On external interventions to improve village-level governance and development: The DRC Edition

This is from an excellent JDE paper by Humphreys, de la Sierra and der Windt:

We study a randomized Community Driven Reconstruction (CDR) intervention that provided two years ofexposure to democratic practices in 1250 villages in eastern Congo. To assess its impact, we examine behavior in a village-level unconditional cash transfer project that distributed $1000 to 457 treatment and control villages. The unconditonal cash transfer provides opportunities to assess whether public funds get captured, what governance practices are employed by villagers and village elites and whether prior exposure to the CDR intervention alters these behaviors. We find no evidence for such effects. The results cast doubt on current attempts to export democratic practices to local communities.

Here’s a description of the program:

Our study takes advantage of a large UK funded CDR program, called “Tuungane,” implemented by the International Rescue Committee andCARE International in 1250 villages throughout eastern Congo. The program had as a central goal to “improve the understanding and practice of democratic governance ….”

… Over a four year period, the program spent $46 million of development aid, reaching approximately 1250 villages and a beneficiary population of approximately 1,780,000 people. A large share of this funding was used for facilitation and indirect costs, with only $16m, 35% of the total program costs, going directly towards infrastructure. These shares reflect the fact that the main focus of the intervention was institutional change, not the use of existing institutions to deploy development funds.

This very cool paper raises important questions about the role of elites in African development (read it to get a better understanding of the futility of these kinds of “democracy promotion”, too).

It might seem logical to assume that short-circuiting elite power, whether at the local or national level, may lead to accelerated development. However, because a lot of “development” is often elite-driven, an explicit agenda of effective elite disempowerment might actually yield suboptimal outcomes. All else equal, elites are often better organized, better-placed to take risks (on account of having more economic slack), better able to protect their property rights, and routinely deploy the state to further advance their economic interests. $46m in the hands of a powerful and secure elite class might yield jobs in firms that provide economic stability for whole districts. It is also true that less powerful or stable elites are likely to squander it on consumption, quick profit schemes, or stash it abroad.

These observations are not unique to African states.

Overall, when I look at most African states, what I see are a lot of very weak elites lacking social power, constantly unable to bend their societies to their will, and resigned to low-equilibrium forms of  political and economic organization (for example, by being mere middlemen in lucrative global commodity markets). In the case of the DRC, this is true whether one looks at Kabila/Tshisekedi or the leaders of armed groups in the east of the country. The same goes for so-called “traditional” leaders. Throughout the country and in the wider region, such elites lack infrastructural power in profound ways. Importantly for economic development, many often lack the ability to protect their own property rights. Our stylized idea of the nature of societal power relations on the Continent needs some updating. Consider this paragraph:

Eastern Congo is a well-suited environment to examine the adoption of democratic practice in local governance. The state has largely with-drawn from the rural areas of the east and enjoys low legitimacy. Local governance is often described as “captured” by traditional chiefs and vulnerable to corrupt practices by state officials. These features are not unique to the Congo. Multiple accounts suggest that in many Sub-Saharan states, colonial rule used pre-colonial institutions to create “decentralized despots” in ways that are detrimental to development.

topographies.jpgAre local elites in the modal African country this powerful? Is this the sense one gets traveling in rural Ghana or Zambia? Do these (mostly) guys look like they are in charge? As the paragraph notes, “traditional leaders” often lack the means to coerce their constituents (the state is largely absent). Despite Mamdani’s persuasive (Rwanda) story, these are not powerful and unchecked “despots” in the standard sense.

At times Africanist scholarship on state/elite society relations can seem schizophrenic: Africa is the land of “imperial” big men elites who can scarcely project their power on account of state weakness (see here, here, and here). Since the early 1990s, a lot of effort has been put into taming the allegedly imperial political elites in the region. Missing in our analyses and in donor programs have been attempts to understand the structural weakness of these same elites and the attendant consequences. The presence of an erratic and parasitic elite class might be the proximate cause of underdevelopment in the region. However, I would argue that a deeper cause is persistent elite weakness in the region. Catherine Boone’s book (see image) is the best I’ve ever read on African elites’ strategies of power projection in a context of state weakness (Boone is easily the most underrated Comparativist of her generation).

The tenures of Africa’s Amins, Mobutus, and Bongos took the form they did in no small part because these were structurally weak leaders (long leadership tenure is not synonymous with state capacity). Throughout their times in office they did all they could to destroy any and all alternative centers of power (including institutions such as legislatures). Their failures reinforced their respective counties’ two publics problems whose legacy is chronic elite weakness that is obvious for all to see. To this day, very few African countries have stable economic elite classes with easily identifiable immovable assets in-country. Most operate like little more than Olsonian roving bandits.

I am yet to see a clear theory that links greater vertical accountability to state/elite capacity. The historical record suggests that democracy works best in contexts with pre-existing state/elite capacity. In my own work, I’ve shown how strong autocratic legislatures beget strong democratic legislatures.

This is not a defense of autocracy. It is a reminder that the processes of state and political development, while related, often run on separate tracks and should therefore be decoupled in programs such as the one above and in our studies.

Mamdani on the African University

Ugandan scholar Mahmood Mamdani has an excellent piece in the LRB on the African University. Here are excerpts:

On the state of affairs:

The African university today is still very much what it was from the start: a colonial project with a monolingual medium of instruction, framed in terms of a European ‘universalism’ from which a large majority of the colonised were excluded.

On language (channeling Ngugi wa Thiong’o — an idea that keeps growing on me with age…):


Makerere University

Is there an intellectual mode of reasoning we can describe as African, in the way Mazrui spoke of a ‘French’ or a ‘Western’ mode of reasoning? Not an ancestral or genetic mode, obviously, but one which weaves together a set of discourses communicated in a common language that presupposes – or suggests – an intellectual community with a long historical formation. Language is our first obstacle here. Most of those of us who have come out of colonialism speak more than one. The languages of colonialism are inevitably languages of science, scholarship and global affairs. Then there are the languages of colonised peoples – languages whose growth was truncated by colonialism. Our home languages remain folkloric, shut out of the world of science and learning, high culture, law and government. There are exceptions. In East Africa, Kiswahili is the language of popular interaction, culture, and official discourse, also the medium of primary and secondary schooling, but not of university education. At East African universities, it has the status of a foreign language, with departments of Kiswahili studies. It is not the bearer of a scientific or a universal philological tradition.

Finally, on decolonizing the university:

What would it mean to decolonise a university in Africa? The East African experience suggests that one answer would be the opposite of what is happening in American and British universities: reducing the cost of a university education, by state grants and subsidies, to make it more inclusive. In the first place, therefore, fees would have to fall. I was at the University of Cape Town from 1996 to 1999; in the years that followed – the heyday of South Africa’s independence – fees began rising. In the second place, there would have to be multilingual projects designed to provide Westernised education in several languages and to nurture non-Western intellectual traditions as living vehicles of public and scholarly discourse in those languages. This is not a demand for a revivalist project, but a call to include the languages of popular discourse, which in South Africa would mean centres for the study of the Nguni and Sotho languages and traditions (the opposite of area studies), and translation units, carrying the best academic literature – global, regional and South African – back and forth between the new linguistic centres and the older faculties. Broadening the referential world of African universities means competence in the languages which embody non-Western traditions.

Read the whole thing here. Mamdani serves a great narrative on the intellectual history of East Africa.

In reaction to Mamdani Chris Blattman makes important points in this thread.

Also, let’s bring Transition back to life.

Rwanda, 20 Years On

Caution: This is not an apology for President Kagame and his autocratic tendencies that have resulted in carnage and death in the DRC, Rwanda and elsewhere.

At a conference last year a US State Department official told a group of us that Rwanda was so polarizing that even at the Consulate in Nairobi the DRC crowd did not get along well with the Rwanda crowd.

It is not surprising why that might have been the case, or why the present analysis on the commemoration of the 20th anniversary of the 1994 genocide remains polarized.


If one just looks at the improvements made in advancing human welfare since President Paul Kagame and the RPF took power (see graph, data from the World Bank) it is hard not to arrive at the conclusion that ordinary Rwandese are unambiguously better off. The country is the least corrupt in the region and has also been consistently ranked top in the ease of doing business. But there is also the side of the Kigali government that most reasonable people love to hate: the murderous meddling in the DRC and the oppression and occasional murder of dissidents at home and abroad. Those who admire what President Kagame has done tend to emphasize the former, while his critics tend to emphasize his autocratic tendencies which have made Rwanda the least democratic country in East Africa (see below, data from Polity). Many wonder if the post-1994 achievements are sustainable enough to outlast President Kagame’s rule.

So is Mr. Kagame a state-builder or your run of the mill autocrat whose achievements will vanish as soon as he relinquishes power?

ImageIn my view, I think that Rwanda is the best success story of state-building in Africa in the last 20 years. I also think that this (state-building) should be the paramount consideration for those who care about the Rwandese people and want to help them achieve greater freedoms. The fundamental problem in states like CAR, Sierra Leone or Liberia has never been the insufficiency of democracy. Rather, it has been the problem of statelessness. The contrast between Rwanda and Burundi is instructive (see both graphs, the two are neighbors with similar ethno-political histories. Rwanda has historically had a stronger state, though. See here and here). Despite the latter being the second most democratic state in the region, it has consistently performed the worst on nearly all human development indicators. Part of the reason for this is that Burundi remains a classic papier mache state confined to Bujumbura and its environs.

May be I am too risk averse. But I am scared stiff of anything that could lead to a recurrence of the horrors of the early 1990s stretching from the Mano River region to the Horn. As a result I am always skeptical of activism that takes state capacity (including coercive capacity) for granted.

With this in mind, the fight against autocratic rule in Rwanda should not come at the expense of the state-building achievements of the last 20 years. The international community and those who genuinely care about Rwandese people should be careful not to turn Rwanda into “democratic” Burundi in the name of democracy promotion. Interventions will have to be smart enough to push President Kagame and the ruling elite in the right direction, but without gutting the foundations of political order in Rwanda.

Absent a strong state (even after Kagame), the security dilemmas that occasioned the 1994 “problem from hell” would ineluctably resurface.

Lastly, I think the level of discourse in the “Rwanda Debate” could be enhanced by the extension of the privilege of nuance to the case. For example, if all we focused on were drones killing entire families at weddings in Yemen or the horror that is the South Side of Chicago we would probably get mad enough to ask for regime change in Washington. But we don’t. Because people tolerate the “complications and nuance of American politics.” The same applies to less developed countries. Politics is complicated, everywhere. And those who approach it with priors of good-or-bad dichotomies are bound to arrive at the wrong conclusions. One need not be a Kagame apologist to realize the need for a delicate balance in attempts to effect political change in Kigali.

Before you hit the comment button, notice that this is neither an apology nor an endorsement of autocracy in Rwanda. It is a word of caution regarding the choices outsiders make to accelerate political change in Rwanda.

Tyranny is not the panacea to underdevelopment. But neither is stateless democracy.

For background reading on the 1994 genocide in Rwanda see Samantha Power’s Problems From Hell; Mahmood Mamdani’s When Victims Become Killers; and Philip Gourevitch’s We Wish to Inform You That Tomorrow We Will be Killed With Our Families.