Thursday, February 27, 2014

What Political Science Has to Say about Turkey

The latest twist in Turkey's ongoing political corruption crisis was the release of a tape purportedly showing Prime Minister Erdogan and his son Bilal in a series of phone conversations discussing how to hide ill-gotten money from the police (transcript here). One could talk about this ongoing scandal, but I'd rather focus on some bigger questions.

The question is this: why does a Prime Minister, and party, which has never won an electoral majority (look it up), have such a stranglehold on power within Turkey right now?

The answer is Turkey's political system. A poorly-designed electoral system feigns proportionality but has given Erdogan's AKP three straight disproportionally large parliamentary majorities. What institutional checks that ever existed in the 1982 constitution have largely been eroded: the independence of the judiciary was undermined (it was far from perfect pre-2007, but it's even worse now), the president is now directly elected, and the military has been defanged (not a bad thing, of course).

But no new checks on the power of the parliamentary majority were introduced. And, as this was happening, most observers in Turkey and in the West were cheering along--so taken in with Turkey's apparent 'democratization' that they failed to see the evident danger of 'elective dictatorship.' And, while the evidence has been building for years (see the 'Sledgehammer' or 'Ergenekon' trials, the high number of imprisoned journalists, the slow erosion of media independence, etc), too many policy elites in Washington, Brussels, and elsewhere weren't interested. Now, suddenly, those Beltway insiders have noticed. Better late than never, I suppose, but advice to close the barn door is far more useful before the horse has run away!

We Can Communicate Our Findings Better...But You Have to Listen!

We could have told policy elites and DC/Brussels insiders that this erosion of rule of law in Turkey was going to end badly.

Political scientists have two primary advantages and contributions to offer to policymakers.

First, we think deeply and carefully about politics. While policymakers and Turkey experts were cheerleading Erdogan's 'democratization', we have been thinking and writing about democracy for years. And what we have consistently argued, in accessible venues, is that democracy is far more than just elections (see Dahl or Schmitter & Karl as just two examples). We have also spent a lot of time thinking and writing about the appropriate balance of majority rule, institutional checks on power, civil liberties, and so on that go into producing the best democracy. See for example the lifetime works of Arend Lijphart (he's got a Wikipedia page, so his work is accessible and a big deal).

Second, we have the broader vision that comes with thinking comparatively about the big questions. Events in Turkey seem unique to people who study only Turkey. But, to people who study comparative politics, these events look awfully familiar. Because we see them in Latin America, South Asia, the Balkans, and so on. Scholars writing about competitive authoritarianism and democratic careening had insight to offer into Turkey's political trajectory under Erdogan.

Unfortunately, the painful truth is that many policymakers and country experts didn't want to hear it. (This is a point that Nick Kristof completely misses in his critique of academia). For the first ten years or so of Erdogan's rule in Turkey, DC/Brussels insiders were keen to highlight Turkey's democratization. There was, inevitably, a lot of 'politics' behind this: the desire to promote a Muslim democratic "model" for Iraq and Egypt, etc. A lot of this was the product of shallow thinking that assumes Muslim countries need some sort of specific "model" of democracy; why the German "model" wouldn't work in a Muslim country is never explained but often assumed (as I pointed out here).

So, up until very recently, the emphasis was on touting the positives of Turkey's 'democratization' while downplaying or ignoring the very real concerns that political scientists would highlight. This report by the Council on Foreign Relations, released in 2012, is a pretty good example. It gets around to mentioning the "worrying developments" of the erosion of democracy, but only after writing that "Turkey is more democratic, prosperous, and politically influential than ever before." Which claim do you suppose got more attention?

One could, with a little effort, find dozens of similar articles in influential magazines and think-tank publications from the 2002-12 time period.

There's a lesson in all of this. We (academic political scientists) can and should do a better job of disseminating our insights, engaging with real-time events, and so on. I was doing none of this during that 2007-12 period, though I grumbled privately about these shoddy, cheerleading articles I kept seeing about Turkey (I kept joking that their official English name was "the mildly Islamist AK Party," because it was always described thusly). Mea culpa.

But journalists, policy experts, and policy makers share a lot of blame for disregarding or cherry-picking our insights as well. And that will mean reaching out, in some cases. As Brent Sasley points out, being a professor is actually our "day job", and one that keeps us pretty busy.

Monday, February 10, 2014

Switzerland's Immigration Referendum

On Sunday, Swiss voters narrowly voted to approve an initiative to reintroduce immigration quotas, including upon citizens from EU member states. Some interesting points about the referendum result:

  • Turnout was high at 56%. This result cannot easily be dismissed as an aberration due to low and uneven turnout.
  • The result broke down along linguistic lines. French-speaking cantons voted to reject the measure, while all but one German-speaking canton voted to approve.
Naturally, this result has led to a lot of questions. One puzzling result that some have noted is that the negative relationship between the number of immigrants in a given canton and support for the initiative. In other words, voters in cantons with lowest percentage of immigrants were most likely to support immigration restrictions! Courtesy of Alberto Nardelli:

The figure is in French, but it is straightforward. The vertical axis shows immigrant population as a percentage, and the horizontal axis shows the "yes" vote for the initiative (the size of each bubble reflects the canton's population). One can see that the pattern moving from top-left to lower-right. The initiative got less than 40% support in each of the three cantons with more than 30% immigrant populations, while it received over 50% support in the vast majority of cantons with less than 23% (the national average) immigrant population.


One small reason is that those immigrants who were eligible to vote probably voted not to restrict immigration. But that can't explain much of the difference.

Some, such as Tyler Cowen, suspect that this reflects a natural limit to how much immigration voters will tolerate. Economist Bryan Caplan thinks the result is evidence that status quo bias prevents acceptance of immigration. Pointing to similar data from the US, he argues that people simply become more accepting of immigrants when they see them (and thus get over their fears/hostility, etc). There is probably some truth to this.

But there is also a deeper, social psychological element in all of this. There is a large and growing body of evidence showing that individuals vary in their predisposition toward xenophilia or xenophobia in ways that vary predictable with other traits (such as openness to experience, respect for tradition and authority, etc). Excellent (and readable!) examples of recent work in this tradition include Predisposed and Our Political Nature.

How does this map out? People who are more likely to support immigration are also more likely to be open to new experiences, to be more accepting of non-traditional authority or lifestyles, and to value individual autonomy. People who are less likely to support immigration are also less likely to be open to new experiences, to reject non-traditional lifestyles, and to prefer social cohesion over individual autonomy. (Note that I'm implying no causal direction between support for immigration and other predispositional traits here). It's not much of a stretch to suggest that the former group of people are more likely to move to major cities (where they will end up being around more immigrants), while the latter are more likely to remain in or move to rural or suburban areas. This is not to say that other arguments, such as Caplan's arguments about exposure, or Cowen's point about backlashes, don't have merit. They do. But I would argue they may well be secondary to more fundamental patterns observed in social psychology.

This is heavy on speculation without individual-level data, of course. Nonetheless, it provides a prediction that helps to make sense of the observed results: support for immigration has less to do with economic interest than it does with a basic orientation towards difference, social cohesion, and members of "out-groups".