1148 The Methods of Growth Econometrics
such variables. This is the usual motivation for using fixed effects in the growth
context, as discussed in Islam (1995), Caselli, Esquivel and Lefort (1996) and Tem-
ple (1999). In more recent work, fixed effects estimators have been used in studying
the effects of distinct events or “treatments,” such as democratization or trade
reform. We will discuss this approach in section 24.5.3.
A particular motivation for the use of fixed effects arises from the Mankiwet al.
(1992) implementation of the Solow model. As discussed in section 24.3, their
version of the model implies that one determinant of the level of the steady-state
growth path is the initial level of efficiency (Ai,0)and cross-section heterogeneity
in this variable should usually be regarded as unobservable. Islam (1995) explicitly
develops a specification in which this term is treated as a fixed effect, while world
growth and common shocks are incorporated using time-specific effects.
The use of panel data methods to address unobserved heterogeneity can bring
substantial gains in robustness, but is not without costs. There are times when the
question of interest precludes a fixed-effects approach, and sometimes the limita-
tions of the data will make it uninformative. Some variables of interest are measured
at only one point in time. Others are highly persistent, and this dependence implies
that the amount of useful information in the within-country variation will be
limited. At one extreme, some explanatory variables of interest are essentially
fixed factors, like geographic characteristics. Here the only available variation is
“between-country,” and empirical work will have to be based on cross-sections or
pooled cross-section time series. Alternatively, a two-stage hybrid of these methods
can be used, in which a panel data estimator is used to obtain estimates of the fixed
effects, which are then explicitly modeled in a second stage as in Hoeffler (2002).
A common failing of panel data studies based on within-country variation is
that researchers do not pay enough attention to the dynamics of adjustment, and
the important distinction between short-run and long-run effects. There are many
panel data papers on human capital and growth that test only whether a change
in school enrollment or years of schooling has an immediate effect on aggregate
productivity, which seems an implausible hypothesis. It would be more natural
to consider education as having a lagged effect, especially once various possible
externalities are considered. Another example, given by Pritchett (2000a), is the use
of panels to study inequality and growth. All too often, changes in the distribution
of income are implicitly expected to have an immediate impact on growth. Yet
many of the relevant theoretical papers highlight long-run effects, associated with
the political process for example, and there is a strong presumption that much of
the short-run variation in measures of inequality is due to measurement error. In
these circumstances, it is hard to see how the available within-country variation
can shed much useful light, at least until better data become available.
There is also a more general problem. Since the fixed effects estimator ignores
between-country variation, the reduction in bias typically comes at the expense
of higher standard errors. Another reason for imprecision is that either of the
devices used to eliminate the country-specific intercepts – the within-groups trans-
formation or first-differencing – will tend to exacerbate the effect of measurement
error.^19 As a result, it is common for researchers using panel data models with fixed