A question with an uncomfortable answer
The luck-of-birth question that runs through this volume has a version that stings anyone who lives where a permit requires a bribe and a policeman is a cost of doing business. If corruption is as ruinous as everyone says, the reasoning goes, then a country soaked in it is condemned, and there is nothing to do but wait for the graft to end before life can improve. Yet the reader asking this may also have watched their own country, or the one next door, grow at a furious pace for years while ranking near the bottom of every honesty league table, and the two facts refuse to sit together. The felt question is not moral, it is practical: does corruption always wreck an economy, or can a place get richer in spite of it, and even through it?
Like the other machinery chapters, this one opens by naming that question rather than a scene, because the answer to it is a piece of apparatus the rest of the volume borrows. That answer is that the word “corruption” hides two different things that behave in opposite ways. One of them prices like a tax and a country can grow while paying it. The other prices like a random, uninsurable risk, and that is the one that strangles an economy. Telling them apart, and seeing why the difference is not the amount of corruption but its shape, is the whole of the chapter, and it is why the title has a real answer: yes, corruption can coexist with rapid growth, but only one kind of it.
Two economies that both call it corruption
Imagine two countries with the same headline problem, that nothing gets done without a payment. In the first, the payment is one payment. There is, in effect, a single hand at the top that collects, sets a known rate, and in return sees to it that the permit is granted and the road gets built, if only because the people at the top want the economy to keep producing the surplus they are skimming. In the second country the payment is many payments, none of them final. Every clerk, inspector, and official along the chain is a separate operator with a separate demand, no one can promise what the next desk will want, and paying the first bribe buys no protection against the tenth. The total taken may be identical. What differs is that the first country’s corruption is centralized and predictable, and the second country’s is fragmented and open-ended, and that difference decides whether a business can function at all.
The economists Andrei Shleifer and Robert Vishny set this out in the language of monopoly. A single monopolist of corruption, they showed, behaves like any monopolist: it wants to set its price where the total take is highest, which means keeping the bribe low enough that business still happens, because the collector shares an interest in the economy staying alive. Break that monopoly into many independent collectors, each grabbing for himself without regard to the others, and the logic inverts. Each new official adds his own demand on top of the last, none of them internalizes what the pile does to the total, and the combined burden climbs until it chokes the activity that everyone was feeding on. They described the second case as corruption by independent monopolists, and saw it in the Russia of the early 1990s, where the old monolithic system for collecting bribes had fallen apart and an investor had to satisfy a queue of ministries and agencies that no longer answered to anyone. The counterintuitive result is that a country with one big thief can be less hostile to business than a country with a thousand small ones.
Why one prices like a tax and the other like risk
The reason the shape matters more than the sum is that investment is a bet on the future, and what kills a bet is not a high cost but an unknowable one. A firm deciding whether to build a factory can absorb a bribe the way it absorbs a tax, a tariff, or the price of steel, so long as it knows roughly what the bribe will be and trusts that paying it delivers the permit. Fold the expected payment into the sums, and if the project still pays, the factory gets built. What no firm can plan around is a cost that arrives without warning, in an amount no one can name, with no guarantee that meeting it ends the demand. That is the difference between a levy and a shakedown. Centralized corruption is the levy: ugly, unfair, a drag on growth, but survivable. Chaotic corruption is the shakedown, and a shakedown does not merely tax an investment, it makes the return on the investment impossible to calculate, which is the one thing capital will not tolerate.
There is an older and more hopeful theory that ran the other way. Writing in 1964, Nathaniel Leff argued that in a country tangled in bad rules, a bribe could be efficient, a way to grease past a foolish regulation and get the goods moving, so that corruption might even help development along. The empirical work of the following decades did not bear the hopeful version out. When Daniel Kaufmann and Shang-Jin Wei looked at whether firms that paid more in bribes actually spent less time tangled in red tape, they found the opposite: the firms paying the most bribes faced the most delay, not the least, because the officials who collect the bribes are the same ones who create the obstacles that make the bribes necessary. Grease and sand turn out to be the same substance. The delay is not a problem the bribe solves; it is the product the bribe is sold to relieve, and there is always more of it to sell. That finding is what leaves centralized corruption looking less like a lubricant and more like a tax that at least holds still.
What the cross-country picture shows, and hides
Line up countries by how corrupt they are seen to be and by how fast they have grown, and the average relationship is the expected one: more corruption tends to go with slower growth. The economist Paolo Mauro established the modern version of this in the mid-1990s, finding across a range of countries that higher corruption went with lower investment, and through lower investment, slower growth. The robust part of his result was the investment channel, that corruption deters the building of new capital; the direct line from corruption to growth was weaker and came and went as the controls changed, which is itself a hint that the story is not a simple one. The chart below is the same kind of picture with recent data, and it carries a caution that has to be read before the dots.
The chart earns its caveat twice over. The measure on the horizontal axis, the Corruption Perceptions Index, is exactly what its name says: a survey of how corrupt experts and businesspeople believe a country to be, not a tally of bribes actually paid, because bribes are secret and no honest tally exists. Perception can lag reality, and a country that cracks down can look corrupt for years after it has changed. And the clean, rich countries on the right grow slowly for reasons that have nothing to do with corruption, since a country already at the technological frontier has less room to catch up. Read past both cautions and the useful information is in the vertical scatter at the corrupt end of the chart. Countries seen as similarly corrupt have grown at wildly different rates, which is precisely what the “always kills growth” story cannot accommodate.
The anomaly, and its limits
The dots that do not fit are a cluster of East Asian economies that grew at record rates for years while ranking among the more corrupt places on earth. China lifted hundreds of millions out of poverty over exactly the decades its officials were, by every account including its own, extracting on an enormous scale. Suharto’s Indonesia grew fast for a generation while the president’s family became one of the richest on the planet. The pattern was pronounced enough that political economists gave it a name, the East Asian paradox, and the explanation most of them reach for is the one this chapter began with. The corruption in these cases tended to be centralized: a ruling party or a strongman’s circle that collected in an organized, roughly predictable way, and that, whatever else it was doing, had a stake in the economy it was feeding on continuing to grow. Investors could price the graft and still build the factory. That is corruption behaving like a tax, and a country can grow while paying a tax.
The paradox has to be handled with the same care as the resource curse in the last chapter, because it too can be overstated. Some scholars argue that when corruption is measured more carefully the East Asian cases look less exceptional, and that these economies grew despite their corruption, not because of it, and would have grown faster without it. That is very likely true, and nothing here says corruption is good for growth. The claim is narrower and it survives the scrutiny: predictable, centralized corruption is a burden an economy can grow under, while chaotic, fragmented corruption is a barrier it usually cannot, and the difference is large enough that the same measured level of graft is compatible with a boom in one country and a collapse in another. There is also a floor beneath the whole distinction that the earlier chapters laid. Corruption of the organized, taxable kind presupposes a state organized enough to collect it in the first place, which is why the centralized case tends to appear where there is real state capacity behind the graft (the chapter on state capacity). Where the state has collapsed entirely, there is no orderly corruption to be had, only the war of every small predator against the rest. And the graft discussed here is graft of the public kind, the abuse of a public office; the private swindle, the fraud that empties a pension or a savings account, is a different animal with its own home later in the series (Volume V, on why nobody went to prison for the scam that took your savings).
Corruption always kills growth.
Oversimplified Moderate confidence
On average, corruption does drag on an economy: across countries, more of it tends to go with less investment and slower growth, and the investment channel is the firmest part of that result. The word that fails is “always.” Whether corruption merely taxes an economy or strangles it depends on its shape. Centralized, predictable graft prices like a heavy tax that a business can plan around, and several economies, the East Asian high-growth cluster among them, grew for years while ranking as deeply corrupt. Decentralized, chaotic graft prices like an open-ended, uninsurable risk, and that is what actually freezes investment. So the claim is not the reverse of the truth, and it is not baseless; it collapses a real but conditional effect into a universal law and loses the distinction that does the work. The confidence is held to moderate by the evidence rather than the ruling. Corruption cannot be measured directly, so the cross-country picture rests on perception indices rather than counts of bribes, and separating what corruption does to growth from everything else that moves together with it is genuinely hard. The shape of the answer, that the kind of corruption matters more than the amount, is well supported; the precise size of the penalty is not.
Sources
- A. Shleifer & R. W. Vishny, “Corruption,” Quarterly Journal of Economics 108(3) (1993), pp. 599–617 — a single monopolist bribe-collector sets a lower total burden than many independent collectors, because uncoordinated officials each add their own demand and drive the total up; corruption is more distortionary than taxation because it must be secret.
- P. Mauro, “Corruption and Growth,” Quarterly Journal of Economics 110(3) (1995), pp. 681–712 — across countries, higher corruption is associated with lower investment and, through it, slower growth (instrumented with ethnolinguistic fractionalization); the robust channel is investment, the direct corruption-to-growth link weaker.
- N. H. Leff, “Economic Development Through Bureaucratic Corruption,” American Behavioral Scientist 8(3) (1964), pp. 8–14 (the “grease the wheels” hypothesis), against D. Kaufmann & S.-J. Wei, “Does ‘Grease Money’ Speed Up the Wheels of Commerce?” NBER Working Paper 7093 (1999) — firms paying more bribes face more delay, not less, because the officials who take the bribes create the obstacles.
- M. T. Rock & H. Bonnett, “The Comparative Politics of Corruption: Accounting for the East Asian Paradox,” World Development 32(6) (2004), pp. 999–1017 — the negative corruption-investment relationship holds among small developing countries but reverses among the large East Asian economies; the paradox is contested (some read the cases as growth despite corruption). Corruption is measured by perception (Transparency International CPI, comparable from 2012).
- Confidence is moderate under the rubric: the pip scores the weaker of evidence directness and construct match, and evidence directness binds here. Corruption is unobservable and proxied by perception indices, and its effect on growth is entangled with everything that co-varies with it, so the magnitude is not cleanly identified. The Oversimplified ruling is secure, since the “always” is refuted by clear high-growth, high-corruption cases; what the evidence cannot pin down is how large the penalty is and how much of it the kind, rather than the amount, explains.
Where the argument goes next
Corruption is a way a country’s own officials can bleed its economy from the inside. The next questions turn outward, to the ways the wider world acts on a country whether it consents or not. When other governments decide to punish a regime by cutting it off, do the sanctions bend the rulers or only the ruled, and do they change a regime’s behavior at all? That is the subject of the next chapter.