Page 5 - Working Paper (Analysis of Political Budget Cycles in Emerging South East Asian Economies)

DDTC Working Paper 0414 DDTC Working Paper 0414
5

procedures for the management. the data such as tax revenue (%GPD), government
• Non-Tax Subsidy: decreasing oil price subsidy; expenditure, unemployment, industrial share of
decreasing electricity subsidy; increase GDP, household consumption and Real GDP are
national program for poor, basic infrastructure, obtained from International Financial Statistics
health, and education; extending small credit (IFS) and Government Financial Statistics (GFS),
access. Other strategies are enhancement of yearly published by the IMF. To simplify the
infrastructure spending and controlling the analysis, we assume that both executive and
deficit within safe limits. legislative elections are positively correlated and
will bring similar effect to other variables. 10
Despite the needs of well-executed fiscal policy,
there are few others important factors needed We convert some variables into logarithmic
to attract direct investment, for instance, legal function, such as GDP, in order to reduce the
certainty, credible monetary policy, political and skewness that might exist. Logarithmically
economic stability, availability of infrastructure, transforming variables in a regression model is a
wood quality of labor force, etc. very common way to handle situations where non-
linear relationships exist between the independent
Another critical issue that is worth exploring is
and dependent variables (Moon & Benoit, 2011).
7
fiscal policy integration beyond 2015 . In 2007, the
11
The sample ranges from 1985 to 2011. Summary
ASEAN leaders affirmed their strong commitment
statistic of the key variables is represented in the
to accelerate the establishment of an ASEAN
Appendix 1.
Community by 2015. The ASEAN Community is
comprised of three pillars, namely the ASEAN We use Two Stage Least Squares and General
Political-Security Community, ASEAN Economic Method of Moments to find out the causal effect
Community and ASEAN Socio-Cultural Community. between election and fiscal performances. The
Essentially for ASEAN Economic Community (AEC), point of utilizing these two dynamic panel models is
it will considerably transform ASEAN into a region consistent estimation of the parameter determined
with free movement of goods, services, investment, in a model. We believe the endogeneity problems
skilled labor, and freer flow of capital. exist in one of the regressors, some of which are
correlated with the disturbance process. 2SLS
Although each member is still having
is nothing more than the Instrumental Variable
sovereignty in their fiscal policy discretions and
(IV) estimator with a decision rule that reduces
rules, they need to adopt there beyond 2015
the number of instruments to the exact number
policies to support the AEC Blueprint. One example
needed to estimate the equation and fill in the Z
is to gradually eliminate tariffs and non-tariff
matrix (Baum, 2006). As discussed, when we have
barriers on agricultural products, processed food,
multiple endogenous variables, we need at least the
and manufactured products. Further, Custom
same number of instruments as the endogenous
administration in those ASEAN – 6 (i.e. Brunei
variables (exact identification).
Darussalam, Indonesia, Malaysia, Philippines,
Singapore, and Thailand) have to accelerate To fulfill requirement of these models, the
to modernization of customs techniques and instruments should satisfy two conditions. The first
procedures to enhance trade facilitation. is that they should not be correlated with the error
term. Second is that they should be correlated with
4. Data and Methodology the endogenous variables. When we have multiple
endogenous variables, the second condition has a
Similar to previous researches, we consider the more complex expression, called a rank condition.
database on political institutions from the World Moreover, the second methodology that we apply
Bank that provides a wide coverage of countries’ is the Generalize Method of Moments (GMM). GMM
political systems and elections between 1975 estimators use assumptions about the moments
8
and 2013 . We create a binary election indicator, of the random variables to derive an objective
Exelec and Legelec, which take the value 1 in function (Baum, 2006). Based on GMM, we obtain
election years and 0 otherwise. We include the parameters estimates by finding the parameters
two indicators to accommodate different political that make the sample moment conditions as true
9
systems across the five economies. The rest of as possible.

7 Currently, the DDTC team is working on a paper in tax related issues president is elected by the winning party which become majority in the
to AEC. parliament
8 Users of the database cited from Thorsten Beck, George Clarke, Alberto 10 We also conducted a collinearity test to check the relationship
Groff, Phillip Keefer, and Patrick Walsh, 2001.”New Tools in Comparative between the two variables. Results will be provided among request.
Political Economy: The Database of Political Institutions.” 15:1, 165- 11 We have also tested the model using structural break since there
176 (September), World Bank Economic Review. was a crisis between 1998-1999 in this region. However, the results are
9 Before the Asian crisis, Indonesia exercised a political system where pretty similar.