Education and the Transfer of Inequality from Generation to Generation
by Peter R. Moock - 1978
Education is considered by some to be so important a determinant of individual well-being in American society that an understanding of the reasons for differences in the propensity to invest in education is crucial to the development of sound and equitable social policy. (Source: ERIC)
The author gratefully acknowledges the critical reading of this article in draft by William Garner, Hope Jensen Leichter, Victor Levine, Joyce Moock, Harold Noah, and Jeffrey Siedenberg.
Human capital economics provides a theoretical framework for analyzing the decisions of individuals to incur costs for education. The model predicts that an individual will choose to add to his or her stock of human capitali.e., to investso long as the discounted value of the returns that accrue over time exceeds or at least equals the value of the present consumption foregone.1 While there are many ways to raise productive capacitysuch as migrating (from an area with a smaller to one with a larger stock of complementary capital, physical or human)2 or marrying (which, despite the tax laws, can yield similar benefits via specialization and economies of scale)3 education is probably the clearest example of human capital investment, and one of the most significant in terms of total resources expended.4 Indeed, the enhancement of an individual's capacity to "do things" (not just market-related activities, but all things in life) is a reasonably accurate if not sufficiently specific expression of what economists and others mean when they speak of "education."5
Looking just at formal schooling and investment in learning on the job, Mincer has shown that differences in the accumulation of human capital may account for as much as 55 percent of the variation in the weekly wage rate of white, nonfarm, nonstudent American males up to age sixty-five.6 The effect of education on "household production" is more difficult to observe, but the evidence we have suggests that education enhances efficiency in nonmarket activities just as it seems to do in paid employment.7
Michael has shown that the effect of additional education on family consumption patterns is similar to the effect of additional money income;8 in other words, a family with more education than another family generally behaves as though its money income were higher, even when money income is actually the same. Leibowitz finds that the educational attainment of the mother of a preschool child is related positively to the IQ of the child when tested later in school, and to the final educational attainment of the child, and she finds these relationships in multiple regression runs in which she controls for other determinants of intellectual development, including the time of the mother spent in caring for the preschool child.9 In sum, education seems so important a determinant of individual well-being in American society that an understanding of the reasons for differences in the propensity to invest in education is crucial to the development of a sound and equitable social policy.
Within the human capital framework, the individual's propensity to invest in the formation of human capital is greater, the smaller the interest cost of financing the investment and the greater the rate of return on each dollar invested.10 As a decision-making model, the human capital framework applies less well, presumably, to children than it does to adults. The very young have not yet assumed responsibility for the way in which resourcestime and moneyare allocated. They are, typically, decision takers, rather than decison makers. For them there are few, if any, alternatives to an externally prescribed course of action.
The transition from helpless infant to responsible adult is a gradual one. There is no single developmental landmark that divides childhood from adulthood, although many communities celebrate the transition at some point in ritual fashion, as with the Christian confirmation or Jewish bar mitzvah. An important rite de passage shared in early life by virtually every modern-day American is entry into school. Although many decisons will continue to be made for the childby parents and, now, in loco parentis by teachers and other school personnelschool entrance marks the first important break with parental authority. To some extent, the individual's progress through the educational system and success thereafter lie, from this point forward, in his or her own hands.
But, by the time a child enters school and assumes partial responsibility for the future, considerable investment has already taken place. In particular, the child's parents have incurred costs in the child's behalf that will affect his or her propensity to invest in education over time. Ignoring innate differences in ability, which the author believes can never be identified with any great rigor, or measured with tolerable precision, we can assert that the greater the preschool investment of the parents in the child's behalf, the greater the individual's incentive to invest in forming additional human capital, because the rate of return on each dollar of investment will be higher, and because the financing cost usually will be lower.
The rate of return on educational investment is high for individuals from advantaged backgrounds because, as students, they are efficient relative to other individuals; they have learned how to learn. Also, at least in the past in this country, they have received a disproportionate share of available subsidies in education (scholarships, fellowships, admission to publicly funded universities), so that the price of a given course of study has been lower on average for individuals from advantaged backgrounds than for others. Finally, because of family connections, they are likely to be more successful at marketing whatever amount of human capital they acquire.
The cost of financing education (i.e., the interest cost on the capital investment) is usually lower for such individuals, which is the second reason for their greater propensity to invest. There are often family funds to cover both tuition and living expenses during the schooling period, whereas less advantaged individuals, to invest to the same extent, would need to borrow funds in the commercial capital market. Moreover, when it is necessary to borrow from sources outside the family, the rate of interest may be lower for those from advantaged backgrounds because of family collateral and reputation and because of these individuals' demonstrated scholastic ability.
What we are suggesting is that children are already highly disparate in capacity by the time they enter school at age six; and that differences in capacity at the point of school entrance, an important step in the progression toward individual emancipation, explain differences in the propensity to invest in the formation of further human capital and, ultimately, differences in income.
How would we go about measuring capacity at the point of school entrance? Since units of capacity cannot be observed directly, and since tests of ability are often considered suspect, our measure may need to be a monetary one. With adults, we could look at differences in their market wage, which can be assumed to measure differences in capacity.11 Except under special circumstances, however, children are prohibited by law from working for wages even if they, or their parents, would choose otherwise. Thus we cannot look to market value as a measure of a child's capacity but must, instead, consider the cost of producing it.
The costly inputs in the production of early human capital are the flow of purchased goods and services and the flow of parental time allocated to the child.12 Considerable research on the determinants of ability in young children has focused on the availability in the home of material resources, which are relatively easy to measure. Some recent studies, however, indicate that the quantity and quality of parental time inputs may be as important, if not more so, than the input of purchased goods and services.13
In the "classical" American family, in which sex roles are sharply definedthe father specializing in market work, the mother in houseworkthe "direct" (money) costs of child care are the responsibility of the male breadwinner, whereas the "indirect" (time) costs fall largely on the woman. This family stereotype, in which there are two parents present, the male engaged in market employment and the female fully occupied at home, can be misleading in light of current facts and figures. Given the rising incidence of marital instability in America, it is predicted that four out often children born in this decade will spend at least part of their childhood years in single-parent families.14 Moreover, as a reflection of the growing participation of all women in market work, the mother of a preschool child today is about as likely to be in the labor force as to be out of the labor force. In 1973, 48 percent of women with children under six spent at least some time in market work during the year; 16 percent worked fifty or more weeks. Taking the child as the unit, instead of the family, 45 percent of children under six were children of working mothers; 14 percent were children of mothers working fifty or more weeks.15
In other words, current data indicate considerable diversity in family organization and employment patterns. The working mother may be an aberration in some people's minds, but she is not so in reality. Despite this, however, the fact remains that significant numbers of women do stay home to care for young children, delaying, interrupting, or terminating valuable market careers. To each of these women (or to the family unit of which she is a member), the benefits of the increased time that staying home allows her to spend with the child must be worth, at least, the loss of market income; otherwise, she would choose to engage in market employment and find cheaper substitutes (e.g., day-care facilities) for her own time in child care. Any attempt, however, to cost the input of a mother's time in caring for her preschool child is fraught with methodological pitfalls.
First, the fact that a particular woman with a preschool child is not employed is not proof that she would be employed without the preschool child. Even if she had been employed up until the child's birth, the decision to bear the child and the decision to drop out of the labor force may have been made simultaneously. There is no certainty that the birth of the child caused her to quit her job. Unless it is true that the presence of the child is the sole cause of her market inactivity, then it is incorrect to count the full amount of the earnings that she forgoes as a cost of the child. But if not the full amount, then what fraction should be assigned to raising a child?
Second, since "earnings forgone" cannot, by the nature of the beast, be observed directly, some approximation of this cost element must be found. Provided that the woman has been employed sometime in the past, we might measure the current value of her time in market work by her last wage. Depending on how long it has been since she quit, however, labor demand conditions are likely to have shifted, and her market skills to have depreciated. Alternatively, we might estimate her wage by observing the average wage of other women who are currently employed and who possess all the relevant characteristics of this woman. Of course, the specification of "relevant characteristics" is a matter of some conjecture.
Third, even if we knew exactly what the woman, were she now employed, would be earning in monetary terms, we would not know her full compensation. An employee's full compensation for work performed consists of his or her own money wage, including all contributions withheld from the worker's paycheck, plus benefits (payments "in kind," such as free or subsidized meals, and "intangibles," such as job prestige or the desirability of geographic location), plus any on-the-job training that is financed out of the individual's earnings flow and the future returns to which belong to the individual. For any particular individual in any particular period, money wage may account for only a small fraction of full compensation.16
Fourth, there are almost certain to be significant economies of scale in child care. Once a family has taken on the responsibility of a first child, the cost of caring for a second child (who is reasonably close in age to the first) is presumably much reduced. What fraction, however, of the family's total outlay of resources on child care is the marginal outlay? On what basis should we allocate the full cost of the woman's time input between the two (or, where there are more than two, among the several) children?
Despite these conceptual and informational problems, studies that would quantify the costs of child rearing are not without precedent. In a study for the U.S. Commission on Population Growth and the American Future, Reed and Mclntosh estimated the average indirect costs in 1969 of a first child in the United States.17 By concentrating on first children, the authors avoid the question of how to allocate costs in the situation of a family with two or more children.
The first step in estimating market earnings forgone is to estimate market time forgone. Noting that married women of childbearing age without children under fourteen years of age performed on average 1,000 hours of market work in 1969, Reed and Mclntosh make the assumption that this same number of hours would have been worked by the typical woman who has a child under fourteen had she not been encumbered by the presence of the child. To get an estimate of work time given up typically by the mother of a child of a particular age, Reed and Mclntosh subtract from 1,000 hours the amount of time worked, as predicted by a regression equation that controls for factors such as race, age, educational attainment, and employment status of the husband.
By this procedure, they estimate work time lost by the mother due to the presence of the child to be 894 hours in the child's first year. Work time lost declines gradually during the child's preschool years, reaching 711 hours in the sixth year. In the seventh year, when the child has entered school, the figure plummets to 380 hours. In other words, the mother of a child in school spends considerably more time in market work on average (620 hours) than the mother of a preschool child (106 hours in the first year, climbing gradually to 289 hours in the sixth year) but spends less time than a married woman without children (1,000 hours).
The next step in estimating earnings forgone is to put a price on each hour lost. For this, Reed and Mclntosh use the average wage of women in 1969, and then, to get separate estimates of earnings forgone according to educational attainment, use the average wage of women in each of four educational categories.18 Espenshade updates their results by applying 1977 wage figures. According to his results, the average indirect cost of a first child from birth to age six in the United States is $20,073 currently. For mothers with an elementary school education, it is $15,347; with a high school education, $18,997; with a college education, $24,752; and with more than a college education, $31,396.19
The research of Espenshade allows us to compare the indirect (time) costs of child care with the direct (money) costs. Again following a methodology proposed by Reed and Mclntosh, Espenshade estimates the current direct costs of a first child in the United States. His figures are for the maturation period from birth through four years of college. According to his calculations, averaged across regions, the monetary outlay on food, shelter, clothing, medical care, educational services, and the like is $44,156 following a low-cost plan, and $64,215 following a moderate-cost plan.20 By prorating these averages, we can estimate the direct costs of a first child from birth to age six to be $13,197 and $20,062, following the low-cost and moderate-cost plans respectively.
The research cited here implies that the market earnings given up by a woman due to the presence of a first child typically equal or exceed the sum of direct purchases made in behalf of the child. The total costs of raising a first child to age six would appear to be somewhere in the neighborhood of $40,000 on average in the United States today. There is additional research suggesting that the marginal cost of each subsequent child is about half that of the first, or $20,000 on average.21 In short, the "price" of children today is quite astoundingly high.
Although the average family may expend the equivalent of $40,000 over six years on a first child, and $20,000 on each subsequent child, a significant minority of American families lack the resources permitting such lavish care. Even in the case of a two-parent family in which both parents are employable full time (35 hours per week) at the minimum wage ($2.30 per hour), the total of their potential pretax earnings is just $8,050 per year, $48,300 over six years. At least four-fifths of this potential earnings flow would be required for child care, to maintain (in dollar terms) one or more preschool children at the level of the national average. Given the many alternative claims on income, this family would be unable to provide child-nurturing inputs to this extent.
A child in this family would have to forgo some of the desiderata enjoyed by the typical American child. The sacrifices would include certain toys, books, and items of clothing deemed "unnecessary" or "too expensive." Living space would be at a premium in this family; there would be small likelihood of a separate room, all to the child's self, as a place for quiet and reflection. Both parents might feel compelled to continue working, even if the only alternative supervision for the child were that of an untrained or disinterested baby sitter. The nutritiousness of diet is often a neglected factor in low-SES (socioeconomic status) homes, due to a combination of ignorance and financial incapacity.
Of course, the hypothetical family just described is not destitute. In many American homes, conditions are worsethe market value of parental time is lower than in this home, and the paring of child-nurturing inputs necessarily more extensive. Particularly hard hit are single-parent families and families in which the principal breadwinner cannot find employment.
It is true that the dollar value of resources expended is not a perfect measure of effective inputs in the production of early human capital. Some families make particularly good use of resources expended. They are "efficient" producers of child care. Other families make poor use; they are ''inefficient" producers of child care. One family may lavish expensive purchases on a child, without much attention given to their instructional or nutritional value or to the danger of "spoiling" the child, while another family sees that every dollar spent counts significantly toward the child's development.
Certainly, all hours of time are not equal. Two women, commanding identical market wages, may quit jobs to care for their respective children, but the one may be an attentive and loving mother, while the other, perhaps unconsciously, resents the interruption of her market career and, by staying home, actually retards the child's development. A third woman may continue to work during her child's preschool years yet, by devoting most of her evenings and weekends to child-rearing activities, spend as much time with the child as she would have spent had she quit her job. Moreover, because she continues her career, her attitude toward the child may be more healthy than it would have been, and therefore her time with the child more productive. A fourth woman may engage in labor-force activity but come home so exhausted or preoccupied each day that she ignores, or, worse, abuses her child.
The wage rate, which reflects quite well productivity in market work, may be a poor measure of productivity in child care. A woman or man may lack the education and training necessary to secure a high-paying job but may, nevertheless, make an extraordinarily good parent. Although the odds are against it, many individuals do move from a position near the bottom of the income distribution as children to a position near the top as adults. We may assume that, in many of these cases, the parents of these individuals, despite having failed to achieve economic success for themselves, have managed somehow to confer the attributes for success on their children.
In sum, there are "qualitative" dimensions of child care that have not been measured in research to date. Despite the importance of qualitative differences, however, it remains true on average that more is better than less. To be brought up in a family with (potential) income far below the median level is to be given a handicap that tends forever after to limit an individual's productive power relative to his or her peers.
We know, from research by psychologists, sociologists, and economists, that capacity at any ageas measured by IQ or achievement test scores, educational attainment, or success in market activities-correlates positively and quite strongly with capacity at any other age. Since equality of inputs in the process of elementary and secondary school education, if not yet a reality, is now generally accepted as a goal of American society,22 the family emerges as the most logical target for assistance in any public program that would reduce inequality in America. Although norms relating to male-female sex roles and to the durability of the marriage contract are undergoing changepresumably they have always done sothe family remains the institution in which most children will spend their most formative years.
In recognition of the family's crucial nurturing role, the Carnegie Council on Children has called for a comprehensive national policy that would guarantee employment for at least one member of every family and, through a reformulated federal income tax, set a floor under every family's income, equaling approximately one-half of the median family income.23 Under this system, the full-time care of a preschool child would be treated as a legitimate alternative to market employment, giving the single parent the option of withdrawing from the labor force, an option that is not financially viable under the present system. Moreover, temporary withdrawal for purposes of child care could not, under the new laws, result in loss of job or seniority within the firm.
By implication of these recommendations, the council holds that the development of capacity in young children depends (and should depend) principally on the inputs of parental time and market purchases made out of family income, and that the particulars of child care should be left to the discretion of individual families (except in the most extraordinary of circumstances, when direct intervention may be called for). The council proposes a system that would enable poor families to provide their children with more of the one input (market purchases) without having to sacrifice any of the other input (parental time).
Poverty, which by the council's definition afflicts one quarter of all American families,24 has a debilitating effect on a child's development. As argued above, the disadvantaged child fails to acquire the learning skills that render educational investment, later in life, profitable. The problem of poverty is society's problem, though many would shift the blame onto the victims themselves. Existing programs for relieving poverty are generally demeaning, the direct descendants of seventeenth-and eighteenth-century poor laws. Moreover, partial programs directed narrowly at the symptoms of poverty are doomed to failure.
Traditionally, American efforts to eliminate poverty and increase economic equality have dealt with everything but the fact of economic inequality and the economic system that allows and perpetuates it. All of these efforts have been unsuccessful: relative poverty has not decreased in over a century in America. As long as our economic system permits millions to live in poverty and as long as our political system is not committed to the elimination of poverty, no programs of personal reform, moral uplift, blame, therapy, philanthropy, or early education can hope to eliminate the enormous harm to the next generation that poverty causes.25
Only by attacking poverty directly and at its source, the family as cradle, can the nation hope to short-circuit the process that results in the replication of unequal distribution from one generation to the next.
1 Because of the "time value of money," reflected in market interest rates, future returns (or future costs) must be "discounted to the present" to make them comparable to present flows. The roots of human capital theory can be traced to the very early writings of political economists. Adam Smith wrote in 1776 of investment in human capital (An Inquiry into the Nature and Causes of the Wealth of Nations, vol. 1 [London: Macmillan and Company, 1869], pp. 106-07) in terms that are fully recognizable in light of current theory. But only in the last 20 years has human capital emerged as a primary concern of professional economists. A landmark occasion in the development of this concern is the 1960 presidential address to the American Economics Association of Theodore W. Schultz ("Investment in Human Capital," American Economic Review 51 [March 1961]: 1-17). For a recent appraisal of the heuristic value of human capital theory, the reader is referred to Mark Blaug, "Human Capital Theory: A Slightly Jaundiced Survey," Journal of Economic Literature 14 (September 1976): 827-55.
2 Larry A. Sjaastad, "The Costs and Returns of Human Migration," Journal of Political Economy 70 (October 1962): 80-93.
3 Gary S. Becker, "A Theory of Marriage, Part I," Journal of Political Economy 81 (July/August 1973): 813-46. Reprinted in Theodore W. Schultz, ed., Economics of the Family: Marriage, Children, and Human Capital, A Conference Report of the National Bureau of Economic Research (Chicago: University of Chicago Press, 1975); and Gary S. Becker, The Economic Approach to Human Behavior (Chicago: University of Chicago Press, 1976).
4 For estimates of the full costs of formal education (schools, universities, and government training programs) alone, see Elchanan Cohn, The Economics of Education (Cambridge, Mass.: Ballinger Publishing Col, 1975), pp. 108-12.
5 We may compare the definition of a historian: . . . education [is] the deliberate, systematic, and sustained effort to transmit, evoke, or acquire knowledge, attitudes, values, skills, or sensibilities, as well as any outcomes of that effort." Lawrence A. Cremin, Public Education (New York: Basic Books, 1976), p. 27.
6 Jacob Mincer, Schooling, Experience, and Earnings (New York: Columbia University Press, 1974), chap. 5.
7 There are at least three interpretations of the positive relationship between education and earnings. First, as already noted, education may render workers more productive. Second, education may simply identify those who are, for whatever reasons, already more productive. This is the "screening hypothesis." For a discussion, see Paul Taubman and Terence Wales, Higher Education and Earnings: College As an Investment and a Screening Device, A Report Prepared for the Carnegie Commission on Higher Education and the National Bureau of Economic Research (New York: McGraw-Hill, 1974); presented in summary form in Thomas F. Juster, ed., Education, Income, and Human Behavior, A Report Prepared for the Carnegie Commission on Higher Education and the National Bureau of Economic Research (New York: McGraw-Hill, 1975). Third, education, like union membership, may sort workers into separate, noncompeting labor groups, for reasons that have nothing to do with productiveness. For a discussion of the third hypothesis, see V. Lane Rawlins and Lloyd Ulman, "The Utilization of College-Trained Manpower in the United States," in Higher Education and the Labor Market, A Volume of Essays Sponsored by the Carnegie Commission on Higher Education, ed. Margaret S. Gordon (New York: McGraw-Hill, 1974). Whereas either of the latter hypotheses could provide an explanation of the education-earnings relationships in the external labor market, neither seems to make much sense in reference to the internal (intra-household) labor market. By far the most reasonable explanation of an observed relationship between education and household production is that education makes individuals more productive.
8 Robert T. Michael, The Effect of Education on Efficiency in Production (New York: National Bureau of Economic Research, 1972); presented in summary form in Juster, Education, Income, and Human Behavior
Transfer of Inequality 739
9 Arleen Leibowitz, "Home Investments in Children," Journal of Political Economy 82 (March/April, 1974, Part II): SI 11-31; reprinted in Schultz, Economics of the Family.
10 Gary S. Becker, "Human Capital and the Personal Distribution of Income: An Analytical Approach" (Ann Arbor: Institute of Public Administration, University of Michigan, 1967); reprinted in Gary S. Becker, Human Capital, 2nd ed. (New York: Columbia University Press, 1975).
11 The capacity, at least, to do market work. One's wage, however, is only a rough measure even of this since (1) an individual's full compensation includes significant nonmonetary components (for a discussion of the importance of the nonmonetary components of full compensation, see Robert E.B. Lucas, "Hedonic Wage Equations and Psychic Wages in the Returns to Schooling," American Economic Review 67 [September 1977]: 549-58; also, Smith, An Inquiry into the Nature and Causes of the Wealth of Nations, pp. 105-06; Samuel Bowles, "Schooling and Inequality from Generation to Generation," Journal of Political Economy 80 [May/June 1972, Part II]: S237-39, and (2) during most of an individual's working life, a portion of full compensation is simultaneously reinvested in training on-the-job, an implicit transaction between employee and employer that registers explicitly in the accounts of neither (see Becker, Human Capital, chap. 2).
12 Since the child's own time has no market value, we will ignore this factor in discussing the formation of early human capital. The omission is based on the unavailability of data and cannot be justified theoretically. On the one hand, it may be true that variation in the quantity of own time across children is small, due to the absence of any market alternative. On the other hand, however, the quality of own time becomes increasingly disparate across children between birth and school entrance, as already stated, and a fully specified production function for early human capital would include the stock of human capital in period t-1 as an input in the production of capital in period t. In other words, one input in the production of capital in any period is the output of this same process in the period before. See Yoram Ben-Porath, "The Production of Human Capital and the Life Cycle of Earnings, "Journal of Political Economy 75 (August 1967): 352-65.
13 For a review of some studies by economists of the impact of parental time inputs, see Peter Moock, "Economic Aspects of the Family As Educator," Teachers College Record 76, no. 2 (December 1974): 92-104; reprinted in Hope Jensen Leichter, ed., The Family as Educator (New York: Teachers College Press, 1975).
14 Kenneth Keniston and the Carnegie Council on Children, All Our Children: The American Family under Pressure (New York: Harcourt Brace Jovanovich, 1977), p. 4.
15 U.S. Department of Commerce, Social and Economic Statistics Administration, Current Population Reports: Consumer Income (Series P-60), no. 98 (January 1975), pp. 110-15.
16 Refer footnote 11 above.
17 Ritchie H. Reed and Susan Mclntosh, "Costs of Children," in Economic Aspects of Population Change, ed. E.R. Morss and R.H. Reed, vol. II of Research Reports, Commission on Population Growth and the American Future (Washington, D.C.: U.S. Government Printing Office, 1972).
19 Thomas J. Espenshade, "The Value and Cost of Children," Population Bulletin 32 (April 1977): 1-47.
21 Boone A. Turchi, The Demand for Children: The Economics of Fertility in the United States (Cambridge, Mass.: Ballinger Publishing Co., 1975), chaps. 3 and 4.
22 The goal is articulated in several recent, state-level court cases. See, for example, Serrano v. Priest, 483 P. 2d 1241, 1244 (Gal. 1971); Robinson v. Cahill, 62 NJ 473, 303 A. 2d 273 (1973); and Rodriguez v. San Antonio Independent School District, 337 S. Supp. 280 (WD Tex. 1971).
23 Keniston and the Council, All Our Children.
24 The council's definition of poverty (income at or below one half the median family income, or $7,373 in 1974 for a family of four) diverges markedly from the official "poverty line" (arbitrarily set at $5,000 in 1974). By the government's more stringent standard, 11.6 percent of the American population (15.5 percent of American children) were "poor" in 1974. (Ibid., p. 27).
25 Ibid., p. 118