Globalization and Government Size: Adopting the Mean Group Estimator Model

Globalization and Government Size: Adopting the Mean Group Estimator Model
KATUMBA, Oscar Caster
Issue Date
KDI School of public policy and management
Conventional pooled estimators such as fixed and random effects models posit parameter homogeneity across countries, but this study employed the Mean Group Estimator Model (Pesaran and Smith 1995), which considers parameter heterogeneity to analyze the average effect of globalization, including the impact of its primary constituents such as trade liberalization and financial openness, on the size of government in a sample of 25 countries between the period 1973 and 2005. While overall globalization`s average effect was insignificant, it assumed strong negative and positive significance at country level. Conversely, both trade openness and international capital flows were negatively and significantly related to government size, rendering credence to the efficiency hypothesis (Liberati 2007), albeit related positive and significant country coefficients also existed. By and large, the study shows that different levels of openness, coupled with different political, economic, and social structures across countries, all serve to uniquely determine the size of government. The latter disparity in openness and structures, may also plausibly explain how governments may act to maximize globalization`s benefits, and also ward off its negative effects. Precisely, this study affirms the import of country heterogeneity, in understanding globalization`s effect on government size, akin to an 8-country study done by Islam (2004
Files in This Item:
Appears in Collections:
Master's Thesis (2013)

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.