De volta ao controle de capitais

O Ed me mandou o link para o blogo do Rodrik com seus comentários acerca das objeções habituais ao controle de capitais. Em itálico meus comentários a respeito.

* * *

Nonsensical arguments against capital controls

1. Capital controls result in corruption and rent-seeking. Well, perhaps yes, sometimes they do. But not always, and surely the type of capital controls you are talking about makes a difference. I have yet to hear anyone make the argument that the Chilean tax on capital inflows led to corruption or that long-standing Taiwanese controls have been overwhelmed by rent-seeking. Our job as policy advisors is to design policies that minimize the risk of corruption while their primary objective is being served. Government regulations on environmental externalities, health, or consumer safety are all subject to corruption and "capture" by the private sector as well, but most economists take this as a reason to think of better-designed regulations, not as a reason not to regulate. People who do not understand this should not be in the business of providing advice to governments.

2. The problem is not with capital flows per se, but with the underlying market distortions that induce risky behavior by financial intermediaries and by borrowers. So policy should target these distortions directly, through appropriate prudential regulation, rather than target the flows themselves. Yes in principle, but no in practice. If one could design the perfect prudential regulatory regime, able to handle all future financial innovations, then indeed we would not need direct controls on capital flows. But if we cannot, and we surely cannot, we need to work on as many margins available to us as we can. That is where the gun control analogy is really helpful. If you could perfectly regulate the behavior of future criminals, you would not need controls on the sale of guns directly. It is people who kill people, not guns--remember? But most of us are reasonable enough to realize that we have imperfect control over the behavior of gun owners and so we think direct gun controls make sense.

Eu acho curioso. Rodrik admite que problema esteja associado às distorções de mercado, que podem ser tratadas, em princípio, por regulação “prudencial”, mas apenas se a regulação for perfeita (“If one could design the perfect prudential regulatory regime, able to handle all future financial innovations, then indeed we would not need direct controls on capital flows”). Por outro lado, ao rebater a crítica sobre a possibilidade de "rent seeking" e corrupção ele diz que “Our job as policy advisors is to design policies that minimize the risk of corruption while their primary objective is being served (…) People who do not understand this should not be in the business of providing advice to governments”.

Há uma clara assimetria: podemos impor controles de capital porque somos (ou deveríamos ser) espertos o suficiente para evitar problemas de corrupção e "rent seeking" (senão não deveríamos sequer entrar no negócio de recomendações de política econômica); por outro lado, (segundo Rodrik) algum defeito inato nos impede de fazer o mesmo do ponto de vista de regulação bancária, já que só a regulação perfeita impediria o problema. Isto é, devemos ser brilhantes para impor controles de capital, mas somos misteriosamente incapazes do mesmo brilhantismo no que tange à regulação do sistema financeiro. Então por que ele está mesmo no negócio de propor recomendações de política econômica?

Eu não conheço o caso de Taiwan (vale a pena investigar), porém, do que conheço do caso chileno, pode até não ter havido corrupção, mas os próprios chilenos abandonaram o sistema, o que me sugere que não era a maravilha que o Rodrik apregoa

3. Capital controls won't work because they are easy to evade. Surely, some leakage is inevitable, but it is paradoxical that the same people who make this argument are those who cry bloody murder at the mention of capital controls. If you can evade capital controls at little cost, you should simply be unconcerned. And if you can evade them only at a cost, well then capital controls are working! Or as my co-author Arvind puts it, ask the people who make this argument whether they will deny that lifting capital controls will cause an increase in the volume of capital flows?

Não acho que o problema seja só que estes controles sejam fáceis de contornar. O problema é que sempre que houver diferenças de preços de ativos haverá um incentivo à arbitragem, pois ninguém deixa notas de 100 euros (nestes tempos de dólar em queda) no chão. Tipicamente será o sistema financeiro que criará a engenharia financeira para evadir o controle e fará grandes lucros sobre isto. Vai aumentar o custo de capital para as empresas (ver abaixo), o que pode ser realmente o motivo dos controles, mas vai criar um jeito fácil de bancos fazerem dinheiro às custas destas empresas.

4. Capital controls will raise the cost of finance to some firms. Duh? That's the whole point of capital controls...

Posted in: