Skip to content

An oil purse is a curse, of course?

This post is by Adam Martin, a post-doctoral fellow at DRI.

In development economics everyone knows that natural resources are a curse. A well-known study by Sachs and Warner found a negative correlation between resource abundance and growth rates, while subsequent studies have shown a negative relationship with democracy.

The Curse enjoys wide appeal. Aid skeptics like that it implicates oppressive domestic government and nationalized industries. Aid supporters are drawn to its emphasis on geography (destiny!) and the indictment of global markets. And on the popular level, no one makes a better villain than oil companies. But popularity doesn’t stop the story from being hot, flat, and wrong.

New research argues that empirical work on the Curse suffers from two interrelated problems. First, it uses dependence (the share of GDP from that resource) and calls it abundance (the stock of a resource in the ground). But dependence in turn depends on institutional quality—if you have sound institutions, natural resources take their place along other industries. If not, natural resources will by default constitute a large share of GDP because poor institutions stifle an advanced division of labor. When you look at cross-sectional data using dependence as a proxy for abundance, it will look like natural resources compromise institutional quality.

That reliance on cross-sectional data is the second major problem. The Curse story does not claim that Nigeria is Britain plus oil, but rather that Nigeria is less democratic than Nigeria would be in the absence of oil. One way to get around this problem is to test whether oil makes country X less democratic using panel data with fixed country effects. That’s fancy econometric speak for taking into account other factors that might make country X more or less democratic—its history, institutions, culture, etc. Fixed effects also allow testing a corollary of the Curse known as the “First Law of Petropolitics”: as oil prices go up, oil-rich autocrats crack down on democracy even more.

Digging into the recent research:

  • Christa Brunnschweiler and Erwin Bulte tackle the first problem. They find a positive correlation between resource abundance and both growth and institutional quality, and argue that it is conflict and poor institutional quality that lead to dependence.
  • Stephen Haber and Victor Menaldo offer a great review of the second problem. They present evidence that even natural resource dependence does not undermine democratization.
  • Romain Wacziarg corrects for both problems, testing for the effects of high oil prices on democracy using panel data. Again, there is no evidence for the Curse.

These studies argue that, while the Curse is plausible, domestic institutions are simply too persistent for it to matter much. Will belief in the Curse likewise prove too persistent in the face of new and better evidence?

This entry was posted in Academic research, Data and statistics, Democracy and freedom and tagged , , . Bookmark the permalink. Follow any comments here with the RSS feed for this post. Both comments and trackbacks are currently closed.


  1. Pierre-Louis wrote:

    The resource curse exists. If the wrong hands get control of the oil (or gas or diamonds) revenues, they can use revenues to stay in power and rule as they wish. Think of Russia, Nigeria, Sierra Leone (diamonds), Brazil (, Angola and the list goes on…

    The resource curse means that resources can sustain bad equilibria. It doesnt mean that all countries with resources are doomed.

    Dont tell me u are not a bit scared things turn sour in Ghana now that oil revenues will start to pour.

    This is not a matter of econometric identification. Simple OLS make the right point here.

    Posted February 2, 2010 at 4:46 am | Permalink
  2. RJS wrote:

    I guess this gives lie to the Dutch Disease theory of natural resources. In retrospect, it was always a bit too simplistic. Good post. No someone alert the Tom Friedmans of the world.

    Posted February 2, 2010 at 5:40 am | Permalink
  3. I struggle to buy any argument that seeks to refute the ill effects (both economic and democratic) of natural resource dependency without even a mention of currency effects.

    Posted February 2, 2010 at 9:51 am | Permalink
  4. Nor inequality implications for that matter.

    Posted February 2, 2010 at 9:53 am | Permalink
  5. Roberto wrote:

    The very idea of a “curse” is problematic and relevant only as a metaphor. I have argued that it is better to think about a trap where an economy can fall or not. The rent-seeking trap (La trampa del rentismo) allows considering other variables: of course abundance and concentration of natural resources rents, institutional strength and social expectations (inequality, poverty and conflicts). For this, I agree, econometrics can be helpful, but not enough and sometimes inadequate.

    Posted February 2, 2010 at 10:09 am | Permalink
  6. Diego wrote:

    What Pierre-Louis said. I agree dependence seems like the wrong variable to look at. But how would one otherwise explain how Chávez & Co., to name just an example, stay in power for so long without resorting to more typically brutal means?

    Posted February 2, 2010 at 10:40 am | Permalink
  7. Chops wrote:

    As a young economist (job market of 2011, stay tuned), I’m bewildered that Sachs & Warner’s definition of resource-abundant was not laughed out of court.

    The most resource-rich countries in the world are Canada, the US, and Russia. Others in the top tier are the Arab Gulf and Norway. In S&W and similar papers, the only ones to make the cut as “resource-rich” were the Arab Gulf, and they show up as part of the problem, because the sample period starts during a high-oil-price era and ends in a low-oil-price era. So much of the growth regression literature hinges on the start- and end-dates! Why not just use levels?

    A better treatment of the problem comes from Stijns, at Northeastern, who deals with these issues, but still shows somewhat of a “curse”.

    Posted February 2, 2010 at 10:57 am | Permalink
  8. Eric Meade wrote:

    This post and the comments simply point to the (necessity but) insufficiency of econometrics as a tool for thinking about development. Many outputs from the statistical analysis that is so prevalent in development circles today are interesting and internally valid, but not particularly meaningful in terms of policy. This analysis intentionally omits the subjective and qualitative aspects of human experience, and thus limits itself to a set of objective levers that operate linearly. If you want this outcome, pull this lever a little. If you want the opposite, push it back a little. Strangely, the levers don’t often work. Another example is the upcoming CGD event where a Ph.D. candidate will show that 45,000 Tutsi deaths can be explained by radio station coverage during the Rwandan genocide. What insight does that give us for policy? Create “no transmit zones”?

    Posted February 2, 2010 at 11:06 am | Permalink
  9. Resource Blessing wrote:

    Does this mean Prof. Easterly will back off the pro-resource curse claims he makes in The White Man’s Burden? They weren’t important to his main points anyways, and they don’t look true anymore.

    Posted February 2, 2010 at 6:50 pm | Permalink
  10. Andy wrote:

    Spot on Eric, I had similar thoughts as I read this. Economics simply cannot capture the complexity and multifaceted aspects of societies and the challenges they face in trying to develop. Indeed, I am very skeptical whenever I see someone trying to “prove” (or disprove) anything by reducing the world to a small group of numbers. Qualitative or ethnographic studies provide far greater insight into these sort of matters, even if they are not “statistically significant”. At least they involve real people.

    Posted February 2, 2010 at 8:35 pm | Permalink
  11. avam wrote:

    Great response by Eric Meade. Development economics – and the worship of numbers to explain complex systems – is always going to come up short. It’s hard not to get frustrated by much of the ‘new research’ coming out – take paper 1 (“They find a positive correlation between resource abundance and both growth and institutional quality, and argue that it is conflict and poor institutional quality that lead to dependence” – no kidding) I have read work similar to this going back a few years. It seems to me one of the big stumbling blocks continuingly facing academia (and related policy) is the herd instinct of different areas of research. All academia seems to suffer from this – and an almost pathological desire to be insular and stick to their own (anthropology is no better).

    Posted February 3, 2010 at 5:38 am | Permalink
  12. My comments on this post are contained in a post, Development or just poverty reduction?, on my blog at:

    Posted February 3, 2010 at 5:33 pm | Permalink
  13. Quiet Griot wrote:

    “The Curse story does not claim that Nigeria is Britain plus oil, but rather that Nigeria is less democratic than Nigeria would be in the absence of oil. One way to get around this problem is to test whether oil makes country X less democratic using panel data with fixed country effects.”

    But the problem with Nigeria is that its political institutions *formed* in the context of natural resource abundance. You can’t use country level fixed effects to account for that, because there wasn’t a country except very briefly before the oil. The approach of the 2nd study in this context- using other West African countries as a baseline- is not very convincing at all (to me anyway). In the context of Nigeria, it’s hard for me to see how any reasonable observer could conclude that oil has not been detrimental to institutional development.

    I don’t agree with Eric’s blanket statement that econometrics is not a good tool for thinking about development. Rather, the problem is more that economists abuse econometrics all the time, and unless you’re an economist yourself (and often not even then) you don’t really have any basis to seperate the wheat from the chaff. Eric’s point is very much a valid indictment of the discipline and the way many economists go about things, but let’s not throw the baby out with the bathwater. A good, hard, honest look at the data is the one of the best tools we have for evaluating a lot of aspects of development. I agree that development economists should stop goofing around with fancy methodological tricks and stick to the basics; this to my mind is probably the most persuasive argument in favor of the recent trend toward RCTs.

    Posted February 3, 2010 at 10:29 pm | Permalink
  14. Eric Meade wrote:

    Sorry Quiet Griot if I was unclear. My intent is not to throw out econometrics altogether, but simply to recognize its limitations. We need to measure outcomes, but shouldn’t we also account for the fact that many of these models assume a Newtonian determinism long abandoned by complexity theory, developmental psychology, and even modern physics? Because we limit our inquiry to linear and quantitative measures, we conclude that poverty is just a set of unfortunate “traps” from which a “once-and-for-all big push” (Collier’s term) can deliver us. We talk about “reform” (objective and institutional), but what we really need is “evolution” (subjective and societal), or rather “co-evolution” between the objective environment and the subjective human society. (See also “social learning” in David Ellerman’s post above.)

    Posted February 4, 2010 at 12:07 am | Permalink
  15. Quiet Griot wrote:

    Eric, thanks for the discussion, I take your points about evolution vs. reform. I still don’t seem, though, how it is econometrics that leads to the conclusion that poverty is a series of traps that require big reforms to escape from. I would characterize the latter as a view that some economists hold, not a logical consequence of the use of econometrics.

    For example, what if someone were to use econometrics to study the conditions that best foster the “social learning” that Ellerman talks about, for example, by comparing the pace of economic development in different parts of China, and then trying to draw conclusions that could inform policy. Would this be objectionable?

    Posted February 4, 2010 at 10:01 am | Permalink
  16. Eric Meade wrote:

    QG, let me try to elaborate. Any person, system, or society will have two types of variables – quantitative and qualitative. For example, some quantitative variables that describe me are my height, weight, and annual income. Some qualitative variables that describe me are my race, belief system, and culture. Econometrics and statistics deal with quantitative variables. Faced with a specimen that has both quantitative and qualitative variables, a strictly econometric analysis will be blind to the qualitative variables, or more accurately, will misinterpret qualitative variables as the manifestation of a set of quantitative variables.
    Imagine, for example, that you met a few Germans, all of whom had blond hair and blue eyes. (Color is presented here as a quantitative measure of light.) An econometric analysis would miss the German-ness of the specimens, and would have to conclude that blond hair causes blue eyes, and blue eyes cause blond hair. (This is not dissimilar from Collier’s discussion of low income and conflict in The Bottom Billion.) If German-ness were for some reason viewed as undesirable (no offense to Germans – just using a metaphor here), then you may decide to intervene to pull the person out of the “German trap.” You might simultaneously dye the person’s hair brown and give him or her brown-tinted contact lenses, thinking that by adjusting these quantitative variables you would eradicate the person’s German-ness. But without a qualitative change in the individual from German to, say, Chinese, the person is still German underneath the brown hair and brown eyes.
    Similarly, I would suggest that a “poverty trap” is a qualitative difference seen through the lens of a quantitative analysis. Because qualitative differences are not allowed by econometric analysis, they must be the outcome of mutually reinforcing quantitative variables, such as low income and high rates of conflict. The conclusion is that if we could just pull societies out of that mutually reinforcing quantitative quagmire (e.g., military intervention leading to peacekeeping, institution building, and economic growth), then all would be right with the world. I’m not ruling out these types of interventions, but they will be successful only if they account for the qualitative, subjective aspects of that society’s culture and evolution. I’d suggest this is what the U.S. has been learning in Afghanistan.
    Regarding your question, it would absolutely be appropriate to use econometrics to study the conditions that support “social learning” across China. That type of analysis is necessary, but not sufficient. I would also want to study the cultural differences across China (ethnography), the developmental differences between and within the populations (developmental psychology), and other qualitative factors that would indicate whether “learning” or “co-evolution” – and not just material development – is actually taking place. My argument is not one of “either/or;” it is “both, and.”
    Similarly, you mentioned the value of RCTs, which is real. But think about RCTs in a health setting: Drug A may be more efficacious on average than Drug B, but I may be genetically unable to produce the enzyme required to break Drug A down into its active components. So for me, maybe Drug B is better.

    Posted February 4, 2010 at 11:24 am | Permalink
  17. I like the idea of a resource trap, because there does seem to be something going on. Australia was, for example, a significantly richer (per capita) country than the US throughout the C19th and early C20th. We adopted a “Deakinite” policy model aimed at (rural) exploiting resource-exporters for urban (union and manufacturing) interests via protection and wage arbitration. The long-term economic results were not good (though not disastrous either, due to the strength of our basic institutional structure). Our economic performance (particularly our comparative economic performance) has improved markedly as that policy system has been largely dismantled.

    But when I look at the oil-rich theocracies ISaudi Arabia, Iran) of the Middle East and compare that to the effect of silver on Iberia, particularly Spain, from the C16th-C18th, there are some strikingly similar patterns: autocracy, de-commercialisation, aggressive religious obscurantism. We forget that medieval Spain was a pioneer of parliamentarism (giving merchants representation via elected delegates was a Spanish, not an English, innovation: the English copied it later). The flood of silver undermined both parliamentarism and what had been a highly commercial society by giving the Crown the financial power to buy off/ignore commercial and other interests that the Dutch and English were forced to incorporate into their political processes. The Crown could also be as religiously obscurantist as they liked: which turned out to be quite a lot–after all, it provided a way of sorting who was “in” and who was “out” (how to be “in”) when it came to handing out the goodies.

    Posted February 4, 2010 at 7:52 pm | Permalink
  18. Adam Martin wrote:

    Thanks for the comments everyone. I was out of town when this was posted, so sorry for not weighing in earlier. Just a few points-

    I’m highly sympathetic to Eric Meade’s response. My interest in these studies was precisely in how easily this myth spread with just a few anecdotes and some bad econometrics. I think the abundance vs. dependence argument is a great example of why we shouldn’t take data as unambiguous representations of reality. Our bias should be toward skepticism of sweeping empirical claims. And we should give all sorts of empirical methods space to operate precisely because it is so hard to get the facts straight.

    The Brunnschweiler-Bulte piece is, by the way, OLS.

    Just a couple points that misunderstood what I was saying:

    -I’m not claiming there are no ill effects associated with natural resources. I’m saying these studies failed to find evidence that resources systematically effect democracy, growth, or development, and most of the popular studies that do have fundamental flaws.

    -It’s also not the case that these studies (or I) claim that there are NO countries that fit the Curse story, but that it’s not the systemic phenomenon other studies had argued for.

    Posted February 4, 2010 at 9:33 pm | Permalink

4 Trackbacks

  1. By uberVU - social comments on February 2, 2010 at 12:20 pm

    Social comments and analytics for this post…

    This post was mentioned on Twitter by bill_easterly: Aid Watch: new evidence shows Natural Resource Curse is a myth

  2. […] länder i Afrika drabbats av. Fast egentligen handlar det ju inte om naturresurserna, utan om institutioner. Jag minns att jag förde ungefär samma resonemang i en trafikstockning i Kenya i höstas, men det […]

  3. […] Martin argues that the ‘oil curse’ is a myth (and has the research to prove […]

  4. By the muslim world…vol. 2 « life as it happens on February 14, 2010 at 1:23 pm

    […] oil = less freedom…or does it? New research suggests that the “oil curse” commonly associated with petrol-rich, authoritarian […]

  • About Aid Watch

    The Aid Watch blog is a project of New York University's Development Research Institute (DRI). This blog is principally written by William Easterly, author of "The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics" and "The White Man's Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and So Little Good," and Professor of Economics at NYU. It is co-written by Laura Freschi and by occasional guest bloggers. Our work is based on the idea that more aid will reach the poor the more people are watching aid.

    "Conscience is the inner voice that warns us somebody may be looking." - H.L. Mencken

  • Archives