + HHooam Amo HoHo o “Hmo.o Hmoo.o AomHo.o A.Ho.o Aqu.o mmo I omo. + moo. + Hooo. I oHo. I Homm.m + mHo. I oIHmoIH Aoo Aoom.mo HoHH.oo Hmzo.o Homo.mo Ammo.mo Am.ommo HoH.HI mHH.HI Hooo.mI mmo. I Hms. + omm.o + m.mmm+ xzom Amo mr OHpm>Lomoo Hm Asoo.o AMHo Hoow.o Ammmoo.o moo. I moo. + Hmoo.m + moooo. I Hom. I oomEmo Aoo Hmoo.o AHoo o HmHo.o .ono.o DHoo o . mHo. I HHo. I mHo. I *HMH.M + mooo. I moo. + HHooom Hoo Anoo.o HoHo.o Hmoo.o Amoo.o Ammo.o Hmoo.o moo. I ooo. + moo. I HHo. I .oHo. I *qu.fl + moo. + oIHszm Hoo AHoo.v AMHm.m, Ammo.o AHHH.mo A_Hm.mo Am.ooHo oHo.HI *Hom.oI moH. I ooH.m+ oom.H+ *o.Hoo+ mo.oo I m2om Hmo mCOHpm>meno m: mHoIH Hmo mozIm mooI: moIm HLHo mzom peopmcoo EmoH .oomH .mCOHmeSOUO mnHonmmzon CH no: pap HupmzocH mUHosmmzon mnp :H mmHmEmm mo pcmEonQEm mo mums ocu pom mpHsmmp COHmmopmop mmpmzcm pmmmH. mprmI IozeII. 2H mqmmH pCmo Coo mm on om moCmonHCme mmmeHoCH .x. lOl AoHH.o Ammm.o Azom.Hv HHHm.o *mmo.m+ mmH. + omH.H I mHm. + m.©mH I HCV Hoo.m AH.mHoo mo.H Am.ono HHo.mHv mmo.o :H.omI mm.H HH.HmI mo.mH I .ooooH+ Hmo AHom.v Hoom.v HomH.o AHom.mo Hoom.v Ammm.Ho *mmo.m+ 3mm. + omm. I *ooH.o I omo. I *oHo.m I m.omm + QIHszm Hov Hommyo Hmom.v AmmH.o Hmom.mo Aon.o AHom.Ho moH. I omo. + *Hmm. + *mom.o I mHH. + *mHo.m I .moHH + mzom Amo mCOHpm>mepo Hm HHmm.Ho AoHo.o AoHo.mo AoHH.o oom. + moH.H+ omo. + ooH.H + H.mom I ocmsmo on Hoom.o Hoom.:o AHHm.o Homo.mo AHoo.v oHo. + mmm.m I omm. + moH.m I mHo. I .HomH + HHooom moo Ammm.Ho Ammo.o Ammm.o AHmm.:o Amom.o Aom:.mo ooH. + .mH. + ooH. I mzo.m I mmH. + oom.m I m.ooo + oIHooIH moo HooH.o mom.o HmMH.V Ho m.mo Aon.o Hoom.mo moo. I ooo. I oom. + ooH.m I HmH. + omH.H I o.ooo + mzom Aoo mCOHpm>Cmmoo m: mmoIm Hmo HQZIH HHoIz moI: HzHo mzom ocoomcoo EmoH .ommH\ommH "mCOHquSOUO mUHonmmzoc CH pOC p39 mnemsuCfl mcHonmmson on» CH mmHmEmm go pCmEHOHQEm no mama on» Com muHsmmC COHmmmCmop mmumsdm pmmmH mwmumIozeIl.mH mqm¢e TABLE l6.—-Two—stage least squares regression results for the rate of employment of males in house- 1950. holds occupations F—LFP URI GINI 4-0 PCNR Constant Item M9 observations (“\J \O \O (Y) I (x- :r (“J +155.5 \ A PC (a) (\J ('1’, ‘J (I V3 C0 LL,\ /'\ ‘ (‘J m \_/ l71.2) ( r4 F4 .005* (\J .793 r o I—Hao (b) /'\ "7‘ H.‘ k) .006) \/ (I) m \J 1023 (Y) ”'3 C) L) C) Q 0 v .013 011 + (”"7 (\J O O O O + l.970 >3 1 0. Sup /'\ CO (3 IV ) i r1 / I U / .\. /'\ O\ 17’ -‘r in ("‘J C) \,./ /'\ \0 Ln (3 r'—i + l.fl73 Demand (1) 37 observations -1.610 I ,«\ I 4*“ (\J 06.1 r—I + PPNR Q) \/ /"‘\ (\J r—I "ID \./ Q [\_ .:V'] \.I x’ ‘\ .C) I—I I"\ C) W 07 ”x I .2 . [ (n (:3 ("1'") H30 :1-1 - (f) LIN if) (’3 \/ ID C3 \,.’ -: 1:) (a) F" 3 IMYV ) (Q LI LIN ("\IJ SUDply N (g) If" 1 i \J '1 ) \o’ (.003) .005 (.025) + (\J OJ 0 *emand ‘ \ A} (0) C) C) at gnifieance aicates Sl . In .Ho>oH oCoo Coo mm me pm mocmOHuHCme mmpmoHoCH * 103 Amoo.o “moo.o Comm.o Hoooo.o ooo. + moo. + HHH. + mooo. I mHo. + ocoeoo Ago Amoo.o AHoo.o HmHo.o AHom.o Amooo.o moo. I Hoo. + Homo. + HmH. + *mHoo. I moH. I HHooom “mo Amoo.o Homo.o Homoo.o Amoo.o AHoo.o AmHm.o *HHo. + moo. I oooo. + moo. I omo. + mHm. + Hmm. I ozzIs “mo Aom.mo Amo.Hmo A mo.mo Ammo.oo Aon.Ho Ao.oooo mm.HHI mHo.H + omm.HI on.o+ mHo.H + H.ommI mH.om + mzom Coo mCOHpm>Lomoo Hm Amoo.o Amoo.o Ammm.o Hoooo.o *HHo. *mHo. *ooH.H + mooo. I oom. I ocogmo Aoo HHooo.o Amoo.o AHHo.o HomH.o Aoooo.o mooo. I mHo. + omo. + *moo.m + *Hmoo. I mom.H I HHooom Coo Amoo.o Amoo.o AmHoo.o Hooo.o Amoo.o Hmmm.o *HHo. Hoo. oooo. + Hoo. I *oHo. + *mHm. + omm. I ommI: Aoo Ammm.Hv Hmom.mo Homo.mo AHmm.oo Amo.mo Am.momo *mom.o I mom.o I AmoH.HI AmmH.H+ mm.oH+ o.Hmm+ m.HomI mzom Amv mCoHom>Comoo m: mooIm Hmo mmoIz mHoIn moIH HzHo mzom osmomcoo swam .oomH "wCOHmeSooo moHon . . ImmSOC CH mmeE mo pCmEonoEm mo moon oCu Com przmoC COHmmmCmoC moCMSUm ammoH owmme039II.HH mqmoH pCoo Coo mm o:» pm moCmonHCme mooMoHoCH * Amo.Hmo Hom.mmo Amo.omo nom.mmo moo.H I mHH.m+ mo.mHI mmm.m + mooH+ ocmEmo Ago Hmoo.o AHHH.mo Ammm.o Homm.Ho Hoom.o mom. + *oHH.mI *Hmo. I *mH .:I *mHm.H + .mmmH+ HHooom Amo Aoom.o Amom.o Hmmm.o AmHm.mo AmmH.o HmmH.Ho *mHo.m + omo. I omm. + *moo.oI *on. I Hmom.mI .mHoH+ omxIC Coo AmHm.v Ammm.o Ammm.o AHoo.Ho AmMH.o Ammo.o Hmom.H + *oHo. I mmo. I omm. I Hoo. I *moo.m+ m.HoH I mZom Amo mCOHpm>L®mDO mm AHHH.mo .AomH.Ho :Ho.mo Asom.mo- , mmm.m + Hmm.HI mom.H+ mmo.m I m.oom I ocmgwo Hoo AMHm.o Ammm.mo HmHm.v Aomo.Ho Ammm.o on. + Hmo.mI omm. I omo.HI *mmH.H + m.omo + HHooom Aoo AHmm.o Ammm.o AHmm.o AHmo.Ho AomH.o _ AHHH.HV . *omo.H + *mmo.HI omm. + *omm.oI mHm. I mm . I m.mmo + ommIz Hoo Ammo.o AHmm.o Hmom.o Ammm.Ho HomH.o Homo.o *moo.m + moo. I mom. I :mm. + moo. + mom. I o.mmH I mzom ,on mCOHpm>Cmmoo m: HHoIH Hmo mHOIz mmoIz moIz HzHo mzom ocoomcoo EmpH .ommeommH "mcoHpmaoooo moHoo Iomzon CH mmHme mo pCmEHOHQEo no opmp map you prsmmC COHmmmaoC mmhmsom pmmoH mmMprozell.mH mqm<9 lOES .Hm>oH pCoo Coo mm me pm ooCmonHCme moomoHoCH * HHmo.o Amoo.o Comm.mo Ammo.o Hmo. + mHo. + mom.H + Hmo. I omm.m I ocmgmo “no Hmoo.o AmHo.o AoHo.o AmHo.Ho Amoo.o moo. + moo. I Hmo. + *HHo.m + ooo. I moo.H I HHooom Hmo AHoo.o AoHo.o “zoo.o AmHo.o Homo.o Ammm.o mHo. + mHo. I moo. I moo. I *omo. + *moo.o + mo:.H I oIHmmIz Hoo Amom.Ho. AmoH.mo Homo.Ho AoHH.mo AoHH.mo Am.mHmo on.H I mom.H I mmo. + t m. + omm.H+ m.mmm+ H.moH+ mzom Coo mCOHpm>Cmmoo Hm Ammo.o ”mom.o Amm.HHo Ammo.o Hoo. I omH. + omH.H I Hmo. + mHm.o I ocosmo moo Amooo.o Ammo.o Ammo.o AHHm.Ho Amoo.o :ooo. + Hoo. I *mmo. + *mmo.m + moo. + mmm.m I Hmooom Aoo Azoo.o Ammo.o AHooo.o ,mHo.o Ammo.o Aoom.o *moo. + Hoo. I mooo. I mHo. I omo. + *Hmm.m + mom.H I oIHomIC Coo Ammo.o Ammm.mo Ammm.o Como.mo Amom.mo Am.HHHo mom. + mmH.o I mmH. I -oH. I Hom. + m.mom+ m.mmH+ mzom Hmv mCOHpm>Cmmoo m: mooIH Hm: mmoIz mmoIz moIz HzHo mZom ocmomcoo EmoH .ommH ”mCOHumodooo moHOCmmdon CH poC pan mppmdoCH mUHOComsoc . mCu CH monE mo pCoEHOHQsm no moon me Com mqumoC ConmmamC mmasom pmmmH mMMmeozeII.mH mqm<9 106 .Ho>oH oCoo Coo mm mCo pm ooCmOHmHCmHm monouHoCH .x. AmHo.o Ammo.o Hmmm.o Amooo.o moo. + moo. + *mmm.m + Hooo. + mmo.H I ocmgoo Ago Amoo.o AHHo.o Ammo.o Ammo.Ho AHHoo.o Hoo. + mHo. + omo. + *mmm.m + mooo. I mmm.H I HHooom “mo AmHo.o Ammo.o AHoo.o AmHo.o AmHo.o Ammo.Ho Hoo. + moo. + moo. + oHo. + mHo. + *omo.m + mmo.m I oIHzmIz Aoo Amm.mo moo.HHv Ammm.mo Ammo.mo Hon.Ho Hm.oHoo mm.HHI mHm.H + omm.HI on.m+ mHm.H+ H.ommI mH.om + mzom Coo mCoHpm>Cmmoo Hm Ammoo.o AmHo.o Ammm.o AHHoo.o mooo. I moo. I *HHH.m + mooo. I mmo.H I ooogoo -Aoo Amoo o AHHo.o Ammo.o Hmom.o AoHoo.o . moo. + moo. + mHo. + *Hmm.o + mooo. I moo.m I HHooom Coo Amoo.o AoHo.o Amoo.o AHHo.o AmHo.o AmHm.o moo. + Hoo. + moo. + moo. + moo. + Homo.o + mHm.H I OIHmmI: Hoo AmHm.Ho Ammm.mo Homm.Ho AHmm.oo Hmm.mo Am.momo *mom.o I mom.m I m H.HI mmH.H+ mm.oH+ m.Hmm+ m.HomI mzom Coo mCOHpm>Cmmoo m: mmoIm Hmo mmoIz mmon moIs HzHo mZom ocoomsoo EmoH .oomH “mCOHpmasooo moHocmmCOC CH no: poo HCpmSUCH moHogomdon . oCu CH mmHmE mo pCmEHOHQEm mo momp me Com mszmmC COHmmmCmmC mohmsom WmmmH mmemIozeIl.om mqmoH pCmo Coo mm me pm mocooHuHCme mmpmoHUCH * AHo.omo “mo.mmv AHH.mmo AHH.Hmo . Hm.oHI omH.m + mH.HHI moo.m + .mmmH+ ocmEoo HCV Ammm.o Hmmm.oo A mm.o AHmm.mo Amzm.o mom. + omm.mI mmm. I *Hmm.m I *mom.H + .mHmH+ HHooom Hmo Azmm.v Ammm.v mm:.v AmNH.mV Ammw.v Am®©.Hv *oom.m + on.H I mmo. I *mmm.mI mHm. I mHm.H I m.omm + QIHmzus How AmHm.o Ammm.o mmm.o HomH.Ho AmmH.o Comm.o *mmm.H + *on. I mmo. I omm. I Hoo. I *mmo.m + m.HoH I mzom Hmo mCOHpm>Cmmoo Hm Homm.mo AHH:.HV AmoH.mo Amom.mo I mom.m + mHm.H I omm. + omm.H I m.on I ocoEoo Hoo Ammo.o AHmm.mo ”mom.o HHHo.mo AmHo.o moH. + mom.oI mom. I mHm.m I *HHH.H + o.Hmm + HHooom HOV Comm.o AHom.o Ammm.o Homo.mo HHHH.V Ammm.Ho *omo.m + *Hmm.H I oHo. + *mmH.mI oHH. I HmH.H I m.mHm + oIHmmIz Hoo AHom.o AHmm.o m m.o Hmmm.Ho HomH.o Comm.o *moo.m + mom. I mmH. I omm. + moo. + mmm. + o.mmH I mzom Coo mCOHpm>Comoo m: HmoIH Hmo mmoIz mmoI: moIs HzHo mzom pcoomcoo sooH .ommH\ommH umCOHpmozooo moHogmmzon CH uOC ozo.mpumooCH moHOCmmson on CH mmHmE ho pCosonoEm mo mme on Com mszmoC ConmmamC mmpmzom pwmmH ommpmlozeII.Hm mqm¢e 108 .Ho>oH oCoo Coo mm oCo pm ooCmoHMHCme mopmoHoCH .x. AmHo.o HHoo.o Amoo.o Hmmm.o HHoo.o *mmo. + *mHo. + *oHo. I Hmo. + Hoo. + omm. I ocmgoo Aoo Amoo.o AmHo.o Homo.o Hmmm.mo AmHo.o *Hoo. I moo. + moo. I mmm.m I *Hoo. + HoH.H I HHooom “mo Ammo.o AHoo.o AmHo.o Ammoo.o Aqu.o HHHo.o Ammm.o omo. + *mmo. + mHo. I mooo. + moo. I moo. + mHo.H + mom. + ooooIH Aoo AHom.Ho “mom o Aomm.o AHmH.o AmmH.o Comm.o Amm.moo HmH. + omm. + mHH. I mmo. + mmm. I mmm. + mH.mo + mo.mo + mzom Amo mCOHpm>L®mDO \Im AHHo.o Hooo.o Amoo.o Ammo.Ho Amoo.o *Hmo. + Hoo. + mHo. I omo. + *mHo. + mmo.H I ocmEmo Aoo AHoo.o Amoo.o AHHo.o Ammo.o Amoo.o *moo. I moo. + mHo. + mHo. I *mHo. + mom. I HHooom Aoo Ammo.o Amoo.o AHoo.o Hmoo.o “mmo.m AmHo.o Ammm.o Hmo. + moo. I *mmo. + moo. I mHo. I moo. I Hmm.H + mmm. + ooooIH Coo AmHm.Ho Ammm.o Hmmm.o AomH.o AmHH.o AHmm.o Hm.mmo mom.HI *Hmm. I *mmm.H+ mmm. I mmm. I mmm. + *m.omm+ Ho.mm + mzom Amv mCOHum>meDO m: mozmm ooozIH mHoIo oosIH HHAI: moIz HzHo mzom pcoomcoo EmoH .ommH "WCoHomasooo MCHmmo ocm .mCHCmoHo .mCoCsmH mCo CH memEmm oCm mmHmE mo pCmEmoHQEo no mme on Com monwoC ConmoCmmC mmpmsow pmmmH owwumIozeII.mm mqmoH pCoo Cog mm oCo om ooCooHuHCme momeHoCH * Amoo.o Amoo.o Amoo.o AoHH.o AHmoo.v *omo. + *moo. + moo. I HHm.H + Iaoo + mmo. I ocoa mo Aco Hmoo.o AmHo.o Homo.o HmmH. mo AHmo. o *moo. I mmo. + mHo. I .:m.m I mmo. + mmm.m I mHooom va AoHo.o Amoo.o Amoo.o Amoo.o Amoo.o “moo.o “moo.m *moo. + *moo. + HHo. I moo. + moo. + mHo. I *omm.H + mmo. I oooqu on Ammm.o Amom.o Ammm.o AmmH. A Hm.o Am m.o Am.moo mmm. + mHm. + Hmm. I Hmm. + Hmm. I Hom. + *o.mmH+ Hm.m:+ mom Amv mCoHum>Cmmoo Hm Amoo.o Amoo.o Hooo.o AHmu.o Hmoo.o *mmo. + ooo. + moo. + oHo.H + *moo. + mmm.H I ocogmo Aoo “Hoo.o AHoo.o AmHo.o Hmmm.o Amoo.o *moo. I mHo. + moo. + mom. + *HHo. + mm:.H I HHooom Hoo Amoo.o Amoo.o Amoo.o Amoo.o Amoo.o Ammo.o Ammo.o mHo. + moo. + *mHo. + moo. I Hoo. I moo. + *moo.m + mmm. I oooqu Hov HomH.v Homm.o AHmm.o AHmH.o Amom. AmCH. o Hm.mmo *omm.HI omm. I *moH.H+ HmH. I Hmo. I Hmo. m+ *m.mmH+ mmm.m + mzom Amv mCoHum>Cmmoo m: mozam ooozIm mmoIm omSIH mmoIs moI: HzHo mzom pcoomcoo EmoH .ommH “mCOHpmosooo mCHomo oCm .wCHCmoHo .zpoCsmH 0:» CH monEmm qu memE mo pcofimOHQEo mo mme on Com mpHSmmC COHmmmamC mopmsom ummmH owmpmlozmll.mm mqm¢e 110 .Ho>oH pcoo Coo mm mCo pm ooCmoHHHCme mmpmoHoCH * AmmH.o Hmmm.o Aomm.o AHHH.V AHmm.o HHH. + mHm. I Hmo. + omH. I *Hom.H + mm.oH + ocoaoo Ago AmcH.v HoIo.Ho Hom4.o Hmom.o Homo.Ho omo. I mmm. + Hmo. I HmH. I mHo.m + H.moHI HHooom Hmo HomH.o Ammo.o Agom.o HomH.o HCmH.H .mHo.o Comm.v *mmm. + Hmm. I mmm. + Hmm. I Hmo. + HmoH. + omo. + mm.m: I oqu HHV AomH.o Hmmm.o Howm.o Amom.o Ammo.o ”Hoo.o HHHo.o 33H. + mmo. + mJH. + mmH. I moo. + HHH. + mHm. + HHu. I mzom Aoo mCOHom>Cmmoo Hm HmHH.o Hmmm.o AmHm.o AHHm.o AHmm.o Hmo. + mmH. I moH. I 3mg. I *omH.H + mo.mH I ocmsmo moo HmCH.o Ammm.o Ammo.o Hmmm.o Aomm.o moo. I mmm. + mHo. I mmm. I *mom.H + mH.mm I HHooom Aoo HHmH.o AmHm o Ammm.o AmmH.o Ammm.o HHmo.o Homo.o *Hmm. + mmm. I Hmm. + HHH. I mmm. + *HHH. + mmm. + Ho.mm I ooonI Hov AmHH.o Ammm.o Ammm.o AHHH.o AmmH.V Homo.o Ammm.o HHH. + mHH. I omm. + Ho . I mHo. + mmo. + mom. + mH.mm + mzom Aoo mCOHpm>Cmmoo m: mozmm Hoo3IH mmoIm oszH mHHIC moIr HmHo mzom osmomcoo EmoH .ommH\ommH .mCoHpmosooo wCHmmo oCm .wCHCmmHo .zCoCsmH mCu CH monEmm oCm monE no uCoE>0HQEm mo moon 0C» Com mpHommC COHmmmaoC mopmsom pmmoH omwme Iozell. am mqmCowoo Hm Amoo.o Ammo.o Ammo.o Ammm.o C.am.mo Como.o Hoo. I *Hmm. + mmo. + *mmH. I anm.m I omo. + omm.m + ocmgoo Aoo “Hoo.o Ammo.o Hmmo.o HHHH.HV ”moo.m . * mo. I moo. I MHC. + OHm.m + H66. + omm. + mHoojm Hov Amoo.o Ammo.o AmHo o Homo.o HmHo.o AHmo.o Ammo.o HoHH.mo moo. I mmH. I omo. I moo. + Hmo. I omo. I HHo. I moo.m + Hmm.m + ooooIm on Homo.o HHmm.Ho HmHm.o Ammm.o AHmm.o Ammo.Ho HHHm.Ho Am.nHo HoH. + mom.HI omo. I *mmm.m+ moo. I Hmo.HI HHm. I *m.mmm+ mm.mo+ mzom Amo mCOHpm>Cmmoo m: oHHHm mozwaoo Hm Amoo.o Ammo.o Amoo.o Ammo.o H.om.ao Ammoo.o moo. + *Hoow + Hoo. + moo. I *mmH.m + mooo. I moH.m I ocoEmo Aoo Amoo.o HmHo.o Ammo.o HHHm.mo quo.o *mHo. I moo. + m-.. + Hmu.m + moo. + I HHaaam Coo Hmoo.o Ammo.o AoHo.o AmHo.o Hmoo.o Homo.o Homo.o HHIm.Ho moo. + *Hmo. + moo. I moo. + oHo. I oHo. I HHH. + *moo.m + mom. I ooooIm Hoo Homo.o Hmoo.Ho Ammo.o HHom.o AHmm.o Amtm.o Hmom.Ho Amm.omo .*HHH. + *omH.mI mHm. + *omo.m+ mmm. + qu. I *mHm.m+ *H.mmm+ H.HHHI mzmm Hov mCOH pm>chmDO m: OHH¢m moz¢m HOUBIE mmqlm_ mCfilm mmJIE mnlfi HfiHm mzom pcmpwcoo EmuH .ommH ”mcoHomQSOoo mCHozo on nmCHCmmHo .HCUCSMH on CH monEom Ho pCmEHoHoEm mo oomC on» Cog onsmmC ConmonoC mmCmsom pmmoH mprmIozeII.mm mqmmeoo Hm .0 .1. (a o. W a. \.I\I.-/ \\.4 o, HooH V AmmH.v ApmH o mmm. m _..g . Ho o *omm. I *mom. + omm. I mom. I MrH. + Hmo. I o.moH+ oomamo Aoo A» m.v .dmu.w HI. .m muI.u Icon.” HHD. I bx. + . . + opt. + mwm. I 66.HH + mHoann Hum Ammo.o \ / *mmm. I * r—{ v A (“\J LI'\ 3 (\J (\J (\I O O V I A (vxko Ct) . I [\ ‘ ‘0 r—4 :) I + ('1 ‘ + ' I 4 + (_) ‘ 1‘1 X) C + ’10 .3 II“ I A) b ._1 I [L A D V AHH6.v A: H.v 66m.m men.o Homm.m Hmo..w HInm.V *Hmm. + H . + 000. I Hem. + mHQ. I HHm. I won. + Hmm. + um.@H + mzum Amv mCOHum>L®mQO 03 OHBdm maz ,H ‘_ :. IAI . J a) ( I \_; r'\ ,. r'"‘,\ r I O \, ‘/ ,8, (7“ I') —\ 0 ,/ I I ' \ .I - l J O '\ * I I 317 [[A‘I I I C) 1:, O ("W V I oopply I V 111+ /”\ 3’: I 'T) " .o r~ I I _} o o \1 r‘ (“I r, "I 54‘ L. u n. /\ I . v 0 s_/ r"‘\ "-L) .‘> .A C ‘_r r"\ .w\ \ . [.. 1 o \. / '\ ‘ I a ‘I c \ . _) r '1 u _,/ 7. \‘v I : I (W I) I I V./ ‘) I . O +. '"x ‘\ X) I_I I 1’ n) I) O C + .‘ . C + I 'I l C I"r\ I I L o n I r—‘\ I (‘.I f l o o . ,, .\ r~ , ‘.‘I I . '7 7I I . H / ‘1 ’ I :I I I I J . O I . ‘) . - O O A, H j ) l I I c I ,. .. .CI'I- l-x.y'v~ i. I. 1 empt-. '\0 K4. P128 '. In?“ dl’l-..b , ole ., L17? - L '1 nstant f" L 0 Item r‘ J 102’} Jat‘ Y‘I‘ A B9 Obse V "\ l I'YN (”5 CI) 0 0 IV" FT: T I I v C v + A a! 7 "J ;\~ r J (\J T I . \J A L3 fix. d3 :‘J n N O O H \/ ‘3') 1“ nwa ) (a I”\ U“. r‘.\_ I”? ("I ('7 (“I CU (- I O I (",1 .IN‘I (A C_ I C.“ (.5 O I \‘I. I * ,~\ [2 ‘J 1’-) " C) \J O l + * r ‘\ L.—\ I‘W I I (f. C) D O O + /‘\ m ("J C) (:3 C) C) C C v + A “I LA 30* I I . J ¥u . I no «‘1 ) \o (3:) V' I-LCD \ 1-: (b) i) Supply ) " ) ( i115 I If)“ J J I: 32) (I 0.! ( L, ) l A * . 4 f\ C) C) r3 0 O 0 v Demand (d) 1 Q2; Idx i) (.00’ Observations * " I; 3‘“ U \ ”.3 -—7‘ (WW 0 . \_, I /-\ IVW\{_I x.) ‘ _) 7 o o ._./ * /"‘x . I) ‘1'1 I" y“: .‘-I r 4 o a ‘_/ + /"\ ’\- I I v‘ _‘_} F3 _, O I v I * A FA“ (‘W Lax \f) r' O O \J + I“ (:1 (J "C -{T\ (V) _T O O \/ + /'\ ,_‘ "‘7‘ L; ‘I 'T «.V—‘\;3 C O \./ + * A #4 (\ O . if) C) r—i '."W ,——J \._/ + to C) ‘O . r‘I . I‘ I If: H A-. .' '34 A (1) \x ‘I II 1“. II I) ;" + .095 * 1.J\—JI‘: r g \ \(“2 I} I‘ll . J ”7) C) O O \_,-‘ vN ‘ ' - U I. I U l + /"'\ ~I '3‘ 31 ' — 7E . ‘ I O l r"\ --4 ‘1 \L‘) ‘v'] r 4 UN 0 I \_J + * -"\ 3") ITKJ‘ O "D I'D C) O O \J + 4 FJ‘ '1 ’1’) /\ ‘40 .4;(J) the laundry, in or the rate of employment , and dyeing occupations 1‘ TABLE 30.-Two-stage least squares regression results “8 clerni RATIO DE HAN “C3 ll (9‘ "I F-LFP (A . P \‘w \u k.) lil— PCNR GINI Constant Item Mg observations .777* J 0') (‘1 C7) .146 +185.0 NR PC (a) ('1 \ . (I) ("4 o .U3I‘.) ( 307) .“MS (.1u0) (.152) ”x 'J .1 ) ‘4 U 0 2 +198.9 —LCDp M (b) LI". Supply (c) li‘ (.170) +1.713 GD (WW 0L2 -1. ,5H1- (1.uu;) +2 U'\ Lfl u'\ -l mand (d) 2116 I —{ (\J r-‘i 37 observations .81M* . 8 .219* 2 + \J P") C +l7u.0 L“ PCN (8) rd : on jury)! U. . ‘1 L‘ I l I... \o (u L.“ (f) (.151) ’7 ,_ - "I r. I .z' 41‘ I 1,, \ Li I 4 (1.1 I l " o v . v, ( (.190) no p4 u) Dem (h) /\ .‘ IT) \ LIN \ I 117 .Hm>ma pcmo pmo mm mop um mocmoaoflcwflm mmpmoflocH * Aoofi.v Aamo.o Amom.ao Asoo.o *smm. + omo. I omm.m + oHo. I smo.a + ocoEmo Ago Asoo.o Amuo.o mooo.o AoHH.Ho “moo.o *mflo. I oflo. I moo. + ooH.H + ooo. + mom.a + zaoosm Amv AoHH.o Aeao.o Asoo.v Aooo.v Aoao.o Aomo.uv Ham. + moo. I zoo. I mmo. I oHo. I oms.H + mom.m + ammIm Aoo Aooz.sv Aoso.ao Aaom.v Ammo.v Aoofi.ao Asm.moo mom.o+ moo.HI *mom. + mom.a+ omfl. I mo.os + oHH.o + mzoo Amo mCOHpm>meoo om Amma.o AoHo.o Aomm.fio Aoao.o now. + oHo. I .*oee.m + sdo. I mmo.d + ocmemo Aoo Amoa.ao Amme.fio “moo.fio Ao.omwv Amoo.mo sma. I soa. I Amsfi. + mo.mm I mmm. + mo.:HI saoozm on Aooo.o Aeoo.o Amoo.v Amao.o Aefio.v Ammo.v . *oom. + ooo. I moo. I *omo. I omo. I oHo.H + moo.m + ommIm Ago AooH.ev Aeo:.o Ammm.o Amao.o AmoH.Hv Amq.ooo omo. + mHo. + mos. + mam. + omm. I oo.:o + o:.mm+ mZoo Amv mcofipm>mmwno m: BHozIm ooozIm oozIm omoIs moIs HzHo mzom ocoomcoo EmoH .omma "mCOHpmasooo mmoHOmemoQ cam mcflmmmpopflmn. map CH mmHmEmm mo ucmEzOHQEm mo mums map pom mpHSmmL scammmpmmp mprSUm uwmma mprmIozeII.Hm mqm¢e in the emales 1‘. 4‘ A results for the rate of employment 0 greSSIOn TABLE 32.—-Two—stage least squares re 1960. hairdressing and cosmetology occupations F-WAIT F—WCOL HOP LFP F— M- I-UE H , GINI PCNR Constant Item M9 observations Li. ('J r" [\- IT r-I +1.510 r1 ‘0 U\C) O\L1\ (\J 1'3 + v +83.95 PCNR (a) A \O H (T\ (\J (D r'I V + A LEW \. :1 Cd r‘i (D O V I A r". KO 0 I I O r—4 O C) C) V I + A C\ (\J m (:) DJ r‘I C) C) C) V I I A f" I Cl) Ln (Y3 m (V) (3 ‘3 C3 . O . \‘I I + /\ (V3 (" I :J O“ l,L-\ Li“ t‘~~ \I) (\J «\J I—I 4T \J + + I -3 C,- I r“! :f m Lf\ (1") (\J #4 Ln + + >3 «T: r' I [11 C). In D. I 3 [:4 U) A A ,Q C) V v v f‘I’W“ 1;/[l 0v; / K “414) o O ( (.O:U) (.OHO) 118 /'\ /'\ Ox: (\1 [D r“ CC) 0 “O Q r—I' m m v :I‘ m +‘ Iv r'\ A 1 KO m C) r I I—1 Lf\ Ln - O H \O \.../ H v I I A Ln : MOD (V) H v + A .3 r» H CO O\\O V + /'\ M :r O'\ r-I If C) r-I r—-I + v /'\ a“) [‘~- CO U“ 0‘) \D 51' L\ m H O\ (\J + +\« A moo (\J H C) C) v I H :3 N [\ O\ (\1 L0 UN + + U) rd (3 C. 0 CU ~I-I It E p 2 <1) (T5 0 Q > m 514 , (D A U) /‘\ ”U .0 (1) v O v [\~ m (\J 0 CA .1? (\J («\J C) C) (.uOd) .01 OI . C I V v I I /'\ /‘\ Ch (\I O CO 0.1 If m '3 Q C) O C) \_/ V I I A A [‘~ "‘4‘\ O‘I 43- CIZ.) O\ \L) ("W (V) r—{ L) H I“ {‘1 ("I (“*1 v v + + A If\ (TI C) r— (T) k: V + (\J UN \0 r—I O (“J + I >> <11 r-4 [1.] (L m (1 I 3 C14 II) A A (H b0 v V “I; (“II CD and -4 Q) {:3 a (b) If U m C) (.020) £1. LI ’\ ndicates I the in .ployment of females Q LO 0\ H \ O \0 Ch E: H (1) Q4 0 o O (1) l.) hairdressing and cosmetology occupations TABLE 33.—-Two-stage least squares regression results for the ra F-WAIT F—WCOL IOP F-f M-LFP UE M ._ PCNR GINI Constant Item U9 observations 391* a” X) H 0 KO (*1 102.2 + PCNR (a) (.066) L(\ (\I 119 f\ (\J O\ \0 £3 \C) C) CD r-I C) O O O \J + I A \C‘ «‘2‘ (1) \f; XL) LP. (“I IX} \? O 0 I ‘~_/ + + A b» IL!) (V \J [‘\- 1;?! (VI {ZN L‘.’\ LIX (‘4 C) O C) V v + + A 4A KC} If) (D U; \ (Y) C) C (3 r4 r‘—"I :T C J O O O I r 'I r"I | | v A A N ‘13 411‘ UN r—I (x') r—I (3‘. C) ID C3 O 0 O O O V \/ + I f‘\ "\ [\ (WW C.) (‘0 ‘1.) (Y) \Q \0 [\a (V‘ C) \O _‘:r if) t7 0 O O O O H \J H \J + + + A /‘\ if E‘~ OJ r-I C7 [PI (\J Ch [‘~- 0 O O {—4 v I + \0 ON (\I L7\ :1” D I O (\J C1) ‘ f) 11' O‘\ (Y) r—I I + I >3 “(1 (I: r-I C LL] 0. m (I) Q. E I :5 (I) [II (0 L3 A A A .0 0 T5 V v v on (\J (.670) lo) 1 m ‘/ b O ( observations (6) 37 CO C \ Li“- R P C N r? In ‘ I \_ ‘LI'II C) " H I.a )CIL' VIC .-.Vr' fl\ I) lf‘. H’I»: (I) C") O I + 2+: \ \E\ _f 1.3 m4 ' J I O (a; 1'4 + A» "‘3 r'I r—I ‘3‘) r‘I C:) U I ‘74] + (‘x I' r. I 1') T O O + \ t?- L. \ Q C O + "I C [ - r_;\ - l I :‘J r'fi LL CL 1‘5 w I“ 5A?) v _.\ 1' . r- : 1"“! i no, 0 I ,.l I v. , M mt~ ..’\ O“. O O I ’I r I + u_/ 1T 7' 7' ) i ‘1 C I *I .' I 4 ,1 ' .1 11 \ ,x‘y ,- O U C”) if \,/ + I I 0 (1x \j 120 \ .Hm\m_ ucmo Low mm )sw wm mggmMHernow umnflncw * Ammo.v Aooo.v Aoom.o Aoao .o moo. . *ofio. + mm”. + mooo. + moo. + ocmgmo Ago Aomoo.o Aooo.o Aooo.o Aoflo.o Aoooo.o . *omoo. s woo. *omo. + moo. + *rmoo. + omm. + oflooom Amo Ammo.o Amoo.o Amofloo.o Amoo.o Amoo.o Ammm.o moo. . moo. + moooo. + zoo. . *mao. + mom. + omm. + oom-fi on Amofi.oo Aomo.o Amo3.o Aooo.flo Aoao.ao Afl.mfloo oom.ou *mom.o+ *Hmm.a+ omm.fl+ mmo.a- mm.mo + mo.Hm . mzoo Amo mCOHpm>meDO \Lm Aomo.o “moo.o Aoom.v Aofioo.o zoo. + *mao. + omo. + mooo. + Rafi. I ocquo Aoo Aofioo.v Amoo.o Aooo.o Aomm.o “mooo.o *mooo. I moo. . *mflo. + Ham. + *omoo. + Hmfi. + ofloosm Aoo Ammo.o “moo.o Amfioo.v Aooo.o Amoo.o Aqmm.o moo. u *ooo. + oooo. + moo. - *mfio. + mom. + ooa. + mom‘s Aoo Aoom.oo Aoom.v Amoo.o Amom.fio Aoom.ao Amm.omo o:o.on *mmo.o+ *ooo.fl+ mmo.fl+ mom. I mm.mmn mm.oou ozoo Amo mCOHpm>meDO m: HHozu: qooziz moon: omo-: mono HzHo ozoo ocoomcoo EmoH .omma umCOHpquooo . mafigmppmn CH mmeE mo unmeaoaqem mo mpmp wcu aom mpHSmmp coammmpwmp mmumsvm pmmma mmmumlozell.:m mqm $4 V Q) A U) 'U ,C) v O [\ m _ nu.33 +83.85 PCNR (e) C) r—( K“ ’W .— (’1 H (:7 t L) OH (‘1 J .0« (\J C) m) H [H .336 '4 1\ BA h- (f) (\J C) (\1 Lfl (K! .000 (.0000) + .273 .0023 C \ / 094= (. J7) f\. (V) l {‘x (J) O + .0006 (.03 I \‘L) . r—i \QT + Demand (h) C) (\J C.) I"\ ON ("'1 ‘3) 7 Q \‘J-‘* JED) /' k (.00 (0 J ) 0’5 Q) mi ’(3 'L: +4 '- barbering in n results for the rate of employment of males 0 .10 1960/1950. OCC‘JDaClOI’IS . TABLE 36.-—Two¢stage least squares regress IAIT I'll- V V 'Ifl M-OPS -UE I \v ii I NR PC Constant Item M9 observations 122 .06U (.068) .165 (.083) I + * A fig A 0;! b-IWW \O N \O r-I Lflr—I O\ (\J v \_.I + + A A * GILT \O O (\I O\[\ mm CO C) (D H C) (\I v v I I + * /"\ ID (\I O) r—I [\Q :r (h 3 OKO [\ (\I I O O O 0 (\Jv v H I I + * A * A a: I" \O [\-\O C) (VI—‘1' (\I Lfl H (\I r-I O «:T v v I + + A r-I 00 (\J <71) 3 (\J L;’\ r~I on on \Q m I“ —.I (7\ v V + I I :k [\ O (\I H + (\I (I) O\ \O \O (\I r—I r-I :7 (\l (\I r-I + + I >, (I: r—I (I1 <11 D. ’2: (D Q. L) I :3 D4 5: U) A A (U ,0 o v \../ .113) (1.383 ( Q u (.12 .222* (.101) + +1.59M* (.271) Demand (d)' 37 observations A * W KO (‘3 MD (1:) UN O O Cd v I + * A * [‘~- (3 (“J H Lf\ In (\1 CO V + + A * C) (.0 C1) (1') CA \0 r—I C) (\J \,l I + /"'\ \0 ll’\ Lfl 0\ P4 Lf\ (\J CD \0 O I O {—I v I I * A (1') LIN Lfl ON Lf\ O.) r~~I O C) v I + A (W (V) I“ \Q 0") (‘W v + I [\l t\ :3” (\J ~T :7 if r—I + + [II 0: <11 2 CD 0 I (1. TE A A (1) Q4 \J v 10 (\J \J Ifi (WW (‘7\ O * r-x CW \0 LIN In :fr—I (TX!!! FILD r—I(\J +v O (“A C) UN I"I ‘ k?) I IV, /\ [\lfl LI“\L(\ C Li’\ r—I‘J —~A (\J 007* A Deman (h) .1 (.347) 3) 4 (. H on \J r—‘I 51) I l ) .-—I 123 .Ho>mfi pcoo Loo mm ono pm ooCMoflLflcmHm mopmoflocH * Amoo.o AoHoo.o Ammfl.o Aoooo.v *oHo. I *omoo. + Hma. I mooo. I oma. + ocoEmo Ago Amoo.o Amoo.v Aooo.o Aomm.v Anoo.o *ooo. + Hoo. + moo. I mma. + moo. I omm. + snooom Amo Amoo.o Afifloo.o Amooo.o Amoo.v Ammoo.v Ammfi.o *moo. I *mmoo. + mooo. + ooo. I mooo. + man. I omm. + omIz Aoo Aoo:.mo ANHH.HV Ammo.o Aoom.mo Amma.mo Ao.mmfio oam.mI *omm.mI *omm.m+ omm.m+ moo.mI m.mmfl+ om.mmI ozoo Amo mCOHpm>meQO mm Amoo.v. Aoooo.o Aoofl.o Aoooo.o I *ooo. I *omoo. + ooa. I mooo. I mma. + ocogoo Aoo AoHoo.v Amoo.o Amoo.o Amom.o Amooo.o mooo. + moo. I moo. + omo. I mooo. + Han. + snooom on Amoo.o “mooo.o Amooo.o Amoo.v Ammoo.o Aoofi.o *moo. I *qmoo. + oooo. + moo. I mooo. + omo. I mom. + omIz Ago Afloo.mo Ammo.o Aoom.o Aomm.mv Amm:.mo Ao.omfio mmm.m+ omo. + mmo.a+ omm.m+ mmm.mI m.onI mm.mm+ mzom Amv mcoopw>pomoo m: .mozoz ooogIe mooIm omoI: moIz HzHo mzoo pcoomcoo EmoH .omma "mcofiquSOOO pflmoms . mono CH mmame no pomeonoEo mo mums mop Lou mpHSmmp coammopwop moLmSUm ummoa mwmpmlozeIl.mm mqm<9 in shoe Ixales .‘ 11‘ e of employment :. who \4 c ,— 'u ults for the ra P835101". P88 TABLE 38.-oTwo—stage least squares reg \va VF: L4 .hiAi‘J ' . .a. a . A o L I I,’\ rx D ‘ I v 'u’ {:14 , >174 [file‘l' Ju..".. L \ H 4 ecu" 4.. v Constan \D r» . r4 ,‘ . —..4 (0 +03 PCNR (a) :C. . ,, ,. Q) . 0.1.“! .4) .\4 W I \‘J I' IV) (11:) I "I (n) r‘ .) n ) .II‘ fl.” .Vv I I") .;7u b- r-I H (a, 'fl. ) w J < O Offi‘ ‘02) \ou ) (V) 0 LI LI I \ LAW I' \j 124 I :r (F) I III. -\ F" .u‘ I f‘) \. __,I [\-‘O (“W K) apply 3 C) Q r-‘I <:) (:3 (I3 C) (7.) (Y) C)\ C) LI\ 0 C‘ C) t ) C) C) O O _/ I m I'YW C3 O l 'U C (0 C1 O H (T) “ SI 37 obse I: 1.) L1 N A, \\. r‘VI \L’) I . u . u +22 {~\_ 1 fl (*3 r—I \/ C) (D A m :0 \ .00 Hy / I x (.00078) .001 .FI‘L-J": (VI (‘1 C) (D r—I (\I )II 0.1";— I‘" ’I Ifi (:3 + Y! J duppj L.) x (Y 3 \O (\J ”4 J .-. If] \JV {'xr J. + 13 bemax (h) (.170) .0034) ( are .i. 0 cent le' hoe 1218 TABLE 39.-Two—stage least squares regression results for the rate of employment of males 1960/1950. repair occupations M A 0 NR P” ‘V Constant Item A9 observations 0) (\J [xi gx) C) +A03.A PCNR (a) (.132) \O (\J i_.T\ (:3 [so CY ) (.061) r-‘I r—I 01(1) [\-- CI) r—Iv !I\\\L) +22U M-SO (b) 125 .“02 (.301) +103.2 + Supply (C) .uiu (.250) + .A3A (.309) -11U.8 Demand d) 37 observations * /'\ \O O\ CO (‘0 \0 F4 * /'\ 010:) r4 (\1 (.er—4 ;I (If) 1' \J (‘~ C") O (I; (\J \O I D 0 +390.7 PCNR (6:) KO C) (IN (‘0 r-‘I bwo J) XE). \\O I——I K \I .3 C.) +312.0 I—so \ A (f) ['1 ._ I f\. (3 \_/ (\I (D .020 + .116 +185.2 Supply ) < C) [\._ O") /'_\ (Y0 1“ (D \J In Ch (\I (NW (\J W KO ‘x0 + +1.031* (.455) —122.8 Demand ) (h CHAPTER VII SUMMARY The data assembled in the previous chapter seems to indicate that for most occupations in the private households industry, the supply of labor is positively related to the GINI index. The findings also suggest that the wage rate is of little relevance in the supply of domestic labor, but that the opportunity cost of employing a family member in the home to do the chores that might be performed by hiring someone else may be a relevant factor in the demand for.live- out domestics. For the laundering, cleaning, and dyeing occupations, there is also some evidence that income distri- bution is an important factor in determining the rate of employment in these occupations. The lodgings and restaurant industries apparently are an important element in the demand relationship. The regressions for F—BEA are largely insig— nificant while those for M-BAR indicate that M-WCOL is an important factor in the demand for barbers, while M—UE is an important consideration in the supply relationship. For M-SO, there is some evidence that both T-WCOL and HANDE are factors in demand. But the evidence for HANDE is somewhat contradictory. It seems appropriate at this Juncture to summarize the results of preliminary research which dealt with the 126 127 relationship between the rates of employment in routine per- sonal services and the condition of the labor market as indicated by unemployment, labor force participation and migration rates. In these regressions it was found that, with respect to M-SO, M—BAR, and LCDO (male or female), there was a high positive association with unfavorable labor market conditions. Many of the occupations in the households industry were also associated with poor underlying labor market conditions. There was no clear cut interpre- tation of the results for F—BEA, F-HHOin and a subgroup of F—HHOout known as babysitters. Perhaps better results might have been achieved by’ respecifying the model. In particular, it might be appro- priate to include PCIO and RATIO as endogenous variables in an expanded system of equations. GINI might also be treated as an endogenous variable inasmuch as it can be expected to be interdependent with some of the wage variables. In view of the seemingly peculiar results for PCNR in some of the equations, it might be appropriate to make PCNR an exogenous variable with the particular RPS rate an endogenous variable along with GINI. This model might be of the form: GINI = x1 + x2 RPS + x3 M-UE + X“ M-LFP + x5 ALT (10) RPS = y1 + y3 GINI + yu CHAR (11) where 10 is the form of the supply function and 11 is the form of the demand function. SELECTED BIBLIOGRAPHY 128 SELECTED BIBLIOGRAPHY Becker, Gary S. "A Theory of the Allocation of Time." The Economic Journal, LXXV (September, 1965), 493-517. Best, Ethel L., and Erickson, Ethel. A Survey of Laundries and Their Women Workers in 23 Cities. Women‘s Bureau Bulletin, No. 78. Washington, D. C.: Government Printing Office, 1930. Bowen, William G., and Finegan, T. A. "Labor Force Partici— pation and Unemployment." Employment Policy and the Labor Market. Edited by Arthur M. Ross. Berkeley: University of California Press, 1965. Bureau of Employment Security. Estimatinngnemployment. Washington, D. C.: Bureau of Employment Security,' March, 1960, reprinted April, 1961. Business and Defense Services Administration. The Laundry and Drycleaning Industry, A Study of Problems and Prospects. Washington, D. C.: Government Printing Of ice, 1965. Cain, Glen S. "Unemployment and the Labor-Force Participa— tion of Secondary Workers." Industrial and Labor Relations Review, XX (January, 1967), 275-297. Community Household Employment Programs. Women's Bureau Bulletin, No. 221. Washington, D.-C.: Government Printing Office, l9U8. Ducoff, Louis J., and Hagood, Margaret J. Labor Force Definition and Measurement: Recent Experience In the United States. Social Science Research Council BulIe- tin 56, 1937. . "The Meaning and Measurement of Partial and Dis- guised Unemployment." The Measurement and Behavior of Unemployment. A Report of the National Bureau of Economic Research. Princeton: Princeton University Press, 1957. 129 130 Eaton, Isabel. "A Special Report on Negro Domestic Service in the Seventh Ward Philadelphia." W. E. Burghardt DuBois. The Philadelphia Negro. Philadelphia: University of Pennsylvania, 1899. Eckaus, R. S. "Economic Criteria for Education and Train— ing." The Review of Economics and Statistics, LXVI (May, 196A), 181-90. Erickson, Ethel. Employment Conditions in Beauty Shops. Women's Bureau Bulletin, No. 133. Washington, D. C.: Government Printing Office, 1935. Farrar, Donald E., and Glauber, Robert R. "Multicollinearity in Regression Analysis: The Problem Revisited." The Review of Economics and Statistics, XLIX (February, 1967), 92—107. Ginsberg, Eli et al. Occupational Choice, An Approach to a General Theory. New York: Columbia University Press, 1951. Hodge, Robert W., Siegel, Paul M., and Rossi, Peter. "Occupational Prestige in the United States: 1925- 1963." Class,_Status, and Power.. Edited by Reinhard Bendix and Seymour Martin Lipset. 2nd ed. New York: The Free Press, 1966. Laughlin, Gail. "Domestic Service." Report of the Indus- trial Commission on the Relations and Conditions of Capital and Labor Employed in Manufactures and General Business. Washington, D. C.: Government Printing Office, 1901. Lindauer, John H. "The Accuracy of Area Unemployment Estimates Used to Identify Depressed Areas." Indus- trial and Labor Relations Review, XIX (April, 19665, 377-89. Manpower Administration. Dictionary of Occupational Titles, 1965. Vol. I: Definition of Titles. 3rd ed. Washington, D. C.: Government Printing Office, 1965. . Dictionary of Ocoopational Titles, 1965. Vol. II: Occupational Classification. 3rd ed. Washington, D. C.: Government Printing Office, 1965. . Selected Characteristics of Occopations (Physical Demands, WorkingConditiono, Training Time) 1966. A Supplement to Dictionary of OccupationalvTitles, 1965. Washington, D. C.: Government Printing Office, 1966. 131 Manpower Roport of the President. Washington, D. C.: Government Printing Office, 1967. Mincer, Jacob. "Labor Force Participation of Married Women." Aspects of Labor Economics. A Report of the National Bureau of Economic Research. Special Conference Series No. 14. Princeton: Princeton University Press, 1962. . "Labor-Force Participation and Unemployment, A Review of Recent Evidence." Prosperity and Unemploy- ment. Edited by Robert Aaron Gordon and Margareth. Gordon. New York: John Wiley and Sons, Inc., 1966. . "Market Prices, Opportunity Costs, and Income Effects." Measurement in Economics,_Studies in Mathematical Economics and Econometrics in Memory of Yehuda Grunfeld. By Carl F. Christ et al. Stanford, Calif.: Stanford University Press, 1963. National Committee on Household Employment. A non-profit organization with headquarters in Washington, D. C. A Handbook for Leaders. n.d. 1965 Handbook on Women Workers. Women's Bureau Bulletin, No. 290. Washington, D. C.: Government Printing Office, 1966. "Opposition of Organized Labor to the Tipping System." Monthly_Labor Review, XXV, No. A (October, 1927),,1—u. Pigou, A. C. The Theory of Unemployment. London: Macmillan and Co., Ltd., 1933. Reiss, Albert J., Jr. et a1. Occupations and Social Status. New York: Free Press of Glencoe, Inc., 1961. Salmon, Lucy Maynard. Domestic Service. 2nd ed. New York: Macmillan Company, 1911. Scoville, James G. "Education and Training Requirements for Occupations." The Review of Economics and Sta- tistics, XLVIII (November, 1966), 387-95: "The Servant Problem." Fortune, XVII (March, 1938), 81-85ff. Smith, Ethel M. "America's Domestic Servant Shortage." Current History, XXVI (May, 1927), 213—218. Stigler, George J. Domestic Servants in the United States 1900-1940. Occasional Paper No. 24. New York: National Bureau of Economic Research, 1946. 132 . Trends in Employment in the Service Industries. Princeton: Princeton University Press, 1956. Strand, Kenneth and Dernburg, Thomas. "Cyclical Variation in Civilian Labor Force Participation." The Review of Economics and Statistics, XLVI (November, 196A), 378-391- Training in Service Occupations Under the Mappower Develop: ment and Training Act. Manpower Research Bulletin, No. 9. Washington, D. C.: Government Printing Office, 1966. Ullman, Joseph C. "How Accurate are Estimates of State and Local Unemployment?" Industrial and Labor Relations Review, XVI (April, 1963), 433-452. U. S. Bureau of the Budget. Standard Industrial Classifica— tion Manual, 1967. Washington, D. C.: Government Printing Office, 1967. U. 8. Bureau of the Census. 1960 Census of Population, Alphabetical Index of Occupations and Industries. Washington, D. C.: Government Printing Office, 1960. . 1960 Census of Population,_C1assified Index of Occupations and Industries. Washington, D. C.: Government Printing Office, 1960. U. S. Department of Labor. Job Descriptions for the Cleaning, Dyeing, and Pressing Industry. Washington, D. C.: Government Printing Office, 1938. . Job Descriptions for Domestic Service and Personal Service Occupations. Washington, D. C.: Government Printing Office, 1939. . Job Descriptions for the Laundry Industry. Washing- ton, D. C.: Government Office, 1937. Verway, David I. "A Ranking of States by Inequality Using Census and Tax Data." The Review of Economics and Statistics, XLVIII (August, 1966), 313-321. "IAAAAAAAAAI