29 March 2014


A Review
George Psacharopoulos
T hat education is a form of investment that can contribute to
individual and social development is not a novel idea. Over
two centuries ago Adam Smith wrote:
A man educated at the expense of much labor and time... may be
compared to one... expensive machine.. . The work which he
learns to perform.. . over and above the usual wages of common
labor will replace the whole expense of his education (1776, p. 101).
Articles on education as investment appeared sporadically in the
first half of this century (for example, Strumilin
1929, Walsh 1935). But it was not until the late Figure 1
1950s that the subject became a separate field of
study-the economics of education. The spur was \
the realization that not all the increases in na- radeipa <
tional output could be accounted for by the tainment
growth of conventional inputs: physical capital, ttainment
labor, and land. The "residual" puzzle in growth m Earnin
accounting was solved by Schultz (1961a) and akAbility Lernn
others, who introduced human capital into the utcome
aggregatep roductionf unction. Cognitive
One way of analyzing the complex links between
education and development is illustrated in
figure 1. The direct relationship depicted by arrow 4 from school
to the labor market-is known as the "external efficiency of education"
and has received most attention in the literature, mainly from
economists. The triangular path (arrows 1, 3, and 6) from family
background to schooling and learning outcomes is known as the
(? 1988 The Intemational Bank for Reconstruction and Development/The World Bank 99
This content downloaded on Fri, 25 Jan 2013 14:43:08 PM
All use subject to JSTOR Terms and Conditions
"internal efficiency of education" and has been studied mainly by
sociologists and psychologists. In recent years, the availability of longitudinal
data sets has permitted some attempts to estimate the full
The first section in this article presents some of the principal findings
on the role of education in development that have been generated
in the past twenty years. The following section deals with a series of
debates that have appeared in the literature. The final section attempts
to draw some policy implications for educational priorities in
developing countries.
The Evidence This review is restricted to the effects of education most directly
related to a common notion of economic development. One is efficiency
in resource utilization, leading to higher income; another is more
equitable distribution of such income.
Labor Market Outcomes
There is widespread evidence that an individual's earnings increase
with each extra year of schooling. The explanation of human capital
theory is that education makes the individual more productive, not
only in the market place but also in the household.
As Welch (1970) and Schultz (1975) have put it,
education has a beneficial allocative effect or helps
the individual to deal with disequilibrium situations.
The empirical core of the human capital school
lies in the crossover shown in figure 2-the tradeoff
between a low level of education and earnings
today versus more education and earnings tomorrow.
Such a relationship has been documented
in practically every country that has data on ageearning
profiles by education. (See the appendix
for examples from the latest Brazilian census.)
Proper discounting of the costs and benefits associated with educational
investment (the minus and plus areas in figure 2) leads to
estimates of its profitability from both the private and social point of
view. Like the rate of return to any other project, the return to
educational investment is the discount rate that sets the net present
value of the net stream of benefits equal to zero. In a state subsidized
system the cost of education to the individual is the earnings forgone
while in school. From a social viewpoint, however, the cost of education
must include all resources used to provide education (for example,
teachers' salaries and the use of classrooms).
Figure 2
More education i- +
Less education
100 Research Observer 3, no. 1 (January 1988)
This content downloaded on Fri, 25 Jan 2013 14:43:08 PM
All use subject to JSTOR Terms and Conditions
Estimates of the returns to education are now available for over
sixty countries. Table 1 gives a regional summary and raises several
noteworthy points.
* The social returns to education in developing countries are at least
as high as any reasonable measure of the opportunity cost of capital
or a social discount rate. In other words, investment in people may be
more conducive to economic growth than investment in machines.
* Rates of return are highest in primary education, followed by
secondary and then university levels. For primary education, unit
costs are small relative to the extra lifetime income or productivity
associated with literacy. For university education, the opposite is
* The same diminishing returns apply across countries: the more
developed the country, the lower the returns to education at all
levels. The high returns to education in low-income countries must
be attributed to their relative scarcity of human capital.
* Private returns are higher than social returns at all levels-a result
of the public subsidization of education in most countries. The
discrepancy between private and social returns is greatest at university
level-which raises issues of equity as well as of how educational
expansion should be financed.
Other studies have estimated rates of return by field of study and
by sex. Table 2 shows that technical education and agronomy are
associated with lower returns than the more general subjects-a counterintuitive
result that is again due to the relatively high unit cost of
technical education (Psacharopoulos 1987a). And table 3 shows that
the rate of return is higher on women's education than on men's.
Although in all societies the absolute earnings of men are higher, the
opportunity. cost of study for women is often lower than for men.
Post a Comment