Journal of Economic Growth, 2: 155–168 (June 1997) c  1997 Kluwer Academic Publishers, Boston. Power, Distributive Conflicts, and Multiple Growth Paths GILLES SAINT PAUL CERAS, DELTA, and CEPR THIERRY VERDIER CERAS, DELTA, and CEPR This article shows that multiple growth paths may occur in a politico-economic model of endogenous growth. This multiplicity is characterized by the coexistence of the low-tax, low-capital-flight equilibrium and a high-tax, high-capital-flight equilibrium. The likelihood of multiplicity is crucially related to the structure of power in society—namely, it is necessary that the politically decisive agents have a greater access to international capital markets than the average in the economy. Keywords: growth, taxation, capital flight, multiple equilibria, redistribution JEL classification: E22, E44, E62, F21, H2, O16, O4 1. Introduction Political instability and distributional conflicts are in general harmful for growth. 1 They are more intense in some countries than others: for example, they are more intense in Latin American countries than in European countries. The former countries have repeatedly adopted populist programs that have created disincentives for growth through capital flight and reduced domestic investment. Yet at some times these countries have been able to adopt sounder policies (and consequently better economic performances) without necessarily affecting their political institutions. One striking example of such policy reversals is Mexico. Up to the late 1980s, Mexico had traditionally pursued high-tax policies. Thus, the average personal income tax rate was 50 percent and the corporate income tax rate 42 percent. In less than two years, these tax rates were both reduced to 35 percent (see Aspe, 1993, p. 108, fig. 2.7). These reforms have coincided with a sharp rise in capital flows. Thus, net foreign reserves rose from −4 percent of GDP in 1988 to +3 percent in 1991 (Aspe, 1993, p. 231, fig. 5.4); similarly, direct foreign investment (in U.S. dollars) almost quadrupled from 1988 to 1991 (Aspe, 1993, p. 165, table 3.11). Last, the economy’s growth performance was much better in the first half of the 1990s than in the last half of the 1980s (4 percent versus 1 percent). 2 This drastic change in policy and performance may be perhaps explained by some exoge- nous shift—let’s say of preferences. Economists are however generally uneasy with such