Summary Glaeser Glaeser et al. - Do institutions cause cause growth? Introduction Survey article revisiting previous attempts to establish causality between institutions and econ growth. Including two approaches in development: 1.
Democracy
and checks on gvt secure PR
2. Human and capital accumulation that start the process. (democracy a consequence of this), example south korea (good institutions here argued to be the outcome of econ growth), Taiwan and china Similarities: emphasize the role of PR, seeing the security of those as a public policy choice. Measurement
of institutions
1. Risk of expropriation by the gvt 2. Gvt effectiveness 3. Constraints on the executive All three, three, by construction, a.
dont describe pol inst they are outcome measures (measure past performance in the first case and quality in the second.
b.
Are
subjective measures reflect what actually happened, rather than permanent rules of
the game. Also since theres been a high reduction of the risk over a very short period of time c.
Rise sharply with the level of development the causality could go the other way around!
d.
Are
extremely volatile (especially in developing countries) cannot measure the durable
characteristics that institutions have. e.
Are
uncorrelated with available constitutional measures of constraints on gvt from electoral
rules or courts. The
first two: a.
confound constraints on gvt with dictatorial choices (since a dictator can choose a good policy), they dont measure institutions, which in their essence are constraints.
The
third: a.
SO:
Also
reflect political outcomes rather than durable constraints
these three are uncorrelated with constitutional constraints on gvt. doubt that these
conventional measures of institutions reflect the deep parameters they want to measure, and that they can be used to establish causality.
Political Discussion
institutions, human capital & economic growth of basic OLS evidence. Confirms: initial level of human capital+ average level of institutions
predicts level of economic growth in the same period of time. Overall,
this evidence is unsupportive of approach 1, and supportive of approach 2 (more basic cause
of growth is human capital). From the OLS regressions, no proof that institutions cause growth as opposed to growth improving institutions. Policies
and growth in poor countries (after 1960)
Looking at empirics. The countries had uneducated populations and were run by dictators in general. BUT:
show dispersion of growth rates inconsistent with approach 1. The evidence is suggestive, but
suggest that focusing on factor accumulation (including growth in human capital) has been more successful than putting constraints on the government.
IV:s Has been used as a way to address the problem that growth might itself lead to better institutions. Discussion
the work of Acemoglu et al with settler mortality and population density. critique: it
doesnt establish a role for institutions. (might have been human capital and not institutions that caused growth.) at an econometric level, predictors of settlement patterns not valid instruments for institutions. Main critiques: a.
Difficult
to assess that the asset being transplanted through settlement was institutions and
not just know-how and human capital. b.
Are
the instruments valid? settler mortality is basically uncorrelated with constitutional
measures of checks and balances, but we would expect the European influence on institutions to be reflected in current rules and procedures. c.
The instruments are correlated with the modern disease environment as well, at least according to Glaeser et al.
d. There is a chance that the IV, settler mortality, influence todays development through human capital, i.e. is correlated with the error term. In that case it isnt a valid instrument. Strong correlation is found btw settler mortality and years of schooling in 2000, which indicates that this might be a problem. Overall: IVs
in literature are even more highly correlated with human capital (today and 1900), and
the specifications predicting econ growth perform better than institutions. The IV approach doesnt tell us what causes growth. Conclusion
Looking at the timing of human capital accumulation and institutional quality. Evidence consistent with example of south korea, which indicates that the causality goes econ growth + human capital accumulation ---- institutional improvement,and not the other way around. The results are instead consistent with a perspective (by D jankov et al) where institutions are an opportunity set based on the human and social capital of the population in a society. Inst are points on this opportunity set, determined by efficiency, history and politics. In this framework, inst only have a second order effect, the first order comes from human and social capital.
Implications Politically constrained gvt not a viable strategy for poor countries to secure PR. Instead they should emphasize econ pol and choices that ensure such security (actual rules), even by dictators. Growth might be feasible even without immediate inst improvement, and in turn the growth is likely to lead to inst improvement. Overall,
current ,measurement strategies of inst has conceptual flaws, but inst may still matter in the
explanation of econ growth. Should focus more on actual laws, rules and compliance procedures. Policy makers should focus more on policies favouring human and physical capital accumulation.