Stop highlighting our differences? #moreincommon

Last night at my local primary school Governor meeting one of the other governors objected to a table showing a disaggregation of recent pupil discipline issues categorised by ethnic grouping. “Should we really be calling children ‘White Other’ or ‘Black Other’?” Turns out these are the standard official government categories offered to students/parents to self-identify with. As a researcher I'm naturally interested in as many descriptive categories as possible to help understand the factors that drive differences in outcomes between individuals, but every time we ask the question we also ask people to think in ethnic or racial or national terms, highlighting our differences not the more in common. 

 

As Chris Dillow wrote recently in an excellent take-down of David Goodheart:

"The thing is, we all have multiple identities: I’m tall, white, Oxford-educated, bald, heterosexual, male, bourgeois with a working class background, an economist, an atheist with a Methodist upbringing. And so on and on. The question is not: what are my identities? But rather: which of these identities matter?

 

… 

 

Even if you accept biological essentialism, the question of which of our multiple identities becomes salient is surely in large part a social construct.

 

...

 

No good can come from raising the salience of racial or ethnic identities."

The issue comes up often in national censuses. For the last South Sudan census in 2008 it was decided it was too politically charged to ask people their ethnic group. Lebanon hasn’t had a census to count the number of Christians and Muslims since 1932. In the UK, the government recently started asking schools for pupil’s nationalities, with a stated aim of allowing for better targeting of support, but leading to widespread suspicion and calls for a boycott. 

 

The first step to thinking about when we should and should not ask for ethnic identities might be assigning some plausible values to the likely costs and benefits of doing so. A new paper by Evan S. Lieberman and Prerna Singh has taken a systematic approach and coded over 1000 census questionnaires for 150 countries over 200 years and whether they ask for ethnic identities. 

"Through a series of cross-national statistical analyses, the authors find a robust association between enumeration of ethnic cleavages on the census and various forms of competition and conflict, including violent ethnic civil war”.

That seems like a pretty high price to pay.

The Political Economy of Fertility

I argue that fertility may be a strategic choice for ethnic groups engaged in redistributive conflict. I first present a simple conflict model where high fertility is optimal for each ethnic group if and only if the economy’s ethnic diversity is high, institutions are weak, or both. I then test the model in a cross-national dataset. Consistent with the theory, I find that economies where the product of ethnic diversity and a measure of institutional weakness is high have increased fertility rates. I conclude that fertility may depend on political factors.

I'm skeptical of the cross-national empirics without having looked at them, because ... they are cross-national empirics (see here for further explanation). But the theory is interesting and sounds plausible.

(And here's the link for the paper by Thorsten Janus, forthcoming in Public Choice)

Maybe grievances do cause conflict after all

Another counter-intuitive freakonomics-finding-fail?

Everyone knows that civil wars are driven by grievances right? Israel-Palestine and all that. Until along came the quants who found no relationship between measures of ethnic diversity or grievances with civil war. The claim was that diversity and grievances are too ubiquitous to be able to predict isolated incidences of conflict.

Until now. A new, more detailed GIS dataset from researchers at Zurich is finding some effects.

Our most recent dataset GeoEPR, which was published earlier this spring, offers a “bird’s eye view of ethnic settlements” around the world … the new dataset offers geo-coded information about ethnic groups’ settlement areas around the world from 1945 through 2005.

Based on this information, we show that excluded and downgraded groups are much more likely to experience civil-war violence. Furthermore, our most recent research, conducted by Lars-Erik Cederman, Kristian Gleditsch at University of Essex, and Nils Weidmann at Princeton University, uses GeoEPR data suggesting that ethnic groups that are either considerably less or more wealthy than the country average are also exposed to higher conflict risks. Such “horizontal inequality” has previously been associated with conflict, but our research is the first to do so globally with the help of geographic information systems (GIS).

From Lars-Erik Cederman at the University of Zurich Political Science blog.

You can download the data here.

On the importance of motivation for results

So the other day I was idly browsing the tinterwebs for some info on the different tribes of Southern Sudan (yes I am that kind of geek). Tribe, ethnicity, language, and religion were all explicitly excluded from the 2006 census and the 2009 household poverty survey because they are too contentious, so I was assuming that hard data is difficult to come across. Not so.

Joshua project is a research initiative seeking to highlight the ethnic people groups of the world with the least followers of Jesus Christ.

And oh the data is copious. And detailed. I am slightly creeped out and utterly amazed all at the same time.