Transcript About Authors

Do Institutions matter? Estimating
the effect of institutions on
economic performance in China
About Authors
• Ying Fang: Associate professor of
economics at the Wang Yanan
Institute for studies in economics.
• Yang Zhao: Xiamen University
Introduction
China
• Landmass is about the size of Europe
• Population is estimated at one-fifth of the world
• Over 300 cities
• 80% of productivity gap among cities, explained by
the investment climate (World Bank, 2006).
Introduction
• How can we understand the large spatial
differences in economic performance?
schools of thought on this question:
1. Institutions
2. Geography
3. Human capital and government policy
• China provides a easy way to compare these
three.
Previous studies on this topic
1. Paulo Mauro:
Ethno linguistic
Fractionalization
Institutions
(Corruptions)
Economic
growth
Previous studies on this topic
2. Hall and Jones
Distance
From Equator
Social
Infrastructure
Growth
Previous studies on this topic
Acemoglu and Johnson
European settler
mortality rate
Institution
Economic
Growth
Reason for Use of instrumental variable:
problem of endogeneity makes it impossible
to establish a causal relationship
1. Reverse causality
Institution
Economic
performance
Reason for Use of instrumental variable:
problem of endogeneity
2. Mutual causation
Economic
performance
Institution
U
• Author’s IV: Enrollment in protestant
missionary primary schools in the early
20th century.
Reasons for choice of IV:
1. History
2. Geography
3. Protestant education: enrollment in protestant
education in 1919
Enrollment in
protestant
missionary
primary
schools in the
early 20th
century.
Institution
(Index of
property
rights
protection)
Economic
performance
(log of GDP
per capita)
Definitions & Historical background of China’s
institutional changes
• Trace of the historical influence of the West on
China’s institutions
• What China’s economic transition means? A
process whereby property rights are given
increasing protection and respect
• What is Institution? Sets of social rules that
protects property rights. Authors focus on
implementation rather than paper policy
• Economic performance: measure of GDP per
capita.
Data collection process
Index is calculated from survey of cities
using questionnaires answered by scholars,
entrepreneurs and randomly sampled
citizens, the index for property right
protection is calculated from this survey.
Empirical Results
• 2SLS estimate of economic performance is
4.23., meaning that a city GDP per capita
would increase by 4.23% if index of
property right increases by 1%
• 2SLS estimates are larger than OLS
estimates which means that OLS estimates
are down-ward biased
• Result is robust when tested for violation of
exclusion
Institution versus geography
• First stage regression of institution on IV
and geographic factors found IV to be
more significant. Second stage regression
of Log GDP on institution and geography
also found institution as more relevant.
Institution Versus Policies
In the first stage regression IV estimate is
more significant and also in the second
stage institution is more significant than
policy variables (dummies for coastal open
cities, government consumption as
percentage of GDP, investment rates in
2004)
Conclusion
1. Institutions are significant in explaining
the variations among Chinese cities.
2. Institution trumps other explanations for
variations in economic performance, that
is, geography and policy.
Critique
• Method of sample selection