Research on the economic origins of democracy and dictatorship has shifted away from the impact of growth and turned toward the question of how different patterns of growth - equal or unequal - shape regime change. This book offers a new theory of the historical relationship between economic modernization and the emergence of democracy on a global scale, focusing on the effects of land and income inequality. Contrary to most mainstream arguments, Ben W. Ansell and David J. Samuels suggest that democracy is more likely to emerge when rising, yet politically disenfranchised, groups demand more influence because they have more to lose, rather than when threats of redistribution to elite interests are low.
Dr. Ben W. Ansell is a professor of Comparative Democratic Institutions at Nuffield College. His work focuses on both comparative politics and international relations. He is also co-editor of the academic journal Comparative Political Studies.
Ansell and Samuels (2014) aim at answering the following question: does inequality cause democratization? They challenge the thesis proposed by Acemoglu and Robinson that authoritarian elites grant democracy to satisfy the redistributive demands of the poor and to decrease the likelihood of revolution. Instead of focusing on the redistributive demands by the poor, Ansell and Samuels focus on the demand for democracy by the rising economic elites who are afraid of their property being expropriated by authoritarian elites and seek more political rights to protect it. As part of their empirical strategy, in Chapter 5, Ansell and Samuels instrument inequality level by lagged inequality level and regional lagged level of inequality. However, it is not clear to me how these instruments satisfy exogeneity and/or exclusion restriction assumption. In addition, I am curious about whether both instruments are weak or not. Furthermore, in Table 5.8, the point estimates of BM Gini fluctuate when different instruments are used, which may either be caused by the failure of identification assumption of their IV or the “local” nature of the IV estimates. Many of these empirical problems are not addressed in the book.