How to Think About Inequality Data and Public Policy

Lately I’ve been thinking about the disagreements among economists about the extent to which inequality has increased in recent decades. I am facilitating a reading group at my university on inequality this semester with some great undergrads, so it has very much been on my mind.

With conflicting data showing different trends, how are we as economists to judge this? How can the general public even have a clue how to judge this?

You may have seen this chart before. It comes from an article in The Economist, which actually does a really good job of explaining the debate over the data if you know nothing about it.

TaxProf Blog

According to some estimates, the share of income going to the top 1% has doubled and is now over 20%. That sounds bad. Maybe we need some more redistribution. Maybe a wealth tax.

But according to other estimates, and taking account of our existing system of progressive taxes and redistribution, the share of income going to the top 1% has not risen at all, is only about 5%. Less worrisome. The existing system of taxes and transfers seems to be doing a pretty good job, or at least no worse than in the last 60 years. No need for a new wealth tax, etc.

So who is right?

Sorry, I don’t have the answer. I think I’m pretty good at digging into economic data (follow me @jmhorp on Twitter for an almost daily dose of data debunking!), but I am no expert in this area. There’s probably only a dozen economists that really understand this data and the trade-offs in different forms of measurement.

So instead of giving you the “correct” answer, I offer you a chance to reflect. Our temptation is to say the “correct” data is the one that comports with our political preferences. If you are a progressive, you probably think inequality is bad and getting worse. Piketty is your man. If you are more of a libertarian, you probably think it’s about the same as recent years. Auten and Splinter must be right!

Stop. And instead, consider how you might view the policy implications of the data you don’t like being the correct data. If you are a progressive, would you still think we need a wealth tax even if the Auten and Splinter data is correct? If you are a libertarian, would you still think things are just fine and maybe we should cut the top tax rate if it turns out that Piketty and co-authors have the real data?

If you answer is the same for the policy implications regardless of what the data say, you might want to check yourself. And if so, why are we even arguing about the data?

Perhaps your answer is “I might have the same policy answer regardless of the data, but there are people out there that are convinced by data.” I think that’s possibly reasonable, and I would like it to be true, but where are these people?

Perhaps the answer is “as a libertarian, I don’t care about inequality so long as the poor and middle class are also sharing in the gains.” Or “as a progressive, I will continue to worry about inequality until the top 1% only has 1% of national income.”

I think these are the normal fallback answers. But really? Libertarians: if the income of the top 1% doubled in a decade, but the bottom 99% increased by 0.5%, you would be fine with this, because at least no one declined? Progressives: you would really support increasing taxes on the rich, despite any downside to this, until incomes were exactly equalized?

Frankly, I don’t believe anyone really holds either of those extreme positions. So surely, the data must matter? We want some reasonably shared benefits from economic growth, but no one really demands that they be exactly equal, right?

So, consider your own biases. Don’t engage in motivated reasoning. And think through how your views might change if you are wrong about the data. Perhaps someday Mother Nature will reveal herself, we’ll have the true inequality data, and we’ll see if we were honest about our reflections.