capital investment. 2 We found that even when other
potential determinants of educational achievement
are taken into account—such as public spending on
education, fertility, infant mortality rates, and per
capita income—poverty remains an important negative factor in explaining low levels of enrollment
among secondary students. It is thus plausible to
assume that the relatively low levels of investment
in physical and human capital that are characteristic
of Latin American societies, compared for example
to Asia, have something to do with the huge differences in inequality levels among these regions.
the distributional implications of adjustment policies are more complex and contentious. This hypothesis has been recently tested by 2007 research I
conducted with Humberto López. We found that the
volatility of GDP growth is not only higher in countries more exposed to external shocks (e.g., terms
of trade shocks) but that, for a given magnitude of
external shock, volatility is indeed higher in more
unequal countries. This is independent of the effects
of autonomous domestic policy shocks (e.g., fiscal
and monetary policy volatility) on macroeconomic
volatility, as well as other potential mitigators of the
effect of external shocks, such as the degree of openness and depth of domestic financial systems.
Differences in economic performance between
Asia and Latin America may also be partially traced
to this potential effect of high inequality. Indeed,
Latin American countries have often found themselves suffering from protracted recessions (for
example, Argentina between 1999 and 2002). In contrast, Asians recovered faster than Latin Americans
from the economic shocks of the 1980s and proved
especiallyresilientafter
t h ebankandcreditmelt-
d o wnsofthelate1990s.
E c onomiststodaylargely
ag reethatsoundinsti-
t u tions—includingrule
o flaw, law, property rights
p r otection,government
e ffi ciency,andcontrolof
c o rruption—canmake
t h edifferencetoacoun-
t ry”’slong-termgrowthand
per capita income. Many
believe that it should be self-evident that democracy improves the quality of this institutional
environment and hence long-term growth rates.
Yet applied research on the relationship between
a society’s political institutions and growth is
inconclusive, despite the obvious fact that most
Economic research has shown that high inequality is a key factor behind high levels of criminal violence in any society. 3 Not surprisingly, crime rates
(as measured by homicides and theft indices) are
much higher in Latin America than in other developing regions. The effects of high crime and violence on growth have been broadly documented in
several cross-country comparative studies. 4 In particular, more firms in Latin
America identify crime and
violence as a major deterrent to investment than in
other regions. Several specific country studies further suggest that growth
rates are severely affected in
periods of increased crime
and violence (e.g., in El Salvador in the 1980s and in
Colombia in the 1990s), and
bounce back when crime
rates notably improve. 5
Similarly, wide income gaps can leave a country
more vulnerable to economic crises and volatility.
Harvard economist Dani Rodrik observed in 1998
that more equal societies find it easier to respond to
adverse external shocks (e.g., currency devaluations
and fiscal retrenchment) than unequal ones, where
