Causes of Tax Reform in Latin America, 1977-95

Abstract

Among the policy changes associated with neoliberalism in Latin America, has played a leading role as it has been crucial not only to price stabili- zation but also to managing economic liberalization. But it also has a larger sig- nificance, since it involved a reconstitution of core state powers, and these could prove useful to any future government that seeks to expand the state's economic role. This paper seeks to determine its causes more precisely by analyzing data from fifteen Latin American countries from 1977 to 1995. Findings show that the definition of tax reform has been remarkably similar across the region with less progressivity, fewer exemptions, a new leading role for the value-added (VAT), and the strengthening of administration. The data analysis then finds is predicted by (in roughly descending importance) past inflation, explicit IMF performance conditions, new administrations, more authoritarian-elected govern- ments, the dominance of the president's party in the legislature, established elec- toral systems, closed-list proportional representation, less polarized party systems, and more numerous parties. Little or no support exists for the causal importance of past changes in gross domestic product (GDP), the constitutional powers of the president, party institutionalization, or partisan balance. The analysis concludes by placing these results in historical context, referring to theories of state forma- tion and the building of institutions in exchange for resources. In the drama of Latin American policy changes during the late twen- tieth century, played a pivotal role. It was crucial to price stabilization, which depended on finding a substitute for the inflation tax; it supported trade liberalization, in that the most common re- form, the introduction or expansion of the VAT, made up for cuts in 1. The first stages of work on this project were funded by the National Science Foun- dation under SPS-9320878. I would like to thank the staff of the Archives of the Interna- tional Monetary Fund for their helpfulness, and Leonora Dodge and Martha Rodriguez for dedicated research assistance. Thanks to Ken Shadlen, Carlos Elizondo, Jonathan Conning, Kurt Weyland, Marcelo Bergman, and three anonymous reviewers for com- ments on previous versions or fragments of this study. Thanks also to Samuel Morley for sharing his index database with me. Details of the data analysis and spread- sheets of key variables are available at http://www.williams.edu/PoliSci/mahon.htm. My entire database is available to any interested researchers upon e-mail request to me at jmahon@williams.edu.