Wednesday, June 22, 2011

Of labour income and shifts in income inequality

This is based on a comment I made here.

There are a host of issues with inequality: of the significance of income versus consumption inequality, for example. But that income inequality in the US has been increasing in recent times, (on some indicators) back towards the levels of the “Gilded Age”, is fairly clear. The effect is particularly concentrated at the very top of incomes.

Paul Krugman divides the history of C20th income inequality in the US into the “long Gilded Age” until about 1938, “middle class America” of the Great Compression from then until the mid 1980s and the “Great Divergence” since. A pattern he has tied it to political polarisation and financial deregulation.

Krugman is arguing that it is basically a creation of public policy. That the New Deal deliberately compressed incomes until the “supply side” policies of the 1980s reversed the effects. I suspect that public policy has less effect than Krugman claims.

It is worth noting, for example, that the US federal income tax is wildly progressive. The top 5% of income taxpayers pay over half of total federal income tax receipts. The top 20% of income taxpayers pay over 80% of total federal income tax receipts. Even given that the income figures cited above are about market (i.e. pre-tax) incomes, the income transfers of the US government work in the reverse direction to that which Krugman is indicating. So, for the “public policy did it” explanation to work, it must be overwhelmed by other aspects of public policy. For example, Krugman might have a point about the finance industry.

Still, there could be other factors are at work driving up capital and driving down labour incomes (relative to each other).

It is easy to nominate social changes that could be increasing income inequality.
(1) Immigration: widening skill differentials and driving down labour income (and therefore putting upward pressure on capital income).
(2) Increased life cycle inequality with the expansion in higher education, as poor students become highly paid professionals later in life, and increasing skill differentiation.
(3) Women entering the workforce, putting downward pressure on labour income (and thus upward pressure on capital income) and increasing household differentiation as high-income women marry high-income men.
In the C19th, there were temperate and tropical zone labour flows. A vital interest of labour politics in North America and the Antipodes was to block tropical zone labour flows competing with them. (Hence the White Australia policy and anti-Chinese immigration sentiment in both regions.) But even within the temperate zone labour flows, a lot was fairly low skill. The effect of the mass immigration to the US was to increase skill differentials and put downward pressure on labour income (and upward pressure on capital income): hence the inequality of the "Gilded Age".

There was a dramatic drop off in migration in the interwar period. After some lag, there was consequently downward pressure on capital income and upward pressure on labour income. In the postwar period, there was high productivity growth and immigration was largely from Europe, and the more educated and industrialised parts thereof. Capital growth more than compensated for labour force growth, so there was upward pressure on labour incomes and downward pressure on capital incomes (relative to each other) with little or no widening of skill differentials. Hence the "great equalising".

Then productivity growth died away, migration became much more "tropical zone", women flooded into the workforce. There was downward pressure on labour income, widening skill differentials and upward pressure on capital income (due to increased relative scarcity). Combine that with increased life cycle inequality (due to mass higher education) and high-income women marrying high-income men and you get flat labour income and increased income inequality. (And yes, I am aware of the claims that the net effect of migration is to increase per capita incomes. Maybe, but it does not increase specifically labour income, which is my point.)

So how much public policy changes drove, or responded, to these underlying patterns, a fascinating question. But I am sceptical how much public policy can counteract these effects.

I also disagree with Krugman’s take on the wider the effect of Reagan. There is little point in comparing Reagan’s performance with the period of high productivity growth. The interesting comparison is with other developed democracies at the same time as his term in office. The "supply side" reforms that started under Carter led to the US pulling away from European countries in per capita income, after the post war decades of convergence in per capita income as European countries got over the effects of Dictator’s War and technology spread in “technological catch up”. Reagan did well under the comparison that counts—similar countries in the same general context.

A lot of factors affect both relative economic performance and trends in income inequality: analysis needs to carefully consider the complete context, rather than focusing on what seems congenial to blame or credit.


  1. The top 20% pay over 80% of tax receipts because they control 85% of the nation's wealth.

    Saying something as vapid as "But the top n% don't pay n% of the tax, it's so regressive and unfair!!!" implies that in your world a fair tax system would be one where Bill Gates pays the same tax as Manuela Cortez, cleaner at the Holiday Inn, Scottsdale AZ.

  2. So they roughly pay in proportion to their wealth. I was making no implication of unfairness, merely pointing out that the US federal income tax system was strongly progressive (higher incomes pay a high percentage), so the tax system (and tax cuts) is a poor candidate to explain any increase in income inequality. (The US federal income tax system seems to me to one of the better ones in the developed world:far too many bells and whistles, but that is normal.)

  3. This comment has been removed by a blog administrator.

  4. Lorenzo:

    "It is worth noting, for example, that the US federal income tax is wildly progressive."

    This is absolutely false in both historical and comparative terms. In 1944 the top income tax bracket in the US was 94%, in 1960 it was 89%, in 1970 it was 70% and now it is 35%.

    Presumably you are aware that most comparable western nations have significantly more progressive income tax brackets with much higher top bracket rates.

  5. Mel: you are making the elementary error of confusing tax rates with tax incidence. The rate of tax tells you very little about the incidence. In the UK, for example, the Thatcher Government's lowering the rates of income tax led to an increase in the amount and proportion of tax paid by the top income earners. The bottom 50% of US federal income tax payers pay a much lower percentage of total tax than does the bottom half of taxpayers in other Western countries.

  6. the contents are good to be used as reference for my thesis