Page 72 - Education for Development:George Psacharopoulos University of Illinois at Urbana-Champaign, USA

8 ᐿ࿲ၾආӉj઺ԃٙɢඎ

question: First, the private rate of return, that compares the costs and benefits of

education as incurred by and realized by the individual student who undertakes
the investment. Second, the social rate of return that compares costs and benefits
from the country-as-a-whole or society’s point of view. The main computational

difference between private and social rates of return is that, for a social rate of
return calculation, the costs include the state’s or society’s at large spending on
education. Traditional social returns to education are called “narrow-social,” and

returns that include externalities “wide-social.” The distinction between narrow
and wide social returns is more than theoretical. By adding externalities to the
narrow-social returns, one can reach diametrically opposite policy conclusions,
e.g., if primary and tertiary education have differential externalities, by

considering the latter the ranking of profitable education investments could be
changed.

A meta-analysis of over 1000 rate of return estimates in over 100 countries
showed that primary education exhibits the highest returns, followed by
secondary and higher education (Psacharopoulos and Patrinos, 2018).
Since the costs are higher in a social rate of return calculation relative to

the one from the private point of view, social returns are typically lower than a
private rate of return. The difference between the private and the social rate of

return reflects the degree of public subsidization of education. Hence, public
subsidy to education is shown to be regressive.
Several macro-studies have produced results consistent with the micro
evidence. A one-year increase in the average years of schooling of the labor

force raises output per worker between 5% and 15% (Topel, 1999) and is
associated with a 0.30% per year faster growth rate. Krueger and Lindahl

(2001) found a macro-estimated rate of return to schooling between 18%
and 30%. Sala-I-Martin, Doppelhofer and Ronald Miller (2004) found that a