Does small-scale bribery matter? An empirical puzzle

Transparency International has released its new report on corruption in Latin America. One of the things that I like about the survey is that it asked whether respondents have personally had to pay a bribe. 

Let’s call it “retail corruption,” as opposed to “wholesale corruption” paid by businesses or stolen from public coffers. In theory, that strikes me as a more accurate measure than simply asking about overall perceptions.

The problem is that the survey gives some unexpected results. Is Trinidad really that clean? If it is — and it very well might be, asking people I know reports very low levels of retail corruption — does that imply that the retail corruption does not matter? After all, Trinidad suffers from very high rates of wholesale corruption. So does Brazil. Conversely, Chileans and Uruguayans report levels of retail corruption on a par with their Paraguayan and Argentine neighbors ... but corruption appears to play a much smaller role in economic life in the former two.

Consider Uruguay. 22% is a very high number of people needing to pay a bribe for public services! (Albeit not to use the courts or public utilities.) But in overall corruption perceptions it is between Japan and Estonia, slightly less corrupt than France. The country certainly does not feel particularly corrupt on the ground. Of course, it is consistent with the survey that Uruguay should feel much less corrupt than Mexico; but it also feels much less corrupt than Brazil! In a brief trip to Brazil, I came directly in contact with venal officials several times; equally brief trips to Uruguay have yielded nothing. 

Anecdotes are dangerous, of course. But the disjunct between the data and broader survey results, as well as personal experience, makes me wonder. Either the survey cross-section fails to capture actual differences or small-scale bribery is less damaging than I would have thought.  

So many questions, so little answers. I wish somebody would take the bribery survey and include the United States and Canada, to give us a baseline. 

Comments

[Tzintzuntzan]

Does Transparency International distinguish between retail corruption as viewed by foreigners, and retail corruption as viewed by natives?

Because you said that your trips to Brazil had you dealing with retail corruption in ways that Uruguay trips didn't, and world perception is that Uruguay's not as corrupt as Brazil. But Uruguayans report the same levels of retail corruption as Brazilians do.

Maybe in some countries, nobody dares demand bribes from foreigners. (In other countries the reverse would be true, and foreigners would be shaken down more.)