social security) or because of the resistance of those who would lose out
(democracy, income tax), lacking supporting changes (the tax revenue needed
to pay for professional bureaucracies) or prejudice (female suffrage) (Chang
2003).
Second, in historical terms ‘good institutions’ have tended to follow rather
than being a pre-condition for rapid economic growth, industrialization or
technological upgrading. Khan notes that measures of the quality of bureau-
cracy, rule of law, expropriation risk and contract repudiation by government
were little different between successful East Asian countries and poorly
performing developing countries in the mid-1980s. Fast-growing Indonesia
scored the same as Burma or Ghana, and South Korea, Malaysia, and Thailand
the same as Côte d’Ivoire. The corruption index of 54 countries created by
Transparency International shows that between 1980 and 1990 rapidly grow-
ing East Asian countries had corruption scores little different from other devel-
oping countries (Khan, 2002). There is much broader evidence to show that
now-developed countries also had poor institutions during their initial transi-
tions to rapid growth (Chang, 2002). The UK in 1820 had a broadly similar
per capita income to India today but did not have many institutions and organ-
izations found in contemporary India, including universal suffrage, a central
bank, income tax, generalized limited liability, a modern bankruptcy law, a
professional bureaucracy and securities legislation (Chang, 2003). An interest-
ing comparison which allows for a historical dimension is that from Johnson et
al. (2007) who compare current institutional quality in African countries with
those countries in East Asia in 1960 that subsequently experienced rapid
economic growth. They support the findings of Chang and Khan and find that
the average institutional quality for East Asian countries experiencing
sustained growth after 1960 was about the average for all developing countries.
They find that constraints on the executive, economic risk and trade openness
are today better in Africa, while the control of corruption, primary education,
infrastructure (quality of roads and telephone access) and costs of doing busi-
ness are similar in Africa to those East Asian countries in 1960. Only the extent
of currency overvaluation and health conditions are significantly worse in
Africa. Africa today has less inter-state conflict but more internal conflict;
measures of enthno-linguistic fractionalization (an indicator of the potential
for conflict) are significantly worse in African countries than in East Asia in
- This historical comparison allows us more optimism than is often the
case for Africa as it shows that breaking away from a poor institutional legacy
is possible and has been achieved by many others.
Distinguish policy goals and policy means
We may have decided upon our policy priority, perhaps to improve the supply
of skilled and educated labour, or perhaps to boost the profitability of invest-
ment. Desiring an outcome, however, does not give us any information on how
294 Patterns and Determinants of Economic Growth