Top Incomes in the Long Run of History

"What’s missing from the Goldin and Katz book is that they really don’t look at all at what’s going on in the top 10%. That’s a very important part of the story, and one that requires more attention. Some very interesting things have been going on in the top 10%, and especially the top 1%. Atkinson, Piketty and Saez have really been pioneers in this and this article is an overview of much of their research. What the paper shows is that concurrent with the increase in the college premium and inequality between a median worker and a worker at the bottom, there has been an even sharper increase in the share of the top 10% and top 1% in national income in the US, Canada, UK and so on. For me, it’s more relevant for the figures, to know what’s been going on. Occupy Wall Street has brought the 1% to the attention of the wider public, but it was Atkinson, Piketty and Saez who brought it to the attention of the academic community over the last decade. But it’s definitely worth reading their explanations, because some of the things they say are important. For example, they emphasise that it’s very difficult to account for these figures with the standard labour supply, labour demand explanation that Goldin and Katz emphasise. That’s not going to work, and we really have to think about things like social policies, progressive taxation and the politics of it. Yes, this pattern is very remarkable and the Anglo-Saxons have led it. You also see it in many other countries, including France, Italy, Spain and so on, but it’s not as pronounced as in the Anglo-Saxons. It is also important to note that this U-shape applies to the picture that emerges from non-top inequality as well, such as 90th percentile vs 10th percentile or the college premium. For example, Goldin and Katz’s book shows that the college premium was higher in the early 1900s than it was in the 1940s and 1950s. Then it remains stable for several decades, and then it starts increasing again in the 1980s. They use the data from the tax records, so they have data on the top 10%, the top 1%, the top 0.1% and so on. The two they choose to focus on are the top 10% and, especially, the top 1%. Sometimes they even give numbers for the top 0.1%. There you really see striking differences between countries. If you look at the 0.1%, it accounts for almost 8% of national income in the US, when in many countries it’s more like 1% or 2%. It’s also increased dramatically in the US – in the 1950s it was about 3%. I personally don’t think that’s the main thing, though it certainly played a role. It played a role for capital income. When you look at the top 0.1%, many of them are capital earners. So if you tax capital heavily, then the rich are not going to have as much capital left and capital income is not going to be as unequally distributed. There is a very mechanical effect from taxation there. But there are two other, more subtle, effects from taxation. One is that more progressive taxation – higher taxes at the top – may discourage people from working very hard and putting in effort. That will reduce their earnings and thus inequality. That may be inefficient, but it’s one of the things that happen when you have high taxes. Secondly, it might change the way in which people bargain with their companies and engage in “rent-seeking” activities in order to increase their pay or their bonuses. In the extreme – and I don’t think this contributes a lot, but just to illustrate – if top incomes were taxed at 99%, then no CEO would be tempted to do semi-illegal things in order to increase his pay, because there would be nothing to gain from doing so. If the top tax rate is 30%, on the other hand, and CEOs get pay from options, they may be tempted to do things like the Enron CEO, Kenneth Lay, did, because they get a lot of money in return. So while high tax rates at the top may inefficiently reduce these people’s labour supply, it may also reduce their rent-seeking activities."