EDUCATIONAL OUTCOMES IN OECD COUNTRIES
product-market regulations have been shown to contribute to differences in short-run
growth among rich countries, we show that neither set of institutions helps us understand
differences in long-run growth rates within the group of OECD countries.
In part, this analysis also enters into the debates about the relative importance of
institutions and human capital as stressed by Glaeser, La Porta, Lopez-de-Silanes, and
Shleifer (2004). Our results strongly favor cognitive skills, our measure of human
capital, over institutions as a fundamental source of OECD-country growth differences.
While cognitive skills can be developed in a variety of ways, we focus on some
fundamental questions about school policy. Educational policy differs considerably by
country, with different countries pursuing very different investment approaches. At one
level, it is seen in very different levels of tertiary education. For example, while
Australia currently has 86 percent of a cohort entering tertiary education, Norway has 66
percent, and Italy has 53 percent (Organisation for Economic Co-operation and
Development (2009a)).
Indeed, Vandenbussche, Aghion, and Meghir (2006) argue that tertiary schooling is
key for developed countries (see also Aghion and Howitt (2006)). They develop a
model where countries close to the world technology frontier should invest in colleges
and universities in order to move the frontier out through innovation. Developing
countries on the other hand should invest in more basic education since they will grow
by imitating the technologies of more developed countries. This conclusion represents
one of the few pieces of policy advice coming out of growth modeling that apply to
developed countries. The evidence, however, comes entirely from consideration of
school attainment and not the more refined measures of human capital used here.
The central issue of Section 4 is whether different levels or types of skills have payoffs
that differ between developing and developed countries. We separately measure basic
and top skills, based on the micro data of the international achievement tests, and then
consider the implications both of high skills and of tertiary schooling on growth. The
results suggest that basic skills have substantial growth payoffs in OECD countries, and
that if anything, the return to top skills is lower, not higher, in the OECD - i.e., opposite
of the innovation/imitation hypothesis. We do not find a specific role of tertiary
attainment for OECD growth once the focus is on long-run growth and direct measures
of skills are taken into consideration.
Results of the regression analyses suggest a strong and robust effect of cognitive skills
on economic growth in OECD countries, but the regression analyses by themselves do
not describe the path of benefits from any program of changing the skills of the
population. In the next part of the paper (Section 5), we therefore perform projections of
the economic value of a set of education reform alternatives for each OECD country
under different parameter assumptions. Our focus is on understanding the dynamics of
economic impacts of educational programs. Three elements of the dynamics are
particularly important for our consideration: programs to improve cognitive skills
through schools take time to implement; the impact of improved skills will not be
realized until the students with greater skills move into the labor force; and the economy
responds over time through the development and implementation of new technologies.