First, if it’s very uncertain how tax policy is going to affect innovation, why does that imply that taxes on the wealthy should be low? One assumption would be that innovators are very sensitive to the taxes imposed on them, and that if they are taxed too much they will just stop working (and live off their already-accumulated wealth, I suppose). But the opposite assumption – that, to the extent that people are motivated to work for monetary reasons, it’s generally to achieve a certain level of consumption – seems at least as plausible. And it implies that higher taxes would make those motivated to innovate by the promise of financial reward more likely to work harder – because they’d have to work harder to reach that level of consumption. (And those who are motivated just by the desire to innovate, and those who are motivated just by the desire to see their bank balance go up, would be unaffected by their tax rates – innovating is still the only way to innovate, and earning money is still the only way to make the bank balance go up.)
If we knew that, at any given level of taxation, higher taxes would depress innovative activity, I could see how skepticism about our ability to predict how far it would be depressed would militate in favor of low taxes. But if we are uncertain about the sign of the effect at a given level of taxation, how do you conclude that low taxes are always preferable?
It seems to me that the consequences of high taxation that we really need to worry about are uneconomic efforts to recharacterize income as something non-taxable, and the risk of high-earning individuals changing jurisdictions to avoid tax (which is a particular concern for states with high income taxes neighboring states with low ones – Massachusetts versus New Hampshire, for example). But, again, it seems to me that simply declaring that you can’t possibly estimate these effects doesn’t get you anywhere.
Second, Adam Ozimek’s argument is, basically, that there are positive externalities to the efforts of very innovative people that cannot be captured by them because they aren’t part of their product. Thus: even if Steve Jobs captured exactly 100% of the value he added to Apple, he can’t have captured 100% of the value he added to Apple’s employees, whose human capital was upgraded by working at Apple, value they will be able to monetize when they go off to start their own firms. Nor can he have captured 100% of his contribution to the creation of the massive cluster of innovation that is Silicon Valley.
But I’m not clear why Ozimek thinks this should be the case. If we posit a truly efficient market, then Apple employees know that they are getting experience at Apple that they will not get elsewhere – and Apple will know they know this. And Apple will therefore offer lower salaries than other firms that do not provide such experiences. Doctors who want to work in Boston often get paid less than those who are willing to work in Fargo in part because Boston has better weather and sushi restaurants – but largely because Boston is a huge medical hub and Fargo isn’t. That is to say: getting a job in Boston will do more to upgrade your human capital than getting a job in Fargo will.
I’m not saying I believe everyone earns 100% of the value of their marginal product. I’m just saying that I don’t see why Ozimek is so confident that high earners are an exception, and earn less than they contribute, for the reason he gives. And given that it’s pretty easy to find real-world examples of high earners who make more than their marginal product because they have found ways to extract rents, it strikes me as a little peculiar to take this particular tack against Krugman’s argument.
Third, it seems to me that Krugman doesn’t address the real implicit reason why people refer to high-earners as job creators: to whit, because they are good at deploying capital.
It’s not true, after all, that if Steve Jobs, say, earned exactly 100% of his marginal product that therefore nobody else benefited from his innovations. That would be true if he took his earnings with him to Mars, and never recirculated them into the terrestrial economy. But people don’t generally take their accumulated wealth to Mars; they usually spend or invest it.
Assuming no effect whatsoever on incentives, taxing high earners transfers their earnings to other people – either to the government, which then makes the decisions about spending or investment, or to the citizenry at large if the money is simply redistributed. If you do the former, you’re substituting the government’s views on how optimally to spend or invest the money for those of the high-earning individual. If you do the latter, you’re substituting the citizenry’s.
Saying, “these people are job creators” is another way of saying, “I think these people will be good investors, will deploy their wealth in ways that will further increase the productivity of society.” It’s basically saying: these are the right people to bet on if you want to deploy capital optimally.
(I want to stress: I’m not making this case; I’m just stating what I think the case is. I’m not sure what a good empirical test would be of the proposition.)
I think the “textbook economics” answer to this would be: if these individuals are genuinely better at deploying capital, capital – whether they own it or not – will flow into their hands to be deployed. A rich entrepreneur doesn’t actually need his own wealth to start a new business – he’d be able to borrow any money that was taxed away and make the same investments he would otherwise. If the government taxes a bunch of money away from great investors, that doesn’t actually reduce the amount of capital they have to deploy – because the money will flow back to these investors and out of other investors’ hands, and it’s most marginal investors who will lose their access to capital, to be replaced, effectively, by the government.
This question, “who will be a better investor” is not so simple a question to answer. It matters greatly what you mean by “better” – better for whom, and over what time horizon, and under what circumstances. Which is why I keep coming back to the question of what our spending priorities – not just the government’s, but our whole society’s – ought to be. But at this point I think we’re out of the textbook.
In my own opinion, if you want to encourage innovation, I wouldn’t focus on keeping taxes low; I’d focus on keeping them simple, because complexity rewards large firms with the wherewithal to do the financial engineering necessary to optimize their tax position, which in turn gives them an unearned competitive edge. And more generally, I’d focus on what other barriers incumbents have erected to prevent disruptive innovation and to extract rents. The existing patent system, which imposes a huge burden on small, young firms, is one good example of where to look. Most broadly, we should be trying to reduce the friction associated with innovation, rather than focusing on the monetary returns to innovation.