One way to model dictatorship

Says the BBC:

Iraqi politics is still a zero-sum game, and one in which the Sunni Arabs feel themselves doomed to be the losers.

Last week I analogized the situation to repeated games of Chicken.

Iraqi consociationalism woes

Iraq’s Sunni vice president vetoed the hard-won electoral law this week. More about that from Barak Hoffman and Matthew Shugart. The current impasse boils down to the apportionment formula. Sunni and Kurdish politicians think the deck is stacked against them.

So is Iraqi consociationalism coming apart, or is this mundane sectarian brinksmanship?

Iraq is our latest experiment in exporting consociationalism. The Iraqi state is built on explicit recognition and institutionalization of combatant ethnic and sectarian groupings. A closed-list PR system funnels these groups into their respective political parties, and, as we saw this week, governing requires the consent of a member of every ethnic group. Now, Iraq’s constitution is not explicit about this. Articles 66-75 set up a semi-presidential system with a unitary executive whose job is to sign legislation. Then there are the so-called “transitional provisions,” which basically divide the presidency among three people elected by 2/3 vote of the legislature. This all but guarantees that the Presidency Council will include one Shiite, one Sunni, and one Kurd, as it does now.

The problem with consociationalism is that, for it to work, elite politicians have to (1) control the combatant groups they represent and (2) desire compromise. The US constitution’s Framers approached institutional design from the safe assumption that politicians are nihilistic power maximizers. This led them to emphasize checks and balances and to riddle the American political system with veto points. Consociationalism, on the other hand, began as a category of observed behavior.¹ It was not the result of a deductive exercise. If game theory is good for anything, it’s good for designing institutions. We begin with an assumption about the preferences of key players (dictatorship by one’s ethnic group > killing each other > dictatorship of the other ethnic group), we choose a desired goal (violence prevention), and we proceed accordingly. Modeling a situation in this way certainly does not lead us to institutions that depend on mutual good will.

The veto of the new electoral law is entirely consistent with the institutional context. That is, we expect outcomes like this one when we run Iraq’s social profile through the consociational machinery of its democracy. I don’t think we are witnessing a constitutional crisis, at least in as much as “crisis” implies an extraordinarily stressful event. From another perspective, Iraq is in a perpetual state of constitutional crisis.

Let me go out on a limb with some predictions.

First, the election will happen, even if a little bit late. Hashemi’s veto is just another round in the ongoing game of chicken that defines constitutional decision-making in Iraq. Brinksmanship and eleventh-hourism have characterized most moments of important political choice since 2003. Why would preparing for the next national election be any different?

Second, there is nothing “transitional” about the “transitional Presidency Council.” What rational group would agree to give up its veto?

Third, Iraqi democracy will not consolidate any time soon. Recall that a car wreck is one solution to a game of chicken. We are more likely to see a dictatorship or civil war in Iraq than we are a stable, electoral democracy.

1. Ian Lustick has a good article about this.

No poutine for the Queen

Canadian Prime Minister Steven Harper will not invite Queen Elisabeth II to Quebec’s quatercentenary birthday party, according to Le Devoir:

The Canadian government refuses to invite Queen Elisabeth II to the 400th anniversary of Quebec’s founding in 2008 for fear of of provoking the anger of certain francophone inhabitants opposed to the royalty, La Presse reported yesterday. (translation mine)

The Conservatives have an interesting relationship with Quebec swing voters. At any given federal election, the Bloc Québecois wins most seats, but a handful are roughly in play between the Liberal Party of Canada (LPC) and Conservative Party of Canada (CPC). Protest votes in pro-Ottawa ridings have put the CPC (and its PC predecessor) over the top twice in recent history: 1984 and 2006.

In 1987, PM Brian Mulroney (PC) negotiated the failed Meech Lake Accord, which would have extended Quebec veto power and constitutional status as a “distinct society.” In the following year, his party won 63 of Quebec’s 75 ridings. (The Bloc was not yet on the scene.)

Last year, the ten CPC pick-ups in Quebec bolstered Harper’s 21-seat margin in the Commons. The challenge now is to consolidate gains. Symbols must matter with this slice of the electorate. After all, the scandal that drove the Liberals from power involved their using tax dollars to hang Canadian flags up around Quebec. Snubbing the Queen is a good way to reinforce the CPC-LPC distinction. Unlike Meech Lake, though, this move probably won’t push confederal glue to its limit.