Political Economy

April 12, 2013
Alberto Alesina of Harvard University, Organizer

Patrick Francois, University of British Columbia; Ilia Rainer, George Mason University; and Francesco Trebbi, University of British Columbia and NBER

How Is Power Shared in Africa? (NBER Working Paper No. 18425)

Francois, Rainer, and Trebbi present new evidence on the power sharing layout of national political elites in a panel of African countries, most of them autocracies. They present a model of coalition formation across ethnic groups and structurally estimate it with data on the ethnicity of cabinet ministers since independence. As opposed to the view of a single ethnic elite monolithically controlling power, this analysis shows that African ruling coalitions are large and that political power is allocated proportionally to population shares across ethnic groups. This holds true even after restricting the analysis to the subsample of the most powerful ministerial posts. The authors argue that the likelihood of revolutions from outsiders and the threat of coups from insiders are major forces explaining such allocations.

Francesco Passarelli, Università Bocconi, and Guido Tabellini, IGIER

Emotions and Political Unrest

Passarelli and Tabellini formulate a general theory of how political unrest influences public policy. Political unrest is motivated by emotions. Individuals engage in protests if they are aggrieved and feel that they have been treated unfairly. This reaction is predictable because individuals have a consistent view of what is fair. This framework yields novel insights about the sources of political influence of different groups in society. Even if the government is benevolent and all groups have access to the same technology for political participation, equilibrium policy can be distorted. Individuals form their view of what is fair taking into account the current state of the world. If fewer aggregate resources are available, individuals accept a lower level of welfare. This resignation effect in turn induces a benevolent government to procrastinate unpleasant policy choices.

Christian Dippel, University of California, Los Angeles

The Transmission of Colonial Institutions: Evidence from the 19th Century Caribbean

Despite wide agreement on the colonial origins of comparative development, there is little empirical evidence on the actual transmission of colonial institutions over time. Dippel provides evidence for this transmission, using variation in a concrete set of colonial institutions provided by a series of constitutional changes in 17 post-Abolition British Caribbean colonies, where local elites voluntarily curtailed or abolished their elected parliaments. These cessions of local elites' de jure powers are explained by increasing voter turnout of freed slaves and by increases in electoral turnover. Consistent with models of de jure and de facto power, the elites' cession of de jure powers tilted public expenditures and taxation in their favor, permanently putting the Caribbean on a path of non-democracy, low public good provision, and low state capacity

Eliana La Ferrara, Bocconi University

Does Interaction Affect Racial Prejudice and Cooperation? Evidence from Randomly Assigned Peers in South Africa

La Ferrara exploits the random assignment of roommates in double rooms at the University of Cape Town to investigate whether having a roommate of a different race affects inter-ethnic attitudes and cooperative behavior. She finds that living with a roommate of a different race has no significant effect on prejudice on average. She does observe increases in the frequency with which respondents talk about race and report to have experienced discrimination. In terms of social interactions, exposure to a roommate of a different race increases the number of study mates of a different race, but not the number of "best friends" of another race. In a prisoners' dilemma and trust-games analysis, she exploits variation in the race of the roommate and in the race of the game player with whom that individual is (randomly) matched. She finds that black students who live in mixed rooms are more likely to cooperate in the prisoner dilemma game, while white students make higher transfers, both as senders and as receivers, in the trust game when they are paired with someone of a different race. Overall, the random allocation policy seems to have affected social and cooperative behavior to a larger extent than it affected prejudice and implicit attitudes.

Raymond Fisman, Columbia University and NBER, and Yongxiang Wang, University of Southern California

The Mortality Cost of Political Connections

Fisman and Wang study the relationship between the political connections of Chinese firms and workplace fatalities. The worker death rate for connected companies is five times that of unconnected firms; this result also holds when they exploit executive turnover to generate within-firm estimates. The connections-mortality relationship is attenuated in provinces where officials' promotion is contingent on meeting safety quotas. Fatal accidents produce negative returns at connected companies and are associated with the subsequent departure of well-connected executives. These findings emphasize the social costs, as well as the firm-level benefits, of political connections.

Gergely Ujhelyi, University of Houston

Civil Service Rules and Policy Choices: Evidence from US State Governments

Ujhelyi studies the policy impact of civil service regulations, exploiting reforms undertaken by U.S. state governments throughout the twentieth century. These reforms replaced political patronage with a civil service recruited based on merit and protected from politics. He finds that state politicians respond to these changes by spending relatively less through the reformed state-level bureaucracies. Instead, they allocate more funds to lower level governments. The reallocation of expenditures leads to reduced long-term investment by state governments.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us