Some fallacies about the costs of anti-corruption

June 1st, 2011

I am rarely surprised when those who are not economists make egregious economic errors. I am only moderately surprised when those errors are in the Financial Times. I suppose I should get used to it.

Here’s a little basic economics. Economists, whether talking about banking or housing, trees or oil, always follow a set of fundamental principals. These include very basic ideas. For example, when conducting cost-benefit analysis, count the entire range of relevant benefits and costs. Say you want to do an economic analysis of a project to build a bridge. You count all the costs—labor, materials, etc.—and all of the benefits—shorter driving/commuting times, decreased congestion, etc. You don’t leave any out. Intuitive, right? Well, no, because a lot of people get it wrong.

The latest example of this sort of error was in the FT today:

If there’s a cost to corruption, there’s also a price to pay for fighting it. That is one way of looking at why the Philippine economy grew by a modest 4.9 per cent in the first quarter from a year ago – at the lower end of the government’s target range of 4.8-5.8 per cent and below median analyst forecasts… the deceleration was made worse by sharp cuts in government spending as the 11-month old administration of President Benigno Aquino III …froze many big infrastructure projects for review [of corruption]…Government statisticians calculated that if the new administration had allowed public spending to grow at the same pace as last year, GDP growth in the first quarter could have been a much higher 5.4 per cent. If outlays had been frozen at last year’s levels, growth would have been 5.1 per cent.

Okay so the argument here is that there is a cost to fighting corruption. I agree with that, partially. There are economic costs of fighting corruption—most obviously the cost of the personnel involved in anti-corruption efforts or the lost business revenue (if any) to companies who are precluded from engaging in it. I also agree with the (relatively hidden) sentiment in the article, which is that excessive bureaucracy, even when well-intentioned, can stymie economic growth. This is particularly true in developing countries.

But I do have a bit of a problem with the statement that lower growth GDP as a result of decreased government spending is a “cost” of anti-corruption. The government funded projects haven’t been cancelled. They are delayed. Money for those cancelled projects would (likely) go to another. Delays (and bureaucracy in general) are economic costs, yes. I don’t want to downplay that. But the cost of forestalling a project is the loss in benefits between implementing the project today and implementing the project at the delayed date—not the loss of the entire project. Going back to that bridge example: if we finish the bridge in a year from now, instead of today, the cost of the delay is the year’s loss of benefits—not the entire loss of benefits. The loss of economic growth now may be offset but acceleration later when the government adjusts. Or it might not. But failing to present these facts is misleading.

The article also completely misses the economic costs of corruption itself (or in other words, the benefits of controlling it). While the costs are many and I don’t have space to discuss them all, there is one in particular relevant to this story. That is the opportunity cost. An opportunity cost is the cost of any activity is measured by its foregone alternative. In the case of corruption they are obvious; if a large infrastructure project, for example, is rife with graft, it is more expensive. That added expense is money that could be well spent on other projects, which would encourage more economic growth.

The Financial Times asks: “Is the price worth paying? …  If corruption must be controlled, the arguement goes, so must be the struggle against it. Aquino and his officials should give the suggestion a try.”

Quite frankly there is a really excellent chance the price is worth paying. But the answer to that question requires an accurate description of the relevant economic conditions. That’s something this article failed to do.


Written by Ann Hollingshead

Follow @FinTrCo