« Like a Poem?!? &!$%$@! | Main | Convexed »

Monday, December 07, 2009

Federalism as Creating a Minor League for Democracy: The Case of South Africa

Roger Myerson, a Nobel Laureate in economics, has proposed an interesting justification for federalism as a device to promote democratic accountability in new democracies. I am wondering if the Democratic Alliance's recent success in the West Cape elections in South Africa might provide a case study of the Myerson theory in action.

Here's a (perhaps inaccurately) over-simplified version of Myerson's theory (which is a rigorously formal theory as befits a Chicago economist and mathematician). Voters in regimes that have become newly democratic face a problem in assessing challengers to incumbent elected officials holding national office. Even if those officials are manifestly crooked, how is the voter supposed to determine whether the challenger is any more honest? After all, talk is cheap. But federalism can help by providing a sort of "minor league" for politicians in subnational government. Subnational politicians can credibly demonstrate their honesty by their performance in subnational office and then use this reputation to compete against national incumbents for national office. The incumbent must respond to the credible signal sent by the subnational official's track record by cleaning up his or her own act. Thus, federalism might improve democratic performance at all levels of government by reducing the barriers to entry for political competition.

Myerson himself applied his theory to Iraq -- in particular, the Provisional Authority's failure to hold local elections in June 2003, which Myerson believes may have deprived voters in Iraq of critical information with which to evaluate candidates for national office. But perhaps South Africa provides another example. The domination of the political system by the ANC in South Africa creates opportunity for corruption. Certainly the current ANC leader, Jacob Zuma, hardly exudes Mandela-like honesty. But how can any rival credibly signal to the voters that they are more honest? Enter federalism: The Democratic Alliance, a centrist party that forms the main opposition to the ANC, has under 20% of the seats in the National Assembly but won control of the Western Cape in the Spring of '09. On Myerson's theory, the DA's provincial triumph and their subsequent use of provincial power might force Zuma and the ANC to improve their performance.

But a lot depends on whether the South African Constitution protects provincial power with sufficient safeguards so that Helen Zille (the DA premier) can actually acquire a track record. In this sense, federalism is not protected by the national political process, as the old Wechsler-Choper theory maintained: Rather, the national political process is protected by federalism. (I should add that one of my students is writing about this issue of federalism and political competition in South Africa, so tips, citations, opinions, etc, from the readership would be gratefully welcomed by both him and me).

Posted by Rick Hills on December 7, 2009 at 10:49 PM in Constitutional thoughts | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Federalism as Creating a Minor League for Democracy: The Case of South Africa:


A "minor league" is important for democracy, but not primarily for the reasons that Myerson identifies. Over and over again in the post-Colonial literature one sees a cycle of nations declaring independence on a Western model, a very thin elite familiar with these kinds of institutions trying to run the modern constitution put in place, and subsequent collapse in favor of military regimes, one party regimes, theocracies or some other manner of dictatorship or monarchy.

For example, Sudan, on paper, had a Western style court system, but had only scores of trained lawyers to run it, and they were spread thin amongst the judicial, legislative and executive branches. Their ability to rule quickly waned. Conversely, those countries that managed to mostly retain their initial constitutional regime in the wake of independence (e.g. India) had large, well developed cadres of educated professionals to operate those institutions.

A "minor league" is important because political leaders need on the job training to learn how to run government entities within a constitutional framework. Incompetence is at least as big of a problem in emerging democracies as corruption, and inability to get things done through normal channels as a result of incomptence is an important motivator for the corruption that does happen. One of the important reasons that political machines (e.g. Boss Tweed) emerged in the wake of mass immigration to the U.S. from autocratic Catholic Europe is that immigrant populations didn't know how to work through the system so they created work arounds instead.

One of the reasons that one party systems and military regimes often emerged in the wake of Western style systems, is that those systems have lots of low level leadership positions to teach future leaders how to be political leaders. "Mr. Smith Goes To Washington" is a beautiful story, but it is a fairy tale, not an evidence based study regarding what produces good politicians.

Ideally, emerging democracies need not only a "minor leagues" but also an amateur league in the form of civil society from the high school and college level and beyond to develop the kind of seasoned leaders needed to run a national government with minimal competence.

There are many ways to establish a record of integrity without serving in junior elected office (think Nelson Mandela and Ghandi), Myerson's hypothesis notwithstanding. But there are very few ways to learn how to be a political leader without actually spending time in leadership positions doing it.

Posted by: ohwilleke | Dec 10, 2009 2:18:39 PM

Maybe you are right, Brian. But the informational demands of Myerson's model are minimal: Myerson argues only that voters needed to be jolted from a state of cynical, jaded ennui by the idea that one of the two candidates (in our example, Helen Zille) is better than horribly corrupt. I doubt that rational ignorance would prevent such signal from trickling down to the marginal voter (assuming that voters otherwise do not show up or distribute themselves randomly). The game-theoretic response is that the corrupt incumbent -- in this example, Zuma -- has to counter with some improvement in his performance.

But we'll see: Zille's aiming to take over the government in the next election cycle.

Posted by: Rick Hills | Dec 8, 2009 10:48:54 AM

I buy that subnational governments (or entities -- see my paper on charity) offer opportunities for political entrepreneurs to build track records for competing with national officials. But I don't think it's likely the mechanism is by "credibly demonstrat[ing] their honesty."

Given rational ignorance among, no claim of honesty or strong performance is likely to be verifiable by the general public -- and therefore not credible. What lower-level officials can do, though, is credibly signal to attentive interest groups their willingness to appeal to those groups -- e.g., by taking positions in favor of those groups it would be costly to walk back. Maybe one such group is a honest-government alliance, but I doubt the rents one would get from such a group would match what others could "pay."

Posted by: BDG | Dec 8, 2009 9:20:44 AM

The comments to this entry are closed.