How should we tax the hottest money?
Economist Anton Korinek spoke on the topic at this week’s Institute for New Economic Thinking workshop in Bangalore, India. Inevitably, it prompted some thinking on my part on the connection of finance to our trade and investment deals.
Anton’s paper outlines different policy options for countries hoping to avoid capital crises. In particular, he argues for a system of differentiated tax rates – applied before a crisis hits – based on the risk profile of the financial service provided.
The tax rate is estimated using the following parameters: what type of return an investor in a specific asset could hope to make after a crisis-induced devaluation, what type of externality that asset would impose on society, and the frequency of past crises.
Looking at the case of Indonesia in the 1998 Asian financial crisis, Anton notes that investors that bought dollar-denominated assets in the country in the lead-up to the crisis made out like gangbusters following the crisis-induced devaluation. Say you used 1 rupiah in 1997 to buy a 1 dollar asset. A year later, say a dollar can buy many more rupiahs. That lucky investor has made their dollar-denominated return, plus many more rupiahs when they convert their asset back. In contrast, those investors that purchased rupiah-denominated assets or assets (like FDI in retail sector) reliant on vibrant domestic sales do much worse. The numbers he uses are a 218% excess return on dollar-denominated assets, 100% return on inflation-indexed rupiah assets, and 44-66% return on non-adjusted rupiah stocks and bonds. These distinctions in rates of return form the first component of the calculation of the optimal tax rate
Anton then calculates the externality on the basis of a formula described on page 31, essentially – the hit national income takes because of the crisis. For the above assets, he puts these externalities at 30.7%, 14.1% and 6.2%-8.9%.
Finally, the externality is adjusted for the frequency of crises. If Indonesia has one crisis every 20 years on average, then the externality tax is divided by 20. This results in taxes on holders of the above assets of 1.54%, 0.71%, and 0.31% respectively.
Anton’s main justification for his approach is that it is neatly tailored to the more problematic forms of activity. This type of neat-tailoring is seen by most economists as more efficient than generalized policies (such as prudential measures) that impose costs even on market actors with non-problematic assets, even though they may have better ideas with what to do with their money. This cost on the latter group is a consequence of a less efficient policy approach. (Anton writes more about this here, in a paper for the IMF.)
Anton’s ideas would seem to produce smart policy, and wouldn’t require too much administrative capacity to pull off. In other words, all but the weakest states (who can’t even collect taxes) would be able to implement it.
His proposals pose some interesting challenges given the current provisions of trade and investment agreements, which feature various potential constraints on financial services policy space, as I write about here in a book through Boston University.
First, Anton notes that many capital control schemes invite evasion – evasion that is greatly facilitated by financialization and the emergence of derivatives. He suggests outright banning credit default swaps and other instruments in order to discourage innovation. However, as I note in this paper with economist Jayati Ghosh, such bans would likely violate market access provisions of the World Trade Organization’s services agreement.
Second, differentiated taxation could trigger liability under national treatment (or ‘fair and equitable treatment’) rules in investment treaties. After all, investors in credit-default swaps are taxed at higher rates that investors in local currency assets may be more likely to be foreign. If so, differentiated taxes, reserve requirements or quantitative control measures could be the basis for an investment treaty compensation claim by foreign investors. Given this, it may be useful to increase Anton’s tax in order for governments to create a reserve fund for their capital controls and other reserve funds. After all, if governments’ efforts to control externalities from harming society expose society to further financial liability through investment lawsuits, then further internalization of these lawsuit-related externalities is needed.
Of course, harsher taxation of foreigners with bad assets would prompt further liability, potentially creating a liability spiral, which brings me to my final thought.
Considered in the light of the above, Anton’s paper implies that trade and investment agreements might be privileging less efficient policies. Many trade agreements have ‘carve-outs’ for generalized prudential measures, but not for Anton’s targeted policies. (How effective the carve-outs are is a questionable proposition, as I write in the paper with Jayati, but that is another point.) This feature of trade agreements is especially ironic, given that these treaties are typically justified for their benefits of enhanced efficiency.