Brian Riedl and Kyle Pomerlau have both argued in NRO that a new top marginal tax rate of 70 percent on income over $10 million is unlikely to raise much in the way of new revenue, and I’m inclined to agree with them. If your goal is to raise enough new revenue to finance a massive expansion of social-insurance programs and redistribution writ large, you’d really need to raise taxes on middle-income households, as Charlie observed earlier this week. But that, of course, would be politically perilous. If your goal is to raise a meaningful amount of new revenue from the highest-income U.S. households, you’d need a more comprehensive approach, e.g., moving towards mark-to-market taxation (much easier said than done, but bringing an end to the “step-up in basis” would be a start), reforming corporate taxation by further tightening transfer-pricing rules, etc., to stymie the advent of new tax-avoidance strategies.
But what if raising revenue isn’t really the point? To some thinkers on the left, high marginal tax rates are seen less as a revenue-raising measure than as a means of altering the pre-tax distribution of income. For argument’s sake, let’s stipulate that we ought to care about household income inequality. Why care about inequality before taxes and transfers in particular?
The U.S. is often singled out as an outlier among market democracies when it comes to household income inequality. It is worth noting, however, that when we examine inequality before we account for taxes and transfers, a measure known as market-income inequality, the U.S. doesn’t stand out nearly as much. Germany’s market-income distribution, for example, is very similar to that of the U.S. The difference is that Germany’s post-tax-and-transfer income distribution is far more egalitarian. Some other European market democracies, such as the Netherlands, are considerably more egalitarian even before we account for taxes and transfers.
So what’s the difference between societies that start out pretty egalitarian before taxes and transfers and remain so after we account for them, and those that start out pretty inegalitarian and require heroic levels of redistribution to get to the same egalitarian place when it comes to disposable income? If all that matters is the distribution of disposable income, regardless of how you get there, there is no reason to prefer the Dutch status quo over the German. Perhaps the Dutch situation — not a lot of inequality to begin with, taxes and transfers mostly just shuffling money around — is more sustainable than the German one, as those on the wrong end of redistributive policies are less likely to feel burdened. Whether or not that’s true, one can certainly have a preference for a more egalitarian distribution of pre-tax income, which is where the idea of punitive taxation comes in.
One story I’ve often heard on the left is that low marginal tax rates don’t just nudge high-income professionals and corporate executives and the like towards working long hours or moving to more productive regions, where they can command higher incomes; it can also tempt them to engage in self-dealing behavior that could prove destructive, cutting corners to accumulate as much as possible. With higher marginal tax rates, there’d be less incentive to go quite that far, as you’d only capture so much of the proceeds. Nostalgists might point to midcentury America, where corporate chieftains faced much higher marginal tax rates, at least on paper, while at the very least talking a good game about their public-spiritedness. If you believe this story, and I should say I’m not convinced, it’s not hard to see why ultra-high marginal tax rates might appeal to you, even if they wouldn’t boost revenues (or even if they’d decrease them).
What I find striking about this line of argument is its resemblance to the case for immigration restriction. It is possible to accept that labor migration, regardless of its average skill level, can expand the economic pie while still favoring a more restrictive or more selective immigration policy on the grounds that, say, the composition of the immigrant influx has bearing on pre-tax inequality (as a general matter, a high-skill influx will tend to mitigate domestic inequality while a low-skill influx will tend to exacerbate it, in part by changing the composition of the overall workforce). In the case of high marginal tax rates, you can say that yes, high marginal tax rates might represent a hit to GDP, but the benefit of that unrealized economic activity would have flowed to high earners, so let’s not shed any tears. In the case for immigration restriction, you can say that yes, limiting low-skill immigration might represent a hit to GDP, but the benefit of that unrealized economic activity would have flowed almost entirely to the migrants themselves, so it is misleading to suggest that incumbent citizens are bearing a significant cost. In either case, the hit to GDP is seen as less important than some other consideration; it is not enough in itself to settle the argument one way or another.