Why does striking oil so often make countries poorer and less free?

The question a lottery ticket raises

The question that runs under this whole volume, whether the shape of a life was mostly settled by where it began, takes an odd turn for anyone born over a fortune. If your country is floating on oil, or diamonds, or copper, the luck of your birthplace looks like a winning ticket, and the natural expectation is that the ground beneath you should have made you rich. For a great many people it did the opposite. The reader this chapter is written for may be sitting in a country that exports a fortune in crude every day and still cannot count on the lights staying on, and is asking the plain version of the question: the wealth is real, so where did it go, and why did the country around it not get freer as it got richer?

This is machinery the rest of the volume leans on rather than a lived scene, so it opens by naming the puzzle rather than a morning. Across the second half of the twentieth century, the countries that leaned most heavily on resource exports tended to grow more slowly than countries with nothing much under the ground, and many of the biggest oil exporters became not more accountable to their citizens but less. That pattern earned a name, the resource curse, and a large and quarrelsome literature. This chapter sets out what the record actually shows, why a windfall can leave a country poorer and less free through two quite different channels, and why the story the evidence tells is not that oil is a curse but that a windfall magnifies whatever state and rules were already there.

A pattern that looked like a curse

The finding that started the alarm is a correlation. In the mid-1990s the economists Jeffrey Sachs and Andrew Warner lined up countries by how much they leaned on natural-resource exports at the start of the period, measured as the value of primary-product exports as a share of the whole economy, and by how fast they then grew over the following two decades. The more an economy leaned on resource exports at the outset, they found, the more slowly it tended to grow afterwards, and the relationship survived the usual controls for starting income, trade policy, and the rest. Almost none of the countries that were most dependent on resources in 1970 went on to grow quickly. The chart below shows the same shape with a longer runway.

Two things are true of that picture at once, and holding both is the whole task of this chapter. The tilt is genuine: the countries most reliant on selling what came out of the ground really did tend to grow more slowly, and some of the largest oil states are poorer per head today than they were two generations ago. Venezuela, which in the early 1970s had the highest income per person in Latin America, saw real income per head fall to under half of its old level. Nigeria, Africa’s biggest oil exporter, where crude supplies the large majority of export earnings and government revenue, has almost the same income per person it had in the early 1970s. At the same time the spread around the tilt is enormous, and the exceptions are not small countries no one has heard of. Norway pumps oil from the North Sea and is among the richest and most democratic places on earth. Botswana built one of the longest growth records in the world on diamonds. A single line drawn through the points hides more than it shows.

The first channel: a windfall breaks a bargain

To see how a windfall can leave ordinary people poorer and with less say, start with the state, because that is where the money actually lands. A country does not strike oil the way a person wins a lottery. The rent, the pure surplus left after the cost of getting the resource out, accrues to whoever holds the wellhead, and in almost every case that is the state or those who control it. The previous chapters built the tool for seeing what happens next. A functioning treasury normally has to raise its money by taxing the people, and the need to tax is what forces a government to bargain: to account for the money, to tolerate the scrutiny that comes with asking for it, and to keep enough of the population content to keep paying (the chapter on why some states can tax and others cannot). A large enough flow of resource rent dissolves that need. A government funded by a pipeline does not have to ask its citizens for anything, and a citizen who is not being taxed has lost the oldest lever there is for demanding a say.

This is the rentier-state argument, and the political scientist Michael Ross put numbers to it. Studying 113 states from 1971 to 1997, he found that the more a country relied on oil and mineral exports, the less democratic it tended to be, and that the effect reached well beyond the Middle East and did not show up for ordinary farm exports. The proposed mechanism is the one the diagram draws: rents let a government spend on patronage and hold taxes low, and both dull the pressure for accountability that taxation would otherwise create. It is worth being exact about how firmly this leg stands, because it is the more contested of the two. Using long runs of history within each country rather than comparisons across countries, Stephen Haber and Victor Menaldo found no robust link between resource wealth and authoritarian rule, and in some versions of their test resource income went with more democracy, not less. Others read their own data back the other way: Jorgen Andersen and Ross argued that the null holds only for the era before the 1970s, and that a distinct oil-and-autocracy pattern appears after about 1980, once governments in the developing world took over the rents that foreign firms had previously carried off. The fiscal-bargain logic is clean and old; whether it bends a country toward autocracy in general, or only in some places and some decades, is genuinely unsettled.

The second channel: the rest of the economy is priced out

The other way a windfall can impoverish is purely economic and needs no bad ruler at all. When a country suddenly earns a flood of foreign currency for its oil, that money is converted and spent at home, and the spending bids up the country’s wages and prices, its currency, or both. What they share is the effect that matters: the country becomes a more expensive place to produce. Imports turn cheap, which feels like a blessing, and everything the country makes for export costs foreign buyers more, which is not. Factories and farms that used to compete in world markets find they no longer can, and they shrink. The economy narrows toward the one thing the world will buy at any exchange rate, the resource itself, and loses the industries that might have employed people and built skills long after the deposit runs dry. Economists call this Dutch disease, after the name a 1977 article in The Economist gave to what happened to Dutch manufacturing after the Netherlands developed a giant natural-gas field.

Between them, the broken bargain and the priced-out economy explain how a country can sell a fortune every year and grow poorer and less free while doing it. The rent detaches the rulers from the ruled, and the exchange rate hollows out everything the country made before the discovery. Neither mechanism requires anyone to be unusually wicked or unusually foolish. They are the default tendencies of a large windfall meeting an ordinary state, which is why the pattern shows up so often that it looks like a law.

The measure that turns the curse inside out

It is not a law, and the reason is buried in how the original pattern was measured. Sachs and Warner sorted countries by resource exports as a share of the economy, and called it resource abundance. It is not abundance. It is dependence: a country shows up as resource-rich on this measure not when it has a lot in the ground but when its economy has little else going on, so that the resource looms large in the total. That distinction sounds like hair-splitting and is the hinge of the whole debate. A country can be dependent because it is poor and undiversified, in which case the low growth and the dependence may share a common cause rather than one driving the other. The economists Christa Brunnschweiler and Erwin Bulte made this the center of their re-examination, separating true abundance, the value of the resources a country actually holds, from dependence, the share of the economy they make up. When they did, the curse changed shape: dependence turned out to be an outcome of weak institutions rather than a cause of slow growth, and measured abundance, if anything, went with faster growth and better institutions, not worse. Their title called the curse a red herring, which is stronger than the evidence can fully carry, and their own abundance measure has in turn been criticized. What survives the exchange is narrower and firmer: the headline correlation confounds a resource with the weakness of the economy around it, and cannot by itself tell you which way the causation runs.

This is why the scatter earlier is a trap as well as a fact. The very countries that turned a windfall into lasting wealth, Norway above all, appear on the dependence measure as only modestly resource-reliant, because their economies were already large and varied enough that oil was one earner among many. The countries that were captured by the windfall appear as the most dependent, because there was little else left. The measure that seems to show resources dragging countries down is partly recording, after the fact, which countries had strong enough economies to keep the resource in proportion. Directness of evidence is exactly what the resource curse lacks: it is a correlation whose central variable is tangled up with the thing it is supposed to explain.

Same oil, opposite countries

Put the confound together with the exceptions and a cleaner account appears, the one most of the field has settled toward. Resources are neither a blessing nor a curse on their own. They are a multiplier, and what they multiply is whatever institutions were there to meet them. The economists Halvor Mehlum, Karl Moene, and Ragnar Torvik framed it as a contest between two kinds of arrangement. Where the rules reward production, a windfall is invested and the country grows richer; where the rules reward grabbing, the same windfall is fought over and the country grows poorer. Whether striking oil is a lottery win or a curse depends on which set of rules the discovery lands in, and that was mostly decided before the first well was drilled.

The contrast between Norway and Nigeria shows the two directions the same discovery can take. Norway found oil in the North Sea in the late 1960s as an already-wealthy country with strong courts, a capable treasury, and a habit of public accounting. It chose to save most of the proceeds abroad in a sovereign fund, now the largest in the world at well over a trillion dollars, precisely so that the money would not flood the currency and hollow out the rest of the economy, a deliberate answer to Dutch disease. Nigeria found oil at independence with weak institutions still forming, and the rent became the prize in a long contest for control of the state rather than a fund for its citizens. Botswana did with diamonds what Norway did with oil, and became one of the fastest-growing economies in the world for a generation, because at independence it built the courts and the fiscal discipline to treat the windfall as public revenue rather than private spoils, the institutional story the earlier chapters traced (the chapter on colonialism’s legacy). The resource did not decide these outcomes. It revealed and magnified a decision that had already been made about who the state answered to. What a petro-state crisis looks like when the rent collapses, and a currency and a government go down with it, is a story the volume on how economies break takes up (Volume V, on how a money can die); here the point is only that the discovery itself did not write the ending.

Where the argument goes next

A windfall is one test of whether a state answers to its citizens or to whoever holds the wealth, but it is not the only one. Even without oil, a government can be captured in smaller, everyday ways, through the bribe and the kickback, and the same question returns in a different form: does that capture always ruin an economy, or can a country grow in spite of it, and even through it? Why some corruption merely taxes growth while other corruption strangles it, and why a cluster of fast-growing economies got rich while ranking among the most corrupt on earth, is the subject of the next chapter.