Menu Adam R Brown

Notes navigation: Browse by titleBrowse by authorSubject index

Geddes: Politician's Dilemma

Disclaimer. Don't rely on these old notes in lieu of reading the literature, but they can jog your memory. As a grad student long ago, my peers and I collaborated to write and exchange summaries of political science research. I posted them to a wiki-style website. "Wikisum" is now dead but archived here. I cannot vouch for these notes' accuracy, nor can I say who wrote them.

Geddes. 1994. Politician's Dilemma. Berkeley: University of California Press.

In Brief

Leaders, like everybody else, are self-interested. They want to stay in office. Whether they seek policy or office is irrelevant, because they need office to implement policy (more often than not). And if you want to advance, you need to take into account the electoral rules, party rules, and other rules that structure access to government offices (11). So, the size of the selectorate ("franchise") matters greatly. And if you are trying to extend your power over entrenched leaders, you use what power you already have to enlarge the franchise (12). Thus, when it is politically useful to do so, politicians can "be expected to behave autonomously from societal interests" (11).

The politician's dilemma is the conflict between a politician's need for immediate political survival and longer-run collective interests in economic performance and regime stability. In particular, even though a strong, professional civil service would help a country in the long term, an unprofessional, crony-filled civil service will help political leaders survive in the short term.

Research Question

Why would self-interested politicians replace an unprofessional, crony-filled civil service (good for reelection and short-term benefits) with a professional civil service (good for long-term benefits)?

Hypotheses and Research Design


Hypotheses: X1 causes Y1; if X2 happens after Y1, then X2 causes Y2.

In those countries where X1 and Y1 happened, Geddes further looks at the next part of the theory:


Geddes employs a "most similar design" (Mill's "Method of difference"). In this method, cases are picked that are as similar as possible in relevant ways, except that some of the countries get the "treatment" (X1 and, in some cases, X2); these countries are then compared to the control cases.

Cases: Five Latin american countries (Chile, Uruguay, Colombia, Venezula, Brazil) that are similar on key variables that might affect reform: culture, colonial institutional structure, and level of economic development. These cases vary on X1 and, where Y1 happened, on X2.

Chapter-by-Chapter Notes

Chapter 1

By looking at the institutions that structure access to power, we can understand why (self-interested) politicians might occasionally behave autonomously from societal interests (i.e. they will if it is politically useful to do so) (p 11). But, even if politicians have autonomous policy preferences, that doesn't matter much if they can't implement the policy they want. So, you would think all politicians would want to develop state capacity. Whether they in fact do so, however, "depends on whether it serves the immediate career interests of the politicians who initiate the reforms and choose appointment strategies" (14). Thus, the politician's dilemma: Whether to seek short-term political gains/survival or long-term economic gains. Political leaders may genuinely embrace a set of shared national goals, but "the exigencies of political survival" may lead them to "behave in ways that undermine these goals" (19). See example of Allende and the coal mines (17-18).

Chapter 2

Politicians don't provide administrative reform for two reasons: (1) for most ordinary people in developing countries, achieving increased administrative competence and honesty is a collective action problem. Consequently, their interest in reform . . . does not . . . develop into politically compelling demands" (35). (2) The obstacles to collective action are stronger than commonly thought. Cooperative solutions must be better not only than what "individuals can achieve through their own unaided efforts, but [also than] what they achieve by cooperating in small, informal [patron-client] clusters." POLITICAL ENTREPRENEURS might try to gain power by promising public goods. But principal-agent problems prevent them from following through. First, moral hazard: constituents in developing countries often lack enough information to know whether their politicians are following through on this kind of promise. Second, competing incentives: politicians can better ensure their reelection by using the corrupt bureaucracy to reward key supporters. Thus, THE POLITICIAN'S DILEMMA: "Politicians who might otherwise consider offering reforms as a strategy for attracting support will not be able to afford the cost in lost political resources as long as they compete with others able to use such resources in the struggle for votes. . . . A politician might in some circumstances, however, be willing to give up this resource if everyone else were also willing." (42)

Chapter 4

Politicians will only support administrative reform if it helps them electorally. Even political entrepreneurs cannot be counted on to provide reform: they may promise reform, but they will have trouble getting seats in the legislature if they don't use cronyism to give supporters their "due." Geddes gives a series of games to show a few things. First, since larger parties control more patronage, it is likely that larger parties prefer patronage (defection) over merit (cooperation) (Figure 3, pg 92). Thus, minority/challenger parties may favor mutual cooperation, but incumbents actually prefer mutual defection. Second, larger parties will only vote for reform if the electoral gains of supporting reform exceed the electoral gains of using patronage (Figure 4, pg 94). Thus, U.S. civil service reform occurred only when (1) the public got real riled up about it after a disgruntled patron shot Garfield in 1882 and (2) the Republicans got voted out of the House for failing to pass reform; in December, just before the Democrats took their places in the House, they passed reform. (95-6)

Incentives to change from patronage to merit-based system:

  1. Equal distribution of patronage among the two largest parties (so each party has equal incentives to vote for reform, or at least not to vote against it--otherwise, only smaller parties (which have less access to patronage) will have an incentive to support reform);
  2. Legislators have "some additional incentive to eschew patronage, such as widespread pressure for reform from constituents" (99).

Chapter 5

A test of the game theoretic model, partly summarized in the "Variables" and "Design" sections above.

Research by the same authors

Research on similar subjects


Geddes, Barbara (author)Comparative PoliticsState CapacitySelectorate TheoryPublic GoodsPresidentialismClientelismCollective ActionPrincipal-AgentPrisoner's DilemmaLegislatures

Wikisum home: Index of all summaries by title, by author, or by subject.