Max was on a quest. He sought wisdom from The Tax Guru to find the answer to one of life’s pressing questions: what is the best capital gains tax rate. He quested day and night for years, searching the world for the guru who could give him the answer. Finally, as he neared the summit where legend stated the guru lived, he saw the guru’s yurt. Max approached the yurt’s door and knocked. The door opened. Max peered in and saw The Tax Guru sitting at a desk, green eye shades on, meditating over the tax code. The guru stated, in a sure and clear voice: “Enter, my child. What do you have to ask The Tax Guru?” Max, humbled, fell to his knees, and said, “Tax guru. Tell me—what is the best capital gains tax rate? Some say it should be the same as the ordinary income tax rate. Others that it should be zero, at least in theory. Some want it higher, others want it lower. What rate is best?”
The guru replied, “It depends.”
And, it does depend. I recently saw a paper by Ole Agersnap and Owen Zidar, both of Princeton, that relates to what the capital gains tax rate should be. The paper, of course, never says exactly what the capital gains tax rate should be. Rather, it is very precise and exact in explaining what the authors are trying to do. The paper is extremely well done and will be well received and cited in the academic community. But, the paper will very likely get used by politicians who do not fully understand the content to argue that we should increase the capital gains tax rate.
So, what exactly does this paper do? It estimates the revenue maximizing capital gains tax rate. The thinking is that if you increase taxes too much, some people will work less, invest less, realize fewer capital gains, etc., and as a result, you might actually get less revenue from it. There is a sweet spot—a rate high enough to raise a lot of revenue, but, not too high that it discourages too much of the underlying economic activity being taxed (to be clear—the activity is still discouraged, just not enough to make the tax actually raise less revenue).
The question of the optimal capital gains tax rate is a very old question. Very early theoretical work suggested that the optimal capital gains tax rate should be zero. Then, empirical work commenced, and, as far back as 1988, there was already enough done in this area that the Congressional Budget Office wrote a review paper on the issue— a paper about other papers and results. That paper indicated that, “A number of statistical studies have provided strong evidence that realizations of capital gains decline when tax rates on gains are increased. The estimated size of this response of capital gains realizations, however, differs greatly among studies.” Much more empirical work has been done since then, and, new theory has been developed that challenges many of the assumptions that produced the optimal zero percent rate. As it turns out, we have gotten better at getting answers from data in the past 32 years.
So, what does this new paper find?
It finds that the revenue maximizing capital gains tax rate is between 38 and 47 percent. Let’s assume that their methods are flawless and the numbers are correct—what does that mean? A revenue maximizing tax rate means the tax rate that would raise the most revenue possible, but, is not so high that the decline in economic activity undercuts the tax increase. It assumes that the goal of the government is to extract as much income as possible from its citizens, so it can redistribute as much income as possible. I don’t know that, if polled, many citizens would articulate the goal of government in exactly that manner, but that is what the paper, like many other papers, assumes. It is an assumption that is useful and perhaps most importantly, I don’t know of an alternative assumption that would be tractable.
The main problem is that to find the “best” tax rate, you have to know what the objective is. What should the government be maximizing? The happiness/utility of its current citizens?* It’s hard to measure happiness, and it is difficult to justify restricting maximized happiness to only current citizens. Should the government then try to decrease inequality? To create social justice? To maximize GDP per capita on average? To create an environment where people are free to do as they chose without others interfering? To create a just, verdant, and peaceful world?
This is, of course, not really an answerable question. Even if we knew the answer, it is likely the role of government to maximize something that can’t be measured very well, and as a result, you cannot derive the optimal tax rate. Thus, it is just easier to assume revenue maximization, which is why everyone does it, and it is a useful exercise. And, for those into redistribution, it might actually be the right thing to maximize.
In other words, if the maximizing rate is higher than our current rate, that tells us we can have more revenue by enacting a higher rate, which is useful to know if we want more revenue. But, we might not want more revenue. More revenue, over the long term, allows a government with greater scope, which is not the objective of a large portion of the country.
So, back to Max and The Tax Guru. On what does the answer depend? Well—a lot of things, but, principally, on what the role of government is, which then determines what the goal of a tax is. That is a deep question on which it is impossible to get agreement, and, as a result, what the “right” capital gains tax rate is not really defined. Even for a tax guru.
____________________________________
*Technical note: If the government is maximizing combined happiness/utility, and we define utility over consumption, and the increases in happiness you get from consumption increase at a slower rate the more you consume (you get less happiness from spending your 1,000,000th dollar than your 10,000th dollar), this does imply a revenue maximizing tax rate is optimal, because it allows you to take as much as possible from high income people and allows you to redistribute as much as possible to lower income people (who get more utility from consuming it). But, yet again, it may not be the goal of government to maximize total happiness by taking from some to give to others. Rather, some may prefer a government which makes possible an environment where people can maximize their own happiness. Further, to some extent, this is literally assuming that “money buys happiness”, and that people’s utility is defined in large part over consumption.
