Sunday, December 6, 2009

Tagging-based taxation

One of the biggest challenges for public finance has been to design the most appropriate model for individual taxes. Any tax regime has to meet the twin requirement of fairness and efficiency, objectives which economists from the right and left view as conflicting. Left and liberal economists argue strongly in favor of progressive taxation that imposes the major share of the tax burden on the rich and thereby redistributes income. Conservative economists oppose higher taxes on the rich as inefficient since they dis-incentivize effort.

In the search for optimal tax policy design, Chris Dillow points to an old paper by George Akerlof that advocates the use of "tagging" - which tags (identifies) people and then makes specific transfers (or concessions) to them - in taxation. As Chris Dillow writes, "Given that taxes must be raised, therefore, it's best that they be imposed upon productive assets which won't be withdrawn from use if they are taxed". With their focus on inalienable and salient human characterisitics, tags also offer the additional advantage of controlling tax evasion.

Dillow proposes a tax design that would "tag" low-ability (or physically disadvantaged people, who can't work their way out of poverty and are therefore not likely to face incentive distortions with lower taxes or higher transfers) and high-ability (since high ability people would have to pay the same tax even if they work less, there is again no likelihood of incentive distortions) tax payers. Common tags include height, looks, gender or private education.

On the challenge with identifying appropriate tags for these characteristics, he turns to the large amount of recent reasearch on this area. In a famous paper, Greg Mankiw and Matthew Weinzierl had advocated an income tax system that includes "a tax credit for short taxpayers and a tax surcharge for tall ones".

Taxing women less than men, a proposal that even found support in last year's Spanish general elections, has been advocated by Alberto Alesina, Andrea Ichino, and Loukas Karabarbounis. They propose such gender-based taxation as a revenue netural policy by phasing out a variety of other policies already in place that favour women, like quotas, affirmative action, publicly supported child care facilities and care for the elderly.

Dillow also points to strong co-relation of private education with earnings, even controlling for university. And there is also the case for tax credits for the ugly, given evidence that "ugly people do less well at school and earn less than good looking ones - so much so that some turn to crime" and higher taxes on the beautiful given the co-relation between the beauty and intelligence.

And given evidence that ethnic minorities do badly in the labor market, there is also the case for tax credits for them. Is this also the economic rationale for taxation based on on other racial and communal characteristics (or "affirmative taxation" as in India!)?

While conventional progressive taxation does raise concerns about efficiency (in terms of incentive distortions), tagging-based taxation faces the same questions over fairness. In fact, the questions are even more relevant than those raised about regular taxation. For example, a tax on tall people as a tagged category, would unfairly impose a disproportionately larger burden on those tall people who are poor among them (compared to those poor among the short people). Even if have a mechanism to filter them out, the transaction costs associated with that process, would almost surely nullify any advantage with evasion and efficiency that tagging-based taxation enjoys. The only tag, I can think of, that would overcome this objection would be gender-based taxation.

No comments: