The Rwandan empire in Congo

  A “stylized fact” is a fact accepted as true for the sake of argument.  Good stylized facts are true statements that have been simplified in order to allow the creation of tractable arguments.  Bad stylized facts are oversimplified at best, and false at worst. 

One stylized fact is that Rwanda extracted great economic resources from its control of eastern Congo in 1998-2004.  It is accepted.  But is it true?  The answer is not as clear as you might think.

The Rwandan-backed RCD collected fees from local producers.  In Bukavu, merchants who commercialized peasant production were forced to pay a $1,300 fee, in addition to a $3,000 deposit.  In addition, the RCD used military planes to export palm oil to Kisangani.  The rub, however, is twofold.  First, there is not any evidence that the RCD taxes on commerce flowed to Rwanda ... although they did obviously reduce the subsidies that needed to flow the other way.  Second, according to merchants involved in the trade (HBS is very well plugged in to Rwanda) palm oil exports were barely able to cover their costs during the period; Rwanda in fact subsidized the RCD’s profits by providing the planes.

Following the general pattern of the Mexican Revolution, Ugandan and Rwandan troops did not destroy the Sotexki textile plant during their invasion and occupation.  On the other hand, the plant did basically shut down.  The gave an advantage to Rwandan exporters, but it is hard to believe that the additional profits were significant, even in an economy as poor as Rwanda.   

The Sominki gold mines (which also produced coltan and cassiterite) would have been, well, a gold mine, had they not gone bankrupt in the 1980s.  The mine holding company was renamed Sakima, and wound up in the hands of a Canadian corporation, Banro.  In 1998, the DRC expropriated the mine.  (Banro sued Congo at the all-powerful ICSID ... the panel wisely decided to use the fact that Canada had not signed the Washington Convention to punt and let the DRC get away with the expropriation.  Otherwise ... ugh.)  The RCD took over the mines in 1998 and quickly requisitioned 312 tons of coltan and 190 tons of cassiterite.  The DRC gave the concession back to the Canadians in 2003, but the RCD continued to pick over the bones until 2004.  Still, they really were bones, at least in terms of gold; the mines were dead, even by Central African standards. 

So far, so little.  Rwanda “let them [the RCD] exploit the sub-soil for their survival,” and that certainly made it cheaper to maintain control of the area in an international environment that precluded formal annexation.  (There is some evidence that Rwanda wanted to annex the region, which would have completely altered the game, quite possibly for the better.)   That, however, was not a reason for Kigali to continue its occupation; it merely made it cheaper to continue a policy that it wanted to pursue for other reasons.

But that isn’t quite right either.  The U.N. argued that the Congo Desk of the Rwandan Patriotic Army came up with a number of schemes to extract resources from its Congolese territories.  There is even an unconfirmed report of Kagame calling the war “self-financing.”  Thing is, theU.N. report really seems a bit dodgy.  Its estimate of $320 million a year comes from one unidentified source.  It also comes to 20% of Rwanda’s GDP, an four times its official military budget, an enormous sum.  The report gives percentage breakdowns of coltan exports, without mentioning either prices or quantities.  As social science, it is really really bad, inexplicably so.  I could have done a much better job ... but I don’t have reasons to want to find evidence of massive resource extraction. 

I don’t have reasons not to find such evidence either.  In fact, there is solid data regarding diamond profits.  The RPA and RCD apparently set up a powerful monopsony, using force to insure that exports went through their hands.  In 2001, the RCD even studied how their designated tax farmer was performing.  The quota was about $500,000 per month, and the fellow was meeting it ... but the market value of the quantity exported was over $2 million.  The U.N. concluded that the Congo Desk diverted the difference, but in the absence of data on how much was paid to producers and middlemen (and harder price and quantity data) that conclusion is dodgy.  In addition, the maximum estimate of diverted funds comes to a little less than $17 million per year.  That is a lot of money by Central African standards, but it is not $320 million.

In short, it is undeniable that the RCD financed itself from local resources.  It is also undeniable that the RPA got something from the occupied territories.  What we do not know is whether it was anywhere near enough to provide an incentive to prolong the occupation in the absence of security concerns. It really depends on the coltan numbers, which AFAIK we do not have.  Inexplicably.  Here is a pretty bad journalist’s report.  It has numbers, though, and it might be possible to use it to ballpark Rwandan coltan profits, if any. This report seems better, and it might be possible to tease out some real numbers.  I am sure that there is other data available, for anyone with the time and the interest.  (I have only the second at the moment.)

The really interesting counterfactual is whether the Second Congo War and its aftermath would have been as prolonged or as destructive in the absence of an effective taboo against de jure border changes.  And that counterfactual, of course, depends on the amount of resources that could be extracted and the security concerns faced by the relevant parties.  But it is food for thought:  do our post-1945 institutions really contribute to peace and well-being?  I honestly think that they do — but like all stylized facts, it bears the occasional reconsideration.

So ... I ask you all for two things.  (1)  Hard evidence on Rwandan extraction; and (2) thoughts on the impact of the effective end (with a few exceptions, here and there) of de jure annexation?

Comments

[Tzintzuntzan]

Another thing about Rwandan resource extraction is that a lot of the things they took were what, from their point of view, were theirs in the first place. When Kagame's troops took the capital, the fleeing Hutu refugees tried to take everything with them -- such as sawing off telephone cords, ripping the roofs off of houses, etc. In the refugee camps, they'd hock the stuff.

The Rwandans and their proxy allies are probably doing the same (I don't know) -- strip a village down to its literal component parts and treat it like you're a Wall Street liquidator. It's chump change in the grand scheme of things, but a lot of money for a local lieutenant.

[Tzintzuntzan]

About annexation, assuming it was allowed (we know it wouldn't be)...I suspect the immediate result would be another refugee crisis. A lot of Hutus would assume they weren't safe and would flee even further west, all the way to Kinshasa if necessary. That's a problem with this particular crisis, though, not a general principle.

In peacetime...eastern Congo has always served Rwanda and Burundi as the Southwestern US has Mexico: an outlet for surplus population. If you merged the two into one country, I'm not sure which way the migrants would flow. To Kivu, because Rwanda is overpopulated? Or to Rwanda, because that's where the money is?

[Doug M.]

A very slow connection, no academic access, and (most of all) limited time make it difficult to respond in depth. However, here are some first thoughts.

1) The International Court of Justice made a finding of fact that Uganda had indeed looted Congolese minerals and other resources. Uganda is legally obligated to reimburse Congo for this; negotiations have been ongoing for some years now.

2) If Rwanda wasn't occupying those territories to loot the minerals, then why were they? They had no plausible military goal, and were suffering significant damages on the diplomatic front.

3) Circumstantial evidence: let's look at Rwanda's economic indicators.

Their occupation force peaked at around 25-30,000 troops, which was a huge force for a nation of (at that time) around 8 million people -- it's comparable to the US deploying over a million men. On one hand, it was a low-intensity conflict, and next door. On the other, there was constant low-grade fighting, and they were occupying an area roughly the size of California, most of which had to be supplied by air, over distances of up to a thousand miles. Given Rwanda's small, fragile economy, you'd expect to see major signs of stress. Do we?

GDP growth, time series:

1998 8.9
1999 7.6
2000 6.0
2001 6.7
2002 9.4 <- most troops leave Congo
2003 1.0
2004 4.0 <- final peace settlement
2005 6.0

-- So Rwanda grew like gangbusters through this period -- except right after the troops left Congo. Hmm.

Okay, what about foreign debt? Did they borrow to pay for the war?

-- As far as I can tell from brief googling, Rwanda's foreign debt seems to have peaked in 1999, and has been steadily declining since. So, apparently not.

What about using fiscal policy? Did they run a big government deficit to pay for the war?

-- Data insufficient. However, a large chunk of Rwanda's government budget comes from donors. (The figure seems to have peaked at over 60% just after the genocide, and to be currently running around 35% - 40%.) It seems unlikely that donors would have allowed Rwanda to run up a huge fiscal deficit, on their nickel, to fight a war of choice.

Well then, did they print money? How was inflation?

-- Low. I won't do the whole time series, but the average between 1998 and 2005 was about 3.9%, with a single spike of 12% in 2004. If they were printing money, they were being very conservative about it.

So: this was a major expenditure in proportion to that small, poor economy. If the Rwandans were paying out of pocket, the cost should show up somehow. But it doesn't. This strongly suggests that they were not, in fact, paying out of pocket. While circumstantial, this evidence is entirely consistent with the "finance by looting" model.

More in a little while, maybe -- the next few days will be v. busy.

Doug M.

[Noel Maurer]

I dunno, Doug. This kind of circumstantial reasoning leaves causality completely confused. Turn it on its head: Rwanda underwent an economic boom during the period, therefore it could finance the war. It's just as plausible.

Similarly, the ICJ finding by itself doesn't tell a whole lot. The underlying U.N. report would have a lot of legal standing, even though it's incredibly sloppy. There is much evidence that Rwanda (and its proxy) used local resources to finance operations, and that's where the report ends, even if the authors don't seem to realize it.

As for plausible military goals, well, I don't know enough to judge. But I do have enough familiarity with similar situations --- and even a little bit with the specific problem Rwanda faced --- to disagree at the fundamental level of, "Were I a member of Rwanda's executive branch, I would not consider withdrawal in 1998 to be a no-brainer." Perhaps I am just slow.

Having worked in Afghanistan, my expectations of donor-monitoring of government budgets seems to be quite a bit lower than yours. Of course, donors may have been more willing to threaten to pull the trigger against Rwanda ... but that is an empirical question. There is no a priori reason to accept the supposition that donors would be unlikely to provide aid to a country engaged in a war of choice. (And that presupposes that the war was obviously one of choice.)

Finally, as for whether you'd expect signs of stress from military operations in a fragile economy, well, you have made me sad! You know not my academic research? Our empirical findings? The stuff I recently posted about? I get no props from you, my friend. Another haunting, mournful cry goes out across the frozen wastes and icy rivers of Cambridge, Massachusetts.

But believe you me, I understand if you have other things to do over the next few days.

[Doug M.]

Briefly: an economic boom that just happened to end right after they withdrew? Hum.

Obviously a war of choice: Dude. The Rwandans went all the way to the Atlantic Ocean at one point. Later they occupied Kisangani for four years. Kisangani is a thousand miles away from Rwanda, and there were no Hutu rebels within hundreds of miles of it. Militarily, there was zero reason for them to be there.

In fact, in retrospect it looks like the occupation of a vast swathe of Congo was militarily counterproductive. The Rwandans were spread so thin that they weren't quite able to wipe out the last bits of Hutu resistance. That's why some interahamwe escaped into the bush and survived to this day. If crushing the Hutus was their goal, they should have concentrated their forces where the Hutus were, viz., the Great Lakes region.

It's like, your country is being attacked by Yankees fans. So you invade New York... and then conquer every state west to the Mississippi and put a major occupation force into Sioux City, Iowa. It's that level of strange.

Doug M.

[Noel Maurer]

"Hum"? This is where I need to actively repress my Brooklyn origins, because my initial response to that is obscene.

I am going to lay out point-by-point why you should withdraw that dismissive snort. Beside the Brooklyn-related reasons: you know, props, respect.

(1) There is so much variation in the growth rates of poor countries that I have serious doubts a statistical test would show anything. Have there been many one year growth accelerations? Easterly has a very good discussion of the randomness of growth figures in very poor countries; so do (in different contexts) Hausmann and Rodrik.

(2) There is no empirical reason --- and I have done some work on this issue --- to believe that a country (even a poor country) engaged in a foreign war should experience a growth retardation. Causality, therefore, could easily run the other way.

(3) The way to demonstrate the presence of large-scale profiteering (we know that it occurred with diamonds, to a rough maximum of $17m per year) is to check trade flows against domestic production data (which the U.N. claims to have done without providing the data) and to estimate the cost of military operations against the published budget. Of course, that can't demonstrate the absence of large-scale profiteering, but it could conclusively show its presence.

Your circumstantial evidence, however, Doug, is only convincing if you already believe the conclusion.

(4) The Rwandans had security justifications. Your point is that those justifications were not convincing given the extent of their involvement. That makes sense. It also misses the point, which is that Rwandan expansionism was not (for better or for worse) a reason for donors to cut aid ... and the U.S., in particularly, did not cut it off. Figures in millions of current dollars:

1991 40.7
1992 22.5
1993 34.9
1994 54.2
1995 138.2
1996 138.3
1997 90.4
1998 59.0
1999 55.9
2000 38.8
2001 38.5
2002 47.4
2003 47.4
2004 59.1
2005 75.4
2006 103.0
2007 119.1

There might be causality between the aid boom and the end of the war. I do not know. I do know that aid does not (as an empirical matter) cause booms, and I do know that foreign wars do not impede growth. The empirical reason I believe that is my own work. The theoretical reason is the following. Soldiers are paid from tax revenue, and they spend their salaries. It is not clear that they would produce more in the civilian sector, particular in an underdeveloped economy with lots of underemployed resources. Taxing farmers and paying soldiers could, in fact, increase growth in such a situation, depending on how the farmers saved and soldiers spent.

There may have been large-scale profiteering, and that may have been the reason for Rwanda's involvement. That argument, however, is not easy to make, since it requires gathering empirical data that neither you nor I have the time to do. Why is that wrong?

By the way, the internet is a weird animal, so I may be reading more certainty and dismissiveness into your tone than you intend to convoy.

[Tzintzuntzan]

For more on the question of Rwandan motives, Howard French's latest article in the New York Review of Books is worth a read. (I don't agree with everything he says, but it's food for thought.)

[Doug M.]

I can't give this topic focused attention right now, but I did stumble across this recent report:

http://www.congoforum.be/upldocs/Rapport%20%20Exp%20ONU%281%29.pdf

On one hand, it includes both eyewitness and plausible, detailed second-hand reports of the exploitation of gold, cassiterite and other mineral deposits by the FDLR and others. On the other hand, it doesn't seem to cover the macro level; it assumes the minerals are both a money source and a major motivation for the FDLR and other groups (including factions within FARDC, which is supposed to be Congo's own army) without AFAICT running the numbers. Other-other hand, it discusses purchases of military equipment in some detail; obviously money is coming from somewhere.

Anyway, presented FYI.

Doug M.

[Noel Maurer]

Interesting stuff! I have to say, Collier has recently presented some strong statistical evidence that natural resources may contribute to civil war; the general question is still open.

Still, I have a question. Based on a quick skim of the report, it seems to do with the continuing civil war in eastern Congo; not the motivations behind Rwandan intervention. But I'm writing this after a quite-literal three-minute skim. (I know, you have children, but I'm here with my wife's family and can't be too antisocial. I'll be reading it in much detail later today.)

Still, you're more familiar with the DRC context; did I miss something key?

[Doug M.]

No, you didn't. It's not about the Rwandan intervention, but about exploitation of mineral resources today -- by the Congolese Army (which is really a collection of militias) and by other armed groups, including Rwandan "rebel" groups who have since regathered.

But it's the same region, many of the mines are the same, and many of the networks for smuggling were built up during the Rwandan occupation of 1997-2003. If I'm understanding this right, some of the same middlemen who got started under the Rwandans are currently running minerals over the border for the FDLR (among others).

Doug M.

[Doug M.]

There's also this:

http://texasinafrica.blogspot.com/2009/12/show-me-data.html

-- and comments. Interesting discussion, with IMO points for both sides.

(I may keep coming back to this thread periodically as I stumble across stuff.)

Doug M.