Growth and relative living standards : testing barriers to riches on post-war panel data

The main purpose of this thesis has been to estimate and simulate a general equilibrium model of growing small open economies by taking a new testing approach. The effect of business tax and regulation on growth, together with potential effects of government spending on education and R&D, is emb...

Full description

Bibliographic Details
Main Author: Wang, Jiang
Published: Cardiff University 2007
Subjects:
Online Access:http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.584175
Description
Summary:The main purpose of this thesis has been to estimate and simulate a general equilibrium model of growing small open economies by taking a new testing approach. The effect of business tax and regulation on growth, together with potential effects of government spending on education and R&D, is embodied in this model. We argue that regressions of growth on its supposed causes are not on their own persuasive evidence of these causes. Instead we propose to test theories by a two-stage Popperian procedure in which rejection can occur at each stage. The structural model is estimated on post-war panel data for 76 countries and the bootstrap is used to produce the model's sampling variation for the analysis of panel regressions of growth. In the first stage the model as tightly specified must pass an estimation test in its structural form in the second its bootstrapped implications must be consistent with the growth regressions it implies. We test two main classes of growth theory: one is the Incentivist theory in which growth is caused by incentives for people to engage in entrepreneurial activity, the other is the Activist theory where direct government intervention to stimulate particular activities---specifically education and R&D---causes growth. We are able to reject the latter for education at both the structural and the bootstrap levels and for R&D at the bootstrap level, though not the structural. We accept the Incentivist theory at both levels.