Economic growth and economic development 513

Introduction to Modern Economic Growth
capital-skill complementarities in imperfect labor markets can lead to pecuniary externalities. Nevertheless, existing evidence suggests that the extent of human capital
externalities is rather limited–with the important caveat that there might be global
externalities that remain unmeasured. A particular channel through which global
externalities may arise is R&D and technological progress, which are the topics of
the next part of the book. An alternative source of mismeasurement of the contribution of human capital is differences in human capital quality. There are significant
differences in school and teacher quality even within a narrow geographical area, so
we may expect much larger differences across countries. In addition, most available
empirical approaches measure human capital differences across countries by using
differences in formal schooling. However, the Ben Porath model, analyzed in Section
10.3, suggests that human capital continues to be accumulated even after individuals
complete their formal schooling. When human capital is highly rewarded, we expect
both higher levels of formal schooling and greater levels of on-the-job investments.
Consequently, the Ben Porath model suggests that there might be higher quality
of human capital (or greater amount of unmeasured human capital) in economies
where the levels of formal schooling are higher. If this is the case, the empirical
measurements reported in Chapter 3 may understate the contribution of human
capital to productivity. Whether or not this is so is an interesting area for future
research.
The third set of novel issues raised by the modeling of human capital is the
possibility of an imbalance between physical and human capital. Empirical evidence
suggests that physical and human capital are complementary. This implies that
productivity will be high when the correct balance is achieved between physical
and human capital. Could equilibrium incentives lead to an imbalance, whereby too
much or too little physical capital is accumulated relative to human capital? We saw
that such imbalances are unlikely or rather short lived in models with competitive
labor markets. However, our analysis in Section 10.6 shows that they become a
distinct possibility when factor prices do not necessarily reflect marginal products,
as in labor markets with frictions. The presence of such imbalances might increase
the impact of human capital on aggregate productivity.
499