Saturday, December 28, 2013

Should the EU End Accession Talks with Turkey?

As a renewed round of anti-government protests pick up in Turkey in the wake of the corruption scandal, this tweet came across my feed:

This raises a serious question: is it time for the European Union to end accession talks with Turkey? Though I've long supported the hope of Turkey joining the EU, I'm now convinced that the correct answer is "yes".

Before explaining why, perhaps a little review is in order.

The case for Turkish membership in the EU was always something of a long-shot. Geographically, Turkey is only tenuously a part of Europe. Culturally, the hang-ups about Turkey being a predominantly Muslim society that is as much (or more?) "Middle Eastern" than "European" couldn't be ignored. And even when things looked better around 2004-5, Turkey had (and continues to have) serious problems with its standard of democracy and rule of law. So the case for Turkish membership always seemed to depend on improving political conditions; the economic benefits of a young, large, and growing economy; and the symbolic benefits of a largely Muslim country joining.

What Europeans (Don't) Want
In spite of the arguments made in favor of Turkey joining, Europeans were never convinced. The following chart shows public opinion in the EU towards Turkish membership in 2008 (source: Eurobarometer). (Note: "Net support" shows the percentage in support of Turkish membership minus those opposed. A number below 0 means that more respondents in that country oppose than support Turkish membership).

The results speak for themselves. Europeans are against Turkey becoming an EU member on the whole (net opposition of over 20 percentage points), and overwhelmingly so in many countries (including in the two biggest, France and Germany). Support, where it exists, is tepid, except in Romania. (Note that the results omit responses of "not sure" and "don't know", so even where net support is positive, support is still usually below 50%).

This gets to a pretty important reason. Many in the EU complain of a "democratic deficit" in which EU leaders don't listen to what citizens actually want. Here is a good test of that: Europeans don't want Turkey in the EU, so what will EU leaders do?

Engaging Erdogan?
The events of 2013, however, make the case for continuing Turkey's accession talks even more problematic. However one evaluates the whole of the Erdogan government's record on democratization, it is clear that 2013 has seen very visible demonstrations of brutal uses of power against protestors, evidence of high-level and massive corruption, and a willingness to undermine the rule of law in order to maintain power. How can the EU continue accession talks with a regime that is very clearly walking away from the Copenhagen Criteria?

One answer is that the EU needs to continue to "engage" the Erdogan regime to push it to make democratic reforms. This claim makes sense enough if the problem were simply that reforms had stalled, or perhaps if Erdogan were newly in power and needed a push to get started. But, instead, he has been in power for over a decade, and he has actively sought to weaken Turkish democracy in recent years.

The "engagement" argument also threatens a sort of circular logic. When Turkey makes positive changes, it should be rewarded with more engagement. When it fails to, this also shows a need for further engagement. So, is the answer always more engagement? Rather than inspiring Erdogan to push on with democratic reforms, I think it could easily be argued that "engagement" after the events of the past 6 months only serves to undermine the EU's credibility.

What to Do?

It seems difficult to make any sort of case for continuing accession talks with Turkey at this point. It's deeply unpopular with Europeans, it splits European elites, major sticking points (such as the status of Cyprus) remain, and Turkey is backsliding away from the Copenhagen Criteria.

The best option, then, might be to suspend accession talks (rather than pushing forward, as the EU has recently done), in the hopes that a post-Erdogan Turkey might get serious about taking the necessary steps. In the long run, though, it's hard to see a case for continuing the process.

And it's important to remember that ending accession talks does not preclude continuing or even expanding the Customs Union, or pursuing other forms of collaboration.

Monday, December 23, 2013

Is the Muslim World Doomed to Bad Governance?

Over the past two days, Ed Husain, who is a senior fellow at the Council on Foreign Relations, has been involved in something of a twitter debate with several others. It all started with this tweet (find his tweets here):

This comment is surprising for a few reasons. For starters, the "attacks" against the government of Turkey have mostly come from Turks themselves. And why? In the summer, it was mostly about overly authoritarian rule. A newer wave of protests now, in December, center of allegations of widespread and expensive corruption on the part of the Erdogan government. At no point have those "attacking" the government of Turkey claimed that Erdogan didn't win elections fairly, or that the Turkish economy had not grown during Erdogan's 11-year tenure. So it's all a non sequitor, really.

But, about those GDP numbers. Husain's claims about Turkey's GDP, while true, are a bit misleading. I'll outsource the work of explaining why to Dani Rodrik and Emre Deliveli. Short story: Turkey's economy has grown at a healthy clip since 2002, but so has every country that is similar to Turkey economically. In other words, Erdogan's record may be less to do with Erdogan's policies and more to do with broader global market conditions.

But that's just where the fun starts. When others pointed out that the protests are about democracy and rule of law--not economic performance, Mr. Husain responds in a series of tweets:
This is a common argument one hears from people making "political culture" arguments. And, yet, it's a flawed one. First of all, it's quite unclear what "democratic governance of the N. European variety" means. But let's assume it means free and fair elections, rule of law, and protection of individual liberties.

Well, we already know that Muslims can handle the first of those (free and fair elections). Mr. Husain himself uses the fact to argue against those engaging in the latest fad of attacking the Erdogan government. And if we include some aspect of popular participation: well, there have been mass protests for and against various regimes in Turkey, Egypt, and elsewhere. So, Muslims seem to have no problem mobilizing politically.

So does this mean that Islam is contradictory to the rule of law and or civil liberties? I find it hard to see how. Islam imposes strict commandments on its followers. And, among others, members of Erdogan's government have used the language of democracy and civil liberties to justify ending old Kemalist restrictions on religious practice. So what's the problem?

Rather, I suspect that, as with most political culture arguments, this claim about the incompatibility of Islam and democracy is based more on intellectual hand-waving than a serious argument.

Ah! Another one that we commonly hear. "It's been this way for centuries, so it can never change!"

Funny how selective we can be when making arguments about cultural conservatism. Muslim societies have had no difficulty adapting to change in many other respects. The automobile? Television? Blue jeans? Coca-Cola? No problem! In all sorts of economic, social, and technological ways, those "conservative" Muslim populations have problem evolving. Nor, for that matter, do millions of Muslims living in Western societies!

But, somehow, Mr. Husain would have us believe that good governance is a bridge too far for the Muslim world.

On a side note, the "eight centuries" argument is common and appealing, but what sense does it really make? Humans have been organizing themselves into societies and governing those societies for tens of thousands of years. Eight centuries is just a blink of the eye in human history--hardly some insurmountable obstacle. But, at the same time, humans live for only 80 years or so. Why does what happened in the 15th century predestine people living today to some specific outcome that they are powerless to change? It doesn't, of course, and societies all across the world (including Europe) are quite good evidence of that.

Finally: "Turks and Arabs have been at it since the 1200s." Nonstop, inevitable warfare, the entire time? Just like the French and Germans, I guess.
"Can't implant West on East" is the ultimate conversation-ender, isn't it? After all, saying this accuses the other person being (gasp!) a Western imperialist, insensitive to the peculiar needs of of "Eastern" societies. In truth, it's a straw man argument.

But, again, what does any of this actually mean? The blind spots that arguments like Mr. Husain's contain are bewildering. All of the things that Mr. Husain suggests that Muslims must simply learn to accept as a cultural inevitability--brutality, corruption, ethnic conflict, etc--exist and have existed in plentiful supply in the West. It's funny: when Kurdish nationalism threatens to tear apart Iraq or Turkey, one never hears the likes of Mr. Husain arguing that you can't implant the Western concept of nationalism on Eastern societies. Or, for that matter, when touting the economic record of the Erdogan government, one never hears about the incompatibility of Islam and global market capitalism that has facilitated the "Erdogan boom" (despite the fascination with "Islamic finance"). One could give many other examples.

Let's Be Serious

If we can get rid of the appealing but useless political culture arguments, then we can get down to the serious business of diagnosing the problems of countries like Turkey and recommending useful solutions. Thankfully, we have an entire field of study--comparative politics--that addresses such questions. And there are far better explanations for Turkey's ills. One might start by reading HuntingtonMigdalGeddes, Inglehart, or North, or many others.

Turkey's problems are rooted in a lack of sufficient institutionalization and a lack of economic development. For that, we don't need to blame Islam. These are, after all, precisely the same problems that plague Christian neighbors Greece, Bulgaria, and Serbia. And, for that matter, Argentina, Brazil, Indonesia, China, and India. It's nothing that is uniquely rooted in Muslim "culture".

These are serious problems, and difficult to overcome. It's far easier to point out that Turkey's legal system is corrupt and inefficient than it is to remedy the problem. But making facile arguments about the inevitability of the problem, and the failure of any potential solution to that problem, does nothing to help and a great deal to muddy the waters. So, no, Turkey's current political climate is not as good as it gets. But it certainly will be if we accept Mr. Husain's arguments.

Friday, December 13, 2013

European Muslims, Fundamentalism, and Intolerance: A word of caution about some worrying findings

There is a new study out today, which is rightfully receiving a lot of attention. Based on what looks to be an impressive new data collection effort of immigrants from Turkey and Morocco living in six European countries, a report of some early data analysis has some worrying findings.

Specifically, levels of religious fundamentalism and outgroup hostility are shockingly high--and much higher than among native European populations. Over half of European Muslims surveyed appear to be religious fundamentalists, and they tend to be hostile towards homosexuals, Jews, and the West in general. This is troubling, as it suggests that a large portion of Europe's Muslim immigrant community is not at all well-integrated into European society, and it also suggests a large population that may be reception to extremist and violent messages from religious leaders. (See the technical report here).

Here is the scariest figure from the report, which compares levels of outgroup hostility among Muslim immigrants and native European Christians:

Three thoughts:

First, this is not terribly surprising. In June, I posted an blog entry reporting evidence from the European Values Survey of respondents from within Turkey. Though my purpose was different, the findings are similar to those here. When asked about the sorts of people that respondents would prefer not to have as neighbors, substantial majorities of Turks said they wouldn't want to have Jews, Gypsies, or Christians as neighbors. It's not a great leap to imagine the same would be true of homosexuals. So, there is a lot of outgroup hostility in Turkey, and it shouldn't be too surprising that a lot of that apparently transfers to Turkish emigrants living in Europe.

Second, buried in the report is a massive difference among Sunni and Alevi Muslim Turks. While there were high levels of fundamentalism and outgroup hostility among Sunni Turks (45% and 31%), the numbers were much lower among Alevis (15% and 13%). Alevis were in fact hardly different from native European populations. The conclusion here, I think, is that elite messages and family socialization matter. I won't dare try to explain religion here, but suffice it to say that the Alevi faith is more pluralistic and liberal than Sunni Islam as typically practiced in Turkey.

This point is hugely important, because it shows that there is nothing inherently fundamentalist or intolerant about Muslim "culture" (or, god forbid anyone even suggest it, genetics). Rather, the problem is that too many religious leaders within the Muslim world choose to preach fundamentalist and intolerant interpretations of Islam, which in turn influence followers' beliefs. So, if your imam spends every Friday sermon telling you that the Jews and the West are out to get you, there is a decent chance you'll believe it. If, on the other hand, your imam tells you that all people deserve respect, there's a half-decent chance you'll believe that instead.

Finally, though, I have a methodological concern about the survey. Scrolling through the questionnaire, it struck me that all of the questions about religious fundamentalism and outgroup intolerance were structured as statements to which the respondent could either agree or disagree (e.g., "Jews cannot be trusted"). That in itself is not a problem, except that all of the questions appear to have been structured so that agreeing with the prompt indicates fundamentalism or intolerance, while the respondent must disagree in order to indicate pluralism or tolerance.

This suggests to me a well-known problem in survey research called acquiescence bias. This problem emerges from the fact that people generally don't want to appear ignorant and/or disagreeable to the person interviewing them. So, all else being equal, they will tend to agree with statements posed to them as a default position. Thus, if all of the statements "lean" in one direction (such as intolerance), then you will observe higher levels of intolerance among your survey sample just because respondents are more likely to agree rather than disagree with the statements you pose to them. One way to deal with this is to alternate the "direction" of the questions so that, in this case, some are worded to indicate a non-fundamentalist or tolerant position (e.g., "Jews can be trusted").

Having only read the technical report, I can't be certain that I am right about all of this. But, I suspect that the observed level of fundamentalism and intolerance among Muslim respondents would fall by a decent proportion (maybe 10-15 percentage points?0 if proper measures had been taken to guard against acquiescence bias (if they were, then I am mistaken and I apologize, but I can find no evidence of that here). And, while I don't know the literature on this topic well enough to say so with any certainty, I would imagine that acquiescence bias would be greater among less-educated respondents (who will be less likely to have a true "opinion" on many subjects). European Muslim populations definitely have lower levels of educational attainment, on average.

But let's not let our Muslim respondents off too easily. After all, if you asked me a question about whether Jews cannot be trusted, or whether Islam is seeking to destroy the West, nothing short of a deceptive question could prompt me to agree with those sentiments. In other words, my views are firm on the subject, and I'm not going to pretend otherwise for the sake of agreement. Neither, I'd like to believe, would any of my friends (note to my friends: if this isn't true, please let me know so we can stop being friends).

The point being: in order for acquiescence bias to occur, there must be some possibility that the respondent could agree with the statement and that it could be socially acceptable among his/her community to do so. So, what's troubling is that large numbers of European Muslims are in a position to acquiesce to statements that they wouldn't want gay friends or that Jews cannot be trusted. Probably, the number who really believe these things is lower than this survey finds, but it is higher than one would hope to see.

In short, it's pretty bad. But I suspect it's not quite as bad as the report indicates, and it doesn't have to be this way (as the Alevi respondents show).

Thursday, December 12, 2013

Majoritarian Democracy & Political Psychology

In an earlier post, I noted how events in the Czech Republic, Hungary, and Turkey were illustrations of the perils of equating majoritarianism (or, perhaps, electoral democracy) with liberal democracy. In each case, the problem is of an elected leader--presuming that being elected entitles to him to claim knowledge of the "national will"--pushing through policies that are supported by the majority of the population but which are harmful to various political minorities.

Why the problem? On an institutional level, the emphasis on electoral democracy ignores several essential (and related) elements of liberal democracy: the necessity of conflict and compromise, respect for individual rights, and the crucial role of the rule of law.

On a more basic human level, we share certain psychological traits that should warn us away from majoritarian democracy. Namely, many people tend to overestimate the degree that others agree with them, particularly when in the majority (meaning that people also tend to underestimate the degree to which others legitimately differ from them). And many people will quite happily repress others' rights when those other people's beliefs or behaviors are viewed as inappropriate or threatening. What's more, a person's willingness to coerce or repress others increases (paywalled) in stressful situations--precisely the time when cool heads are needed most!

Part of the solution to this problem must be institutional: laws and practices that emphasize majority rule are likely to exacerbate the problems described above.

So, look what happened in Croatia. Last weekend, Croatians voted to implement a constitutional ban on same-sex marriage by a nearly two-to-one margin. As a result of a simple majority vote, the rights of a group of people (who are a distinct minority, and who are certainly viewed as deviants or threats to society by many) have been taken away in a semi-permanent fashion (constitutional bans are not easily overturned).

This highlights a serious problem with the use of direct democracy. What happened in Croatia is reminiscent of what happened in the United States in 2004,  when voters in a number of states similarly acted to ban same-sex marriage. The point should be clear:  

People will happily vote to take away the rights of other people, particularly when they consider those other people to be a deviant or threatening minority!

So, a liberal political system ought not to allow for such a possibility. Direct democracy has its place. But, all too often, the problem with the use of referendums is that they trigger the worst instincts and biases of voters, with the end result of denying others rights or making terribly ill-informed policy judgments.