Uncovering researchers’ implicit bias

I loved this post from the World Bank’s Let’s Talk Development blog on uncovering researchers’ own biases in survey design:

Last year, I was in Nairobi, Kenya … to set up the data collection efforts for a four-country study. One of the goals of this study was to replicate results from lab experiments that suggested poverty is a context that shapes economic decision-making amongst households.

One of our replication questions was a vignette proposed by Sendhil Mullainathan and Eldar Shafir in their book Scarcity. … Shafir and Mullainathan’s findings show that commuters in Princeton, NJ were more likely to say that they would travel to another store for a $50 discount when purchasing a $100 product than when purchasing a $1,000 product. Less affluent individuals at a soup kitchen in Trenton, NJ, however, did not display this kind of inconsistency. For them, the marginal utility of money remained constant, regardless of the cost of the hypothetical product.

With the aim of replicating this question in four developing countries, we took the vignette to … Nairobi. After spending two days field-testing this hypothetical scenario, I was surprised that most of the respondents, particularly those from low-income households, said that they would not travel for the discount. Why was everyone hesitant to travel, regardless of the discount amount? Why are they seemingly exhibiting the preferences of the more affluent in the US?

Somewhat surprised by this discrepancy, I asked multiple respondents’ the reasons behind their choice. One person replied quite plainly, “There is no guarantee that the product will still be there once I go across town. It’s very likely that the product is gone by the time I get there.” Of course! By assuming the availability of the product, we had let our own implicit biases, based on our mental models, influence the design of the question. Since the original question was conducted in the United States, a developed country, implicit in the question was the assumption that availability is generally not a problem. However, for the respondents from less affluent communities, this assumption was not explicit.

Recent articles on conflict

Here’s a handful of interesting articles & books that have passed through my huge pile of unsorted PDFs neatly tagged Evernote notebooks recently.  I’ve included links to ungated versions when available; please let me know if you have access to a free version of any of the gated texts.

  • Chris Blattman’s lecture notes on what American political scientists know about the connection between poverty and violence.  A quick, thought-provoking slide deck.
  • Danielle Beswick on the paradoxes of military capacity building in Rwanda (published version appears to be available for free right now).  Nothing new here if you’ve been watching Rwanda and M23 for a while, but the focus on the risks of a strong military is a useful addition to policy discussions of security sector reform.
  • I haven’t read Severine Autesserre’s Peaceland yet, but it’s high on my list.  Another article covering similar territory to Autesserre’s last book is Jens Stilhoff Sörensen’s piece on the failure of statebuilding.  Key quote: “In its aim to secure, I argue, contemporary state-building and global liberal governance contribute to social and spatial fragmentation in different forms, rather than reconciliation and re-integration.They do so by dismantling previously existing frameworks and introducing market relations where the state has few instruments for attracting cross-sectarian loyalty” (p. 49).
  • Michael Gilligan et al. on how conflict affects social cohesion at the community level in Nepal.  Key point: “We find that violence-affected communities exhibit higher levels of prosocial motivation… We find evidence to support two social transformation mechanisms: (1) a purging mechanism by which less social persons disproportionately flee communities plagued by war and (2) a collective coping mechanism by which individuals who have few options to flee band together to cope with threats” (p. 604)

What explains peace?

In case you missed it, Jon Temin had a great article at Foreign Policy last month asking a critically important question: “Why don’t the policymakers and practitioners looking to end the world’s deadliest wars spend more time studying peace?”  This could mean peace at the country level (he compares Niger’s peaceful relationship with its Tuareg minority to the fraught relationship in neighboring Mali), or within a single country (as shown by the surprising stability of the state of Western Equatoria in South Sudan).  At an even more granular level of analysis, one could look at the case of Butembo – a Congolese city which has remained fairly insulated from conflict despite its location in restive North Kivu province.  But the question in any case is the same: why do some places fall into conflict, while others with similar characteristics manage to avoid it?

There’s a large body of literature in political science looking at cases where civil wars have occurred, but much less looking at war’s absence.  Based on my reading of the conflict literature, here are three factors that the study of peace might start exploring.  (Update, 14 July: read the comments, they’re quite good.  I’ve also added a fourth item here based on feedback from Digitaldjeli.)

  • Regional conflict complexes.  Peter Wallensteen (PDF), Idean Salehyan & Kristian Skrede Gleditsch (PDF), and many others have pointed out that many civil wars are not sui generis, but are linked to conflicts in neighboring countries, often through the mechanisms of refugee movements and state support for armed groups next door.  The canonical example is the way that conflicts in Rwanda have spilled over into and exacerbated conflicts in neighboring DRC.  The obvious question here is why some refugee host countries get drawn into the wars of their neighbors, while others (like Ghana, which hosts a number of refugees from Côte d’Ivoire) manage to avoid this.
  • Ideological collective action.  In The Order of Genocide, Scott Straus finds that Rwandan provinces with administrators who belonged to the ruling MRND party acted quickly to start carrying out genocidal killings after the order came down from Kigali, while administrators who belonged to the opposition were sometimes able to delay the start of violence in their area.  The tactics they employed to do this included organizing self-defense militias for vulnerable communities, threatening to punish people who carried out genocidal attacks, and dispersing groups of men who gathered to start hunting victims.  This, of course, touches on the age-old question: why do some groups of people espouse violent ideologies, while others in the same society do not?  And to what degree are peaceful places peaceful because citizens actively worked for peace, as opposed to simply not having the right preconditions for war?
  • Land tenure policies.  Cathy Boone’s recent book Property and Political Order in Africa argues that places where land tenure rights are assigned by the state are more likely to see both violent and non-violent conflicts scale up to become quarrels with the central government.  By comparison, in places where land tenure is administered by tribal leaders or other local groups, conflicts over land tend to stay “bottled up” at the local level, and are less likely to become national political issues.  Boone stops well short of making the claim that systems of land tenure can explain the prevalence of civil war, but I think there are some ideas here that are worth digging into more deeply.  For example, the highly politicized process by which the state granted land use rights in the Kenyan highlands has created lasting and sometimes violent grievances there, while the politicized process of agricultural collectivization in neighboring Tanzania hasn’t led to large-scale violence (as far as I know).  What mitigated against the violent resolution of land access disputes in Tanzania?  And more generally, are places with tribal or other local systems of land allocation less likely to have civil wars?  This would be an interesting counterpoint to the idea that “tribalism” lies behind many conflicts.
  • Stationary bandits.  Digitaldjeli’s point was that “peace” in Butembo looks more like a protection racket, but the idea that protection rackets can grow into (peaceful, Westphalian) states is actually a classic in the American political science literature.  Mancur Olson (PDF) builds on work by people like Charles Tilly (PDF) to argue that the type of mafioso running the racket matters – “stationary bandits” will protect the people and territory under their control so they can continue to tax them in the long run, while “roving bandits” will steal everything they can from people in the short run, and offer no protection.  Put differently, decisions by political elites can matter a lot for the types of violence that occur within a state.  The million dollar question is why some elites are able to look past the short term gains of roving banditry and decide to make longer term investments in protecting their territory.

At this point I’m actually coming up against the precise problem that Temin highlights: the region I’ve studied most thoroughly, central Africa, is comprised exclusively of countries that have had civil wars, and I’m running out of non-war cases to use for comparison.  What other hypotheses or case studies can you think of that might explain instances of peace in regions seemingly predisposed to war?

Violent struggle and authoritarian durability

I’m only a few months behind the curve on this one – Steven Levitsky and Lucan Way had a very interesting article in December’s issue of Perspectives on Politics called “Beyond Patronage: Violent Struggle, Ruling Party Cohesion, and Authoritarian Durability” (earlier ungated version at SSRN).

This paper argues that institutionalized party patronage — the focus of recent studies by Barbara Geddes, Jason Brownlee, and Beatriz Magaloni — is an ineffective source of elite cohesion. Patronage may preserve elite unity during normal times, but it is often insufficient to ensure elite cooperation during crises. The most durable party-based regimes are those that are organized around non-material sources of cohesion, such as ideology, ethnicity, or bonds of solidarity rooted in a shared experience of violent struggle. In particular, parties whose origins lie in war, violent anti-colonial struggle, revolution, or counter-insurgency are more likely to survive economic crisis, leadership succession, and opposition challenges without suffering debilitating defections. To demonstrate this argument, we compare post Cold War regime trajectories in Kenya, Mozambique, Zambia, and Zimbabwe. Pure patronage parties in Kenya (KANU) and Zambia (UNIP) that were not founded in violent struggle suffered severe defections and fell from power after the Cold War. By contrast, Frelimo in Mozambique and ZANU-PF in Zimbabwe, which were both the outgrowth of long and violent liberation struggles, remained highly cohesive and retained power in the face of powerful opposition challenges and significant economic downturn.

The RPF in Rwanda fits this narrative quite well, and in a 2006 article, Filip Reyntjens noted that the CNDD in Burundi also enjoyed some legitimacy among the Hutu majority because of its role in the civil war. I wonder if this has something to do with Joseph Kabila’s unusual longevity in power, as well.  He doesn’t appear terribly interested in either governing or politicking, but he does seem to lean on his father’s legacy, perhaps getting a boost from any legitimacy he might have earned during the first war.  When I was in Kinshasa in 2009 I remember noting that all of the political posters featured Kabila père rather than the current president.  Would be curious to hear thoughts on this from people who are more familiar with the elder Kabila’s political legacy than I am.

Links of the day

  1. Stability, the International Journal of Security and Development, is a new open-access journal which aims to quickly get relevant academic research to policymakers.
  2. Tom Murphy points out that cancer kills many more people in low income countries each year than does HIV, and Think Africa Press writes that the severe lack of opiates in Africa makes palliative care for cancer victims quite difficult.
  3. Must African presidential aspirants go to prison before they take higher office? (In French)
  4. Videos of cash transfer recipients in Kenya and Mozambique telling their stories in their own words.
  5. A new RCT questions the external validity of RCT-proven results (a political economy story about implementing organizations), and a study from Brazil finds that rainfall fluctuations during pregnancy are associated with changes in infant health outcomes, calling the use of rainfall as an instrument for just about everything into question.