Advertisement

College-Going Behavior as Investment in Human Capital

  • Kazuhiro Arai

Abstract

One of the theories dealing with an individual’s college-going decision making is the theory of human capital. It assumes that higher education is an object of investment, which enhances the individual’s productivity. This theory was formed in earnest around 1960 with the growing interest in economic growth and income distribution, and was gradually refined from that time. Together with this theoretical refinement, a large number of empirical studies have been undertaken in related areas. Today, the theory of human capital is one of the most fundamental concepts in labor economics.

Keywords

High Education Human Capital College Graduate Market Rate Human Capital Theory 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes

  1. 1.
    The basic literature (or survey literature) which discusses various aspects of education investment from the viewpoint of the theory of human capital (as one of the viewpoints) includes Schultz (1963, 1971), Becker (1964, 1993), Mincer (1970, 1974), Blaug (1966, 1968, 1970, 1976 ), Cohn (1972), Psacharopoulos (1973), Rosen (1977), McMahon and Geske (1982), Freeman (1986), and Schultz (1988).Google Scholar
  2. 2.
    The monthly earnings include tax and various allowances such as commuting allowances, family allowances, and overtime allowances. This data is collected every June.Google Scholar
  3. 3.
    Amano (1995) points out that education has both “a visible curriculum” of classroom activi¬ties and “an invisible curriculum” of extracurricular activities which is important for person¬ality development. He suggests that the age has come when we need to understand consciously these two curricula as important parts of education.Google Scholar
  4. 4.
    Welch (1970) indicates the following two effects of education. One is to increase the total output from given resources such as materials, time, and so on. The other is to enable the individual to devise an efficient combination or allocation of production inputs. He says that the latter effect is important to American farmers, because they have a high degree of discretion in the combination or allocation of inputs and this decision must be made con¬tinually. Nelson and Phelps (1966) emphasize that education eases adaptation to technologi cal change. According to Wozniak’s (1987) empirical study on American farmers, education reduces the adoption cost of new technology and uncertainty about the outcome. The empiri¬cal study by Bartel and Lichtenberg (1987) on US manufacturing industries shows that the demand for highly educated workers is large in industries characterized by high rates of innovation, because those workers have a comparative advantage with respect to the adjust¬ment to and implementation of new technologies. Schultz (1975) emphasizes that the acqui¬sition of “allocative abilities” is an important purpose of education. Those who have allocative abilities can perceive, interpret correctly, and undertake action that will appropri¬ately reallocate their resources in response to changes in economic conditions. In his empiri¬cal study on American farmers, Huffman (1977) shows that education actually has such a function. See also Ram (1980).Google Scholar
  5. 5.
    A few interesting findings apply here. Wise (1975) and Jones and Jackson (1990) show in their empirical studies that not only the academic background of college education but also academic records in college have a positive effect on individual earnings. (In economics, a wage is considered to represent individual productivity. See Appendix A.) Hansen, Weisbrod, and Scanlon (1970) have undertaken an empirical study on those who were rejected for military service in the United States. They estimate the relationship between individual earnings and scores in the Armed Forces Qualification Test, which covers word knowledge, arithmetic, mechanical understanding, and ability to distinguish forms and patterns. This study does not directly relate to higher education, but is still informative to some extent in the present context. It shows that the AFQT score was a more important determinant of earnings than years of schooling, though there is a positive correlation between the test scores and the years of schooling. This result implies that the knowledge and skills one actually possesses are more important determinants of earnings than the mere fact of spending time in school.Google Scholar
  6. 6.
    While a wage rate stands for the price of labor service per hour, earnings stands for the reward paid in a certain period of time such as a month or a year, and includes various allowances such as overtime pay, family allowances, bonuses, etc. What is often called wages in Japan are mostly earnings (Nakamura, 1981). In such a case, wages and earnings are substitutes. However, we usually say forgone earnings rather than forgone wages. Economic theory mainly uses the terms of a wage rate and a wage, the latter of which equals the former times the number of hours worked. When the work hours per period is fixed, there is only a difference in measurement between a wage rate and a wage, and thus the two are essentially the same. An income is the sum of earnings and non-labor incomes such as dividend and interest incomes. In empirical studies, family incomes of workers are often represented by family earnings.Google Scholar
  7. 7.
    The opportunity cost is defined as the benefit one loses when one does not use or forgoes an economic opportunity. If one owns land but does not use it, an opportunity cost will arise, because making it a parking lot will generate incomes. If one keeps in one’s safe some money not for immediate use, an opportunity cost will arise again, because lending it to a bank will generate interest.Google Scholar
  8. 8.
    A competitive market is often (implicitly) assumed in discussions in economics. In a com¬petitive labor market, an individual’s wage (rate) equals his/her marginal value product (see Appendix A). Thus, the wage paid to an individual represents how much value he/she produces in the economy.Google Scholar
  9. 9.
    See Weisbrod (1962) for external economies of education. It should be noted, however, that he includes what cannot be called external economies in the rigorous sense of economics.Google Scholar
  10. 10.
    Ben-Porath derives similar properties by introducing a production function of human capital into his work life model which takes into consideration not only investment in school education but also investment in training on the job. Here, the production function of human capital for a particular individual means a function that relates the amount (hours) of his/her human capital service and the amount of purchased input (materials, teachers’ service, etc.) used for production of human capital to the amount of human capital produced.Google Scholar
  11. 11.
    A fundamental proposition of economics is that the allocation of resources in the market will be optimal if each individual (economic agent) acts according to his/her own self-interest. The optimality concept used in this proposition is called Pareto optimality. An allocation of resources is Pareto optimal if an individual cannot be made better off without making another individual worse off. It should be added that this proposition holds only when a number of conditions are satisfied, one relevant condition being that there is no government. Another is that there are no external economies nor diseconomies. It may be possible, however, that the allocation of resources becomes optimal with some of these conditions not satisfied, if the government applies suitable policies.Google Scholar
  12. 12.
    See Chap. 3 of Suzuki (1989).Google Scholar
  13. 13.
    The average annual amount of tuition and fees for private universities was 977,840 yen in 1994 (Ministry of Education, Abstract of Education Statistics ). Information about the Japanese yen exchange rate is available in Appendix C.Google Scholar
  14. 14.
    Very recently, a bank began to offer secured loans of up to thirty million yen with a repay¬ment term of thirty years. (As we see in Chap. 4, one needs very large funds to go to medical or dental school in Japan.) All these types of loans may be only temporary responses to the recent monetary ease, but may also continue to exist in the future.Google Scholar
  15. 15.
    In the United States, there are a large number of dropouts and divisibility of education is actually utilized. In fact, college dropouts earn more than high-school graduates. Additional investment here means an increase in investment in education, e.g., of 10,000 yen or by a week. The marginal private rate of return should theoretically be computed from this incre¬mental cost and the corresponding benefit. Because this is a theoretical argument, each kind of school education is supposed to be minutely divisible.Google Scholar
  16. 16.
    Moreover, Behrman, Pollak, and Taubman (1989) show that there is an inverse relationship between sibship size and sib schooling and earnings similarities. This is because intelligent children with large sibship size have easier access to financial aids and face especially low education costs. This tendency becomes weak if generous financial aids are offered equally as in the case of World War II veterans, who were eligible for educational benefits through the GI Bill.Google Scholar
  17. 17.
    See also Hansen (1963), Slang (1965, 1970), Hines, Tweeten, and Redfern (1970) for how to measure internal rates of return.Google Scholar
  18. 18.
    According to Bailey and Schotta (1972), the private rate of return to graduate school educa¬tion in general is less than one percent. One reason why some go to graduate school even at this low rate of return is that this education generates benefits in the form of future consump¬tion. However, Figa-Talamanca (1974) and Tomaske (1974) are critical of this study.Google Scholar
  19. 19.
    Using another data set, he further considers the following three working conditions: whether or not a worker has discretion in increasing his/her work hours, whether or not a worker has discretion in decreasing his/her work hours, and the stability of income. The results obtained are similar to those in the text.Google Scholar
  20. 20.
    This estimation uses Mincer’s earnings function, which will be discussed below. See also Scully (1979).Google Scholar
  21. 21.
    Haveman and Wolfe (1984) classify benefits of education into twenty categories including that of enhancing productivity on the job. They also allow that more education tends to generate costs such as divorce and work stress.Google Scholar
  22. 22.
    Weiss (1972) compares the variation coefficient of incomes of scientists with Ph.D. degrees with that of incomes of those with B.A. degrees and finds that the former is smaller. The variation coefficient is defined as the standard deviation divided by the mean.Google Scholar
  23. 23.
    Economics usually assumes that workers are risk-averse. This means that given a choice of receiving a wage of 10 million yen with a probability of 50% and a wage of 6 million yen with a probability of 50% on one hand, or a definite wage of 8 million yen, the mean of the previous option on the other hand, a risk-averse individual prefers the latter. Moreover, he/ she prefers the state in which he/she will obtain a wage of 8 5 million yen with a probability of 50% and a wage of 7 5 million yen with a probability of 50% to the first random wage above, though the means are the same. This implies that the first random wage must be evaluated by deducting a larger premium (risk premium) from the mean wage of 8 million yen. See Olson, White, and Shefrin (1979) for rates of return to higher education which take account of risks. Even after considering risks, the rates of return remain quite high, though this result depends on the assumptions used.Google Scholar
  24. 24.
    Other studies which show that the effect of IQ on incomes is small relative to the effect of school education include Ashenfelter and Mooney (1968), Hansen, Weisbrod, and Scanlon (1970), Hause (1972, 1975), Griliches and Mason (1972), Taubman and Wales (1973), de Wolfe and van Slijpe (1973), Griliches (1977), and Cohn and Kiker (1986).Google Scholar
  25. 25.
    According to the estimate by Willis and Rosen (1979), while a typical college graduate’s rate of return to higher education equals 9.9%, the rate of return that a typical non-college graduate could achieve by graduating from college would be 9.3%.Google Scholar
  26. 26.
    Making use of US Navy data of warships and their crews, Horowitz and Sherman (1980) provide a rare empirical analysis of the relationship between human capital and realized productivity. It shows that those crews with high-school education and higher entry test scores tend to keep a ship in better operational condition. This tendency is especially promi¬nent where technologically complex equipment is involved. ( The level of maintenance of equipment will influence the “productivity” of the warship. )Google Scholar
  27. 27.
    This specification is used very frequently in empirical studies, but there are some criticisms. The empirical study by Murphy and Welch (1990) indicates that the specification by a quadratic function does not approximate well the relationship between experience and earnings.Google Scholar
  28. 28.
    There are many works that have examined Mincer’s earnings function. Heckman and Polachek (1974) support Mincer’s earnings function to some extent as the simplest func¬tional form. Blinder (1976) clarifies the assumptions used to derive the earnings function which includes the effect of training on the job, and also considers alternative functional forms. Leibowitz’s (1974b) empirical study based on US data indicates that the assumption of k; = 1 will generate a bias. Lucas (1977b) points out that Mincer’s earnings function is derived only by operating identities and that it has little to do with the market process. Light and Ureta (1995) emphasize that the conventional earnings function yields downward biased estimates of early-career wage growth, and propose a different earnings function with an array of variables that measure the fraction of time worked during each year of the career. Other works include Wallace and Ihnen (1975), Willis ( 1986 ), Lillard, Smith, and Welch (1986).Google Scholar
  29. 29.
    Freeman uses the active job market hypothesis to explain the fact that the influence of economic changes is especially prominent in the wages and employment of young workers. According to this hypothesis, older workers in the United States tend to be protected from the direct influence of economic changes by job security, customs, etc. (In other words, they are said to have jobs in internal labor markets.) Because of this, the influence of economic changes (an increase in the number of college graduates, for instance) concentrates among young workers.Google Scholar
  30. 30.
    In the United States, the rate of return to higher education increased again in the 1980s. Katz and Revenga (1989) insist that this results from the interplay of an increase in the demand for college graduates due to the changes in technology and product demand, a decline in the growth rate of the number of college graduates, macroeconomic factors (increased openness, trade deficits, and labor market slack), changes in institutional structures (the decline in unionization), and so on. See also Blackburn, Bloom, and Freeman (1990).Google Scholar

Copyright information

© Kazuhiro Arai 1998

Authors and Affiliations

  • Kazuhiro Arai
    • 1
  1. 1.Hitotsubashi UniversityKunitachi, TokyoJapan

Personalised recommendations