STOCKS AND DEPRECIATION OF HUMAN CAPITAL: NEW EVIDENCE FROM A PRESENT-VALUE PERSPECTIVE

Responding to a perceived growing interest in human wealth estimates, this paper offers a framework for measuring the aggregate stock of human capital and then implements the procedure for the United States male population age 14 to 75. Unlike previous estimates of human wealth that are based upon historical or resource costs, these estimates measure the capital stock as the discounted resent-value of expected lifetime returns. In the estimation, returns are equated with earnings data from the 1970 U.S. Census 15 percent Public Use Sample for out-of-school males, adjusted for employment and survival probabilities, adjusted for an assumed exogenous growth in future earnings, and discounted at 7.5 percent.
We provide cross-sectional estimates of individual stocks of human capital by age and educational attainment, as well as expected lifetime wealth profiles for individuals by level of education. These individual profiles can be used to obtain direct estimates of age-specific depreciation which suggest human capital is subject to significant and prolonged appreciation before nearly straight-line depreciation begins around middle age. This finding is all the more significant since resource-cost estimates of human capital which must assume a depreciation pattern to obtain stocks have always imposed a much faster rate much sooner.
Finally, an aggregate estimate of the stock of human capital for all males is supplied and its sensitivity to the choice of the discount rate, tax laws, and expected exogenous growth is analyzed. This seemingly-conservative stock estimate is then compared to a much lower resource-cost estimate offered recently by John Kendrick. A discount rate over 20 percent would be needed to equate the two measures. In trying to reconcile the two figures, we raise some new questions about the validity of both approaches for human capital accounting.     

HUMAN CAPITAL INVESTMENT IN HEALTH: A MEASUREMENT FRAMEWORK AND ESTIMATES FOR THE UNITED STATES, 1952–78

Lack of a conceptual basis for measuring human capital investment in health has hampered efforts to expand national accounting systems to include human capital investment. This paper presents a conceptual basis for developing estimates of this health investment, an estimation methodology consistent with the conceptual basis, and preliminary estimates for the United States for 1952-78. While much work remains to be done before comprehensive estimates of investment in health are achieved, it is clear that previous estimates based on answers to the question, “What improves health?” have included some inappropriate expenditures while excluding others that should be included. The conceptual basis presented here leads to a methodology for separating health care costs (not the costs of illness) into maintenance and gross investment. Gross investment can be further separated into net investment and the sum of damages and depreciation but empirical implementation of this step is not attempted here.     

THE ACCOUNTING TREATMENT OF HUMAN INVESTMENT AND CAPITAL

Defining investment as outlays that increase income- and output-producing capacity, the author presents estimates of human investment in the United States 1929–69, comprising rearing costs, education, training, health, safety and mobility outlays. He develops an economic accounting framework to accommodate human investments and research and development in national and sector capital accounts, with appropriate adjustments to the current accounts to provide consistency. The associated balance sheets and wealth statements are also developed.
The wealth and corresponding income estimates are used to compute rates of return on human, non-human, and total capital. In the business economy the average net rate of return on total capital was 10.6 percent in 1969, compared with 10.0 percent in 1929. The average and marginal rates of return on human capital were generally somewhat higher than on non-human capital throughout the period.     

EDUCATION AND THE NATIONAL WEALTH OF THE UNITED STATES*

Human capital theory has motivated a great many empirical investigations into the relationship between education and earnings potential. These studies test refinements of the theory, but do not attempt to value education for the economy as a whole. This study develops series which track human wealth and its educational components for the United States from 1946 to 1980. Three related educational time sequences emerge: (1) schooling wealth, the present value of the current and future contributions of the existing schooling stock to national income; (2) net change in schooling wealth, the amount added to present value in that year; and (3) schooling investment, the present value of the future contributions of the new schooling conducted in that year. One important lesson of this exercise is that the last two series can be quite different as a result of the pattern of appreciation and depreciation of human wealth over the lifetimes of individuals. Moreover, education increases the age of peak human wealth and thus should shorten the period during which individuals save for retirement. This phenomenon may induce a demographic cycle in the nation's savings rate, especially evidenced with the aging of the baby-boom cohort. The magnitudes of the human and schooling wealth estimates are large when compared to financial wealth. For a 4 percent rate, the period-wide average for human wealth is five times-and schooling wealth 2.6 times-the Federal Reserve Board's measure of household net worth. These estimates are naturally sensitive to the discount rate chosen, but show that the gap between human and financial wealth has been widening and that the value of schooling provided in any year greatly exceeds its cost. Schooling represents a form of saving whose value is several times the conventional measure of saving.     

SUGGESTIONS FOR TREATMENT OF HUMAN CAPITAL IN NATIONAL ACCOUNTS—WITH ILLUSTRATIONS FROM INDIAN DATA*

After giving a brief discussion of the biases that exist in the conventional estimation procedures followed in the construction of national accounts, this paper argues for restructuring of national accounts so as to treat human capital formation as investment rather than consumption and suggests that a beginning should be made in respect of schooling. The argument is based on the notion that “investment” or “capital” is that which yields future income streams and also on the rather obvious point that treating as consumption large outlays that really constitute investment distorts analyses of resource allocation, growth and income distribution, and obscures intersectoral relations. It is pointed out that the proposed restructuring of national accounts is more relevant and important for developing countries, many of which are embarked on investment planning. Another major point emphasized is that the input of students’time should be properly measured and included in the estimates of capital formation by schooling. To illustrate what these proposals imply, revision has been attempted of the estimates of (a) educational outlay (or activity in the education sector), (b) gross capital formation, and (c) gross national product, pertaining to the national accounts of a major developing country, namely India, for the years 1960–61 through 1965–66. The modified estimates, though first approximations and covering only a part of the human capital formation and having a systematically downward bias, nevertheless indicate an upward revision of the estimate of activity in the education sector by about 200 to 300 percent, of gross capital formation by about 50 percent and of the gross national product by 4 to 7 percent. These magnitudes show the substantial order of distortion involved in the conventional procedures.     

Four decades of Asian American women's earnings: Japanese, Chinese, and Filipino American women's earnings 1960–1990

This article examines the earnings progress of Asian American women from 1960 through 1990 by comparing their actual hourly wage and salary earnings to simulated earnings. The simulated earnings are obtained by using parameter estimates obtained from human capital models of white women corrected for sample selection bias. Data come from the decennial census Public Use Micro Samples data. American-born Asian American women appear to have made dramatic gains in the 1970s. The 1980s and 1990s show some fluctuations in actual earings relative to simulated earnings between Asian American and white women. These fluctuations may be due to problems measuring experience as opposed to differences in discrimination over time.     

DEPRECIATION, OBSOLESCENCE, AND THE MEASUREMENT OF THE AGGREGATE CAPITAL STOCK OF THE UNITED STATES 1900–1962

This paper is concerned with the sensitivity of estimates of the aggregate capital stock of the United States to the statistician's choice of depreciation method. The usual depreciation charge can be shown to include allowances both for physical deterioration and for obsolescence. If one interprets the gross stock as the stock of surviving assets, then the various net stocks defined by depreciation accounting may be interpreted as a revaluation of these assets by means of an index of embodied technical change. Estimates of the United States capital stock were generated under eight sets of assumptions. These estimates are compared with respect to level, trend, and implications for other aggregate statistical indicators. The conclusion is reached that the assumptions which define a country's stock of tangible capital are of considerably greater importance than has often been supposed.     

Schooling,experience and earnings: The case of an LDC

This is an application of the strict human capital model in accounting for income inequality in an LDC. Using individual characteristics of 1600 male Moroccan full-time employees, differences in schooling and experience explain about 70 percent of relative earnings dispersion. This result is based on the existence of an 'overtaking year of experience' occuring within the first decade of the working life of the individual. Furthermore, an attempt is made to isolate the rate of return to training from the returns to schooling by analysing the earnings of illiterate manual workers differentiated by the level of their skill. The results regarding the relationship between the returns to schooling versus training, the overtaking point, and the explanatory power of human capital variables are remarkably similar to those obtained in advanced countries.     

On-the-job human capital accumulation in a real business cycle model: Implications for intertemporal substitution elasticity and labor hoarding

A substantial fraction of a worker's time at work goes to acquiring human capital. This paper explicitly considers on-the-job human capital accumulation from the perspective of time invested for acquiring skills and learning by doing in an RBC model and shows that the inability to account for human capital accumulation leads to a substantial bias in conventional estimates of intertemporal substitution elasticity.Our main results are based on the standard intuition that the opportunity cost of time invested in acquiring human capital moves procyclically, so that on-the-job time invested in acquiring human capital is “counter-cyclical.” Furthermore, the true wage rate becomes less procyclical, while production hours become more procyclical than total hours at work. The overall results can be viewed as providing a micro foundation for labor hoarding models without adjustment costs.     

Education and income of the states of the United States: 1840–2000

This article introduces original annual average years of schooling measures for each state from 1840 to 2000. Our methodology results in state estimates similar to those reported in the United States Census from 2000 back to 1940 and national, turn of the century estimates strikingly close to those presented by Schultz (Schultz, T. (1961). In N. B. Henry (Ed.), Social forces influencing American education. Chicago: University of Chicago Press.) and Fishlow (Fishlow, A. (1966). In H. Rosovsky (Ed.), Industrialization in two systems. John Wiley & Sons). To further determine the validity of our state schooling estimates, we first combine original data on real state per worker output with existing data to provide a more comprehensive series of real state output per worker from 1840 to 2000. We then estimate aggregate Mincerian earnings regressions and discover that the return to a year of schooling for the average individual in a state ranges from 11% to 15%. This range is robust to various time periods, various estimation methods, various assumptions about the endogeneity of schooling and is in line with the body of evidence from the labor literature. All views expressed here are the authors’ and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System.     

Optimal Schooling Investments and Earnings: An Analysis Using Australian Twins Data

The potential impact that unobserved ability can have on both schooling and earnings outcomes has been modelled by Ashenfelter and Rouse (1998). This paper applies their model to the Australian Twins Sample. The best estimate of the return to schooling for genetically identical (or MZ) twins is 8.9 per cent. Measurement errors in the schooling data are shown to constitute a more serious problem than the omission of the family effects that impact on ability. Individuals from higher ability families receive a lower marginal benefit from their investment in human capital.     

THE TREATMENT OF EDUCATION IN NATIONAL ACCOUNTING1

It is argued that the conventions of an accounting system, such as the S.N.A., are a matter of convenience. The treatment of education as a current expenditure, instead of as a form of capital formation, derives from the Keynesian system, and is not appropriate for dynamic problems of developing countries, where weaknesses in education are often the main “bottleneck” in the process of development. In such countries, expenditure on education clearly yields its benefits mainly in the longer run. To treat this as a consumption item biases policy in the direction of using financial resources for fixed capital rather than human investment, and may cause aid agencies to penalize countries which expand their educational systems. A similar problem arises on other expenditures such as health, but the case for treating them as investment is not so strong. To treat educational expenditure as part of capital formation logically requires two major changes. First education needs to be removed from private and public consumption, and for this purpose a fairly broad definition of what is education should be used. Secondly, the stock of educational capital should be valued. The valuation problems are, however, severe. Variations in cost components make historic cost of little value as a yardstick, and calculations of future returns are fraught with difficulties. Using replacement costs, which seems the best method, involves the construction of education profiles in physical terms which can then be valued by present or by standardised costs. The depreciation of human capital through mortality and retirement can be allowed for by applying national average rates to these physical profiles. Switching educational expenditure from current to capital accounts involves no serious practical problem. However, although there should logically be an allowance for depreciation on human capital, this is not recommended; single monetary measures of educational stock are not very meaningful, and this would involve changing the definition of “net” aggregates. Development of statistics of educational stocks and flows in physical terms—the beginnings of “demographic accounting” fully integrated with the rest of national accounting—is strongly advocated.     

POSTSCHOOL LEARNING AND HUMAN RESOURCE ACCOUNTING

The central concern of this paper is with the treatment of human resources in dynamic applications of capital and growth accounting. Despite many advances, national accounting conventions still give biased profiles of the economy, but the time is ripe for experimentation with measures that can correct those biases and provide a more adequate base for assessment of long-term economic performance and prospects.
In the first section, the logic and feasibility of forward and backward measures of formation of human capital in the simplest case (of full-time schooling) is examined in parallel with physical capital. In a dynamic economy, which is rarely if ever in equilibrium, these approaches complement each other; they are poor substitutes. In section two a number of conceptual and measurement issues are considered with particular reference to human-capital investment periods and the treatment of appreciation, depreciation and obsolescence of human versus physical capital. Here special attention is given to the extended periods of investments in human resources, which overlap with realization of returns, and to the processes and agencies through which postschool investments are made. The last section presents a brief statement concerning asymmetries in disequilibrium biases with respect to the formation of human relative to physical capital. Drawing upon section 1 with regard to forward and backward measures and section 2 with regard to the critical importance of postschool learning, new possibilities in contributions of national accounting to a dynamic analysis of economic development are suggested.     

Another look at the linear q model: an empirical analysis of aggregate business capital spending with maintenance expenditures

Abstract .  The paper revisits the empirical investment literature, which has established that aggregate business fixed investment is not found to be related linearly to marginal or average Tobin's q . The theoretical background is extended here by developing a supply-side model where the depreciation rate of private capital is determined endogenously. The firm can either invest in 'new' capital, which adds directly to the existing capital stock at the presence of convex adjustment costs, or extend the durability of installed capital through maintenance expenditure, which affects its depreciation rate. The model shows that Tobin's q is then a positively related sufficient statistic for both components of aggregate capital expenditures. This central implication is tested empirically using aggregate time-series survey data from Canada on 'new' investment and maintenance expenditures covering the period 1956–93. The estimated relationships produce significant and plausible parameter estimates for the structural parameters of the q model.     

TOTAL INCOMES IN THE UNITED STATES, 1946–1976: A SUMMARY REPORT

Extended accounts of total income and product and associated capital stocks for the United States, in current and constant dollars, are offered for the years 1946 to 1976. They include intangible and tangible capital accumulation and non-market and market outputs in all sectors, services of government and household capital and of unpaid household labor, and opportunity costs of students. Defense and police services are classified as intermediate product; a portion of commercial media services is counted as final product. Expenses related to work are subtracted while the values of employee training and human capital formation and net revaluations of existing tangible capital are added.
Total incomes (TISA) net national product was 50 percent greater than official Bureau of Economic Analysis (BEA) NNP in 1976. BEA gross private domestic investment was only about 18 percent of TISA gross capital accumulation. Intangible investment and TISA net domestic capital accumulation grew more rapidly than BEA net private domestic investment. Household investment has been growing while there have been sharp declines in government investment, particularly in research and development. Contrary to some views of the import of the narrower BEA accounts, total capital accumulation appears to have risen considerably more rapidly than total consumption, 6.3 percent versus 2.2 percent per annum from 1946 to 1976, thus increasing sharply its share of TISA GNP.     

Education, credentials, and immigrant earnings

Abstract.  Using 1981 to 2001 Census data, we study how the human capital of immigrants is rewarded in Canada. We distinguish between years of schooling and degrees obtained in order to estimate 'sheepskin' effects – the gain in earnings associated with receipt of a degree, controlling for years of schooling. We find that immigrant years of schooling and immigrant work experience accumulated before arrival is valued much less than Canadian experience of comparable natives. However, for immigrants the increase in earnings associated with completing educational programs is generally higher than that of comparable natives. We provide both signalling and human capital interpretations of this finding.     

Estimating returns to education using twins in urban China

This paper empirically estimates the return to education using twins data that the authors collected from urban China. Our ordinary least-squares estimate shows that one year of schooling increases an individual's earnings by 8.4%. If we use a within-twin fixed effects model, the return is reduced to 2.7%, but rises to 3.8% after the correction of measurement error. These results suggest that a large portion of the estimated returns to education is due to omitted ability or the family effect. We further investigate why the true return is low and the omitted ability bias high, and find evidence showing that it may be a consequence of China's education system, which is highly selective and exam oriented. More specifically, we find that high school education may mainly serve as a mechanism to select college students, but as a human capital investment per se it has low returns in terms of earnings. In contrast, both vocational school education and college education have a large return that is comparable to that found in the United States.     

Human Capital and Growth Cycles

This paper studies the dynamic interaction between human capital accumulation and economic growth. Capital market imperfections and an indivisibility in human capital investment prevent poor agents from accumulating skills, the acquisition of which positively affects technological progress. More productive technologies in turn require more sophisticated qualification and involve higher training costs. The equilibrium dynamics can be characterized by the joint evolution of productivity growth, the schooling costs, and the income distribution. Under our assumptions, individual incomes follow a non-linear Markov chain. This non-linearity generates endogenous fluctuations of schooling activities and the rate at which productivity improvements occur. We thank an anonymous referee for many helpful suggestions. Support from the German Research Foundation (DFG) under grant KA1519/2-2 is gratefully acknowledged.