Wednesday, December 29, 2010

Are consumption taxes more equitable?

There is no doubt that consumption taxes are more efficient that labor income or capital income taxes, because they do not punish activities one would like to see promoted in an economy (labor supply, investment). But they are widely regarded as unfair, as the consumption share of income is higher for poor people. Hence the implementation of exclusions for essential goods where consumption taxes exist, in order the achieve some tax progressivity.

Isabel Correia claims that switching from income taxes to consumption tax can lead to less inequality even in the absence of lump sum transfers. This is a very counterintuitive result, and this is probably the reason why it made it into the American Economic Review (Yes, I know, I am breaking a trend here). But despite my best efforts, I still do not understand how this could happens, and the article provides very little in terms of explanation. Not only is no intuition provided, but the idea of using Gorman aggregation to reduce the model to a representative agent model seems wrong in this context. If anybody has read and understood the article, please help me here.

6 comments:

mOOm said...

The result without transfers (section II of the paper) depends on labor and capital each being taxed at a constant rate with no tax free amounts/deductions. Income is taxed at an increasing rate due to inequality of wealth distribution. Common prices are needed for the aggregation result hence the constant tax rates.

I didn't read the whole thing. But given this I'm surprised AER published it. Or maybe I shouldn't be.

Vilfredo said...

Even the best journals sometimes publish bad research. An author got lucky. But what is particularly bad now is that is now public that consumption taxes are fair, which may only be true in a very peculiar case. This is very misleading and could have some really bad consequences.

Christos Koulovatianos said...

The source of the seemingly paradoxical result is the dynamic nature of the model which breaks the typical intuition stemming from static models.

Poor agents have perfect foresight in this model, see the tax burden of future consumer taxes and may save more. Proposition 2 in the paper makes conditions behind this possibility this very clear.

On the comment,

"...the idea of using Gorman aggregation to reduce the model to a representative agent model seems wrong in this context.",

please understand that using Gorman aggregation is one of the few known contexts (and the only robust context known to me) within which such a welfare exercise is possible. The reason is, outside Gorman's exact linear aggregation social indifference curves in a utilitarian framework may intersect (see http://www.jstor.org/stable/1906943 ). Apart from Gorman's (1955) intuitive article on this broad subject of welfare intransitivity concerning the comparison of policies (see http://www.jstor.org/stable/i325647 ), Kirman (1992, p. 123-125 - see http://www.jstor.org/stable/2138411 ) explains the (high) possibility of welfare misrepresentation through a picture, while one of the most complete treatments of this subject is by Jerison's 1994 REStud paper "Optimal Income Distribution Rules and Representative Consumers" ( see http://www.jstor.org/stable/2297917 ).

This paper by Isabel Correia is excellent, and, in my opinion, it is an excellent addition for the AER. The reasons are two. First, the policy exercise is undertaken within a Walrasian model with complete markets. Walrasian models pose a clear interplay between scarcity of resources and preference, so values assigned to goods through market activity are logically coherent with a welfare analysis. Second, as explained above, she has made an excellent assumption (Gorman aggregation) in order to make the comparison of policies robust.

That Correia has posed a good question properly is a result of her literacy in general equilibrium and her scholarship.

The results of the paper are very precious, but only as results subject to the model's assumptions: they are very far from guiding policy prescription through solid scientific method. In my opinion, the results of this paper demonstrate the shortcomings of the Walrasion paradigm as a whole. My penny of thought on this broad subject of consumption taxation is that the problem is perfect competition and capital pre-aggregation in the Walrasian model. Both of these assumptions contribute to having no structure-induced markups in the model which could cause a channel of distortion for investment.

In brief, we may lack the correct paradigm for posing this particular question properly. (Yet, see efforts by Egbert Dierker and Brigitte Grodal on this key subject of introducing imperfect competition in general equilibrium here: http://ideas.repec.org/a/tpr/jeurec/v4y2006i2-3p436-445.html )

Christos Koulovatianos said...

One more comment about what we have probably learned from this paper.

The results of the paper, are very precious but only as results subject to the model's assumptions: they are very far from guiding policy prescription through solid scientific method.

In my opinion, these conclusions demonstrate the shortcomings of the Walrasian paradigm as a whole. My penny of thought on this broad subject of consumption taxation: the problem in the Walrasian model is perfect competition and capital pre-aggregation. Both these assumptions contribute to having no structure-induced markups in the model which could cause a channel of distortion for investment.

By the above I mean that consumption taxes (perhaps capital taxes, too), may have to wait until the development of a general-equilibrium theory with "large" firms and markups derived through strategic investments. (See efforts by Egbert Dierker and Brigitte Grodal on this key subject of introducing imperfect competition in general equilibrium here: http://ideas.repec.org/a/tpr/jeurec/v4y2006i2-3p436-445.html )

Anonymous said...

The article offers zilch in terms of intuition or explanation about the limitations of the results. The AER is a generalist journal, and thus read by non-specialists, and it should require a serious amount of explanations. The editor-in-charge for this article was asleep at the wheel.

Economic Logician said...

Thanks for the great comments! I think I learned a lot...