impact on growth in regions other than Sub-Saharan Africa. The broader
implication of this discussion is that growth is not a universal process, cross-
country growth regressions are an intrinsically poor mechanism to analyze
growth and each growth experience should be treated as potentially unique,
that is, as a case study.
Case studies of boom and bust
The use of case-study research rather than statistical methods in economics has
declined, as the former were increasingly regarded as subjective, non-rigorous
and non-replicable (Gerring, 2007:5–6). Several scholars have more recently
attempted to provide rigour for the case-study approach (Sambanis, 2003; Yin,
2003; George and Bennett, 2005; Gerring, 2007).
Barro-style cross-country growth regressions seek to measure the impact of
a change in one variable (perhaps literacy rates) on the dependent variable
(GDP growth). The statistical testing procedure then throws up various criteria
by which a researcher can judge how reliable that estimate is, but the method
does not examine the reasons for the relation. Does literacy have a positive
impact on economic growth because it makes workers more productive or
because it allows firms to use more complex technology? In the latter case
education would not have a positive impact on economic growth unless it was
also possible for firms to acquire that new technology. Case studies can exam-
ine any form of causal process in more detail. Econometrics typically focuses
on a linear (straight-line) relation between a policy variable and economic
growth: for example, an increase in investment will have the same impact on
growth, whether investment is increasing from 2 to 3 per cent of GDP or 32 to
33 per cent of GDP. The ‘big push’ theory, for example, argues that self-
sustaining growth will only occur once investment has reached a critical
threshold and until that point investment is likely to have little impact on
growth.
The use of case studies to examine the causal mechanism is called ‘process
tracing’ (George and Bennett, 2005:205). This method involves the researcher
putting together a detailed narrative inspired by economic theory and
supported by rigorous empirical evidence to analyze a whole sequence of
possible causal/historical processes. All intervening steps between cause and
effect can be taken into account and each step be preceded by a consideration
of relevant economic theory and evidence (Gerring, 2007:181). Comparative
case-study analysis has been useful, for example, in distinguishing the paradox
between deep reform and stagnation in post-Soviet Russia and
moderate/heterodox/gradual reform and rapid growth in 1980s China (Nolan,
1995).
A common criticism of the case-study methodology is that there is no
formal method of case-study selection, leading to a frequent suspicion that
the analytical results will be pre-determined by the choice of the case study:
for example, the case studies of growth among ‘now-developed’ countries
Growth in the Modern World Economy since 1950 53