Edward Conard, a former Managing Director at Bain Capital, has been creating quite a stir with his new book, Unintended Consequences. I haven’t had a chance to read it yet, but I’ve read this article in the NYTimes about it and just saw his pretty epic 3 part Daily Show interview (1, 2, 3) that you shouldn’t miss.
Conard argues that the key to generating economic growth and innovation is to reward risk-takers. We need to get guys to walk away from Google to go out and spin the innovators’ roulette wheel, so any policy (such as high capital gains taxation) that makes leaving Google to spin the wheel less attractive is a really bad idea. He claims that lower innovation, and thus lower productivity growth, in Europe and Japan should provide a stark reminder of what happens when we don’t reward risk-takers enough.
What’s wrong with this argument
1. Omitted Variable Bias:
Japan and Europe are different from the US for many reasons, so comparing one aspect of Euro-Japan to the US and attributing all differences in productivity growth to that aspect can produce misleading conclusions.
2. Which Investments Don’t Happen with Higher Capital Gains?
Conard has a strange economic model in his head about how the innovators’ roulette wheel works. My view is that it seems more realistic to envision a return distribution that governs the potential returns of different investments. To start with the simplest framework, suppose the returns are known beforehand. Drawing a high return is rare, but doubling the capital gains tax rate isn’t going to make investing in the next Facebook or Apple unattractive. Instead, a higher capital gains rate will crowd out “shovel ready investments.” By this I mean investments that haven’t happened yet (and are thus shovel ready) because they produce unexciting and small returns. The costs of higher capital gains within this framework is not preventing the next Facebook or Apple. The cost is preventing mediocre projects (by shifting the threshold determining which marginal investment makes sense). Mediocre projects don’t have first order consequences for productivity, job creation or growth.
Roy model with Risk Averse innovators:
To be fair to Conard, he didn’t describe some imaginary, exogenous return distribution. Instead, he seemed to be describing a Roy model with risk averse innovators who need to be compelled to leave Google or Bain Capital. They are risk averse because they need the 1 out of 100 draw (that one gets after creating a successful startup) to be enormous to compel them to leave. Conard implicitly argues that a 40% capital gains rate makes this much harder than a 0% rate. In other words, they need a 1 in 100 shot of getting $500M rather than a 1 in 100 shot of getting $500M *(1-.4)=$300M. This assertion amounts to a statement about risk aversion.
It’s far from clear that the potential gains from getting these extra marginal workers from Google (who would take 1/100 for $500M but not 1/100 for $300M) are worth the cost of not collecting capital gains taxes, especially if that means not collecting any capital gains that pay for crucial investments in the education system, national defense, and social insurance. In fact, these crucial investments make it much more likely that we maintain an economically mobile society that produces a larger potential pool of innovators. Undercutting investments in social insurance and education is horrible policy even if you only care about generating innovation.
I’m not arguing that getting Google & Bain Capital workers to become innovators would be bad, but I strongly doubt that abandoning a progressive tax system is anywhere close to the right answer to this issue. A system in which one has to win the lottery of birth in order to win the startup lottery is not only inequitable, but also inefficient.
|The Daily Show with Jon Stewart||Mon – Thurs 11p / 10c|