What Is Happening At The Top Of The Income Distribution?

The increase of the share of those on top incomes has not been caused by market forces but is the result of their more favourable taxation regimes they have experienced since the early 1990s. 

Policy Quarterly has just published papers from a symposium on distributional inequality held last June. There are really interesting papers by Geoff Bertram, Phillip Morrison, Bill Rosenberg and Simon Chapple et al which you may want to read for yourself. This summarises my paper on pre-tax top personal incomes, and a follow-up paper on after-tax personal incomes which presents new research.

It is a tricky area because definitions are important and are frequently confused. I looked at the top 10 percent, 1 percent and 0.1 percent of all adults. (Not taxpayers because tax law affects who reports tax. In any case we want to include those who have zero incomes, especially as the proportions changed in the 1970s when more women entered the paid-workforce.) Unfortunately, only the top of the distribution is well recorded by the Department of Inland Revenue whose data I used. I’ve not used the aggregate income the IRD reports because, again, of the poor recording of those at the bottom in the past. Instead I have used the aggregates from the National Accounts. Sorry for this boring paragraph, but you would be surprised how many commentators skip these things and come to conclusions at odds with the evidence. Some don’t even distinguish personal and household incomes or income and wealth.

I can trace the share of before-tax personal income (which almost exactly corresponds to market incomes) of those at the top back to 1937. Before then the IRD data is contaminated by the inclusion of company incomes. The pattern it shows is that the share of those at the top broadly decreased from 1937 to the mid-1980s. That is similar to the conclusion I found in my 1983 book Income Distribution in New Zealand. At the time I was surprised because I began the study with the popular hypothesis that the income distribution in a capitalist economy tends to become more unequal. The evidence from the 1950s to the 1970s was that it did not always.

My response was similar to Keynes’ ‘when my information changes, I alter my conclusions. What do you do, sir?’ So the book discusses how we might explain the falling inequality of before-tax incomes over the period. I suggested the two main drivers were low unemployment and women entering the paid workforce.

From the mid-1980s the shares of before-tax (i.e. market) incomes of those at the top is broadly stable. There are year-to-year variations, and some will see a trend which others will say are not there or are too small to be worth worrying about. Given my 1983 study one might draw attention to higher unemployment and that there were not as many women entering the paid workforce flattening out the previous trend. As a broad generalisation we may collude that the forces driving the post-war reduction in inequality ended so that market shares have been broadly constant in recent years.

The same pattern is not evident in the after-tax data.Unfortunately the data goes back only to 1971. Before then there was a social security tax on incomes which the IRD data does not record.

The story of the disposable income shares is different. There is a slight fall in the share of those on top incomes from the early1970s to the mid-980s followed by a huge jump in their share between 1988 and 1992. (The size depends on which group you are talking about, but the higher the incomes the greater the proportional increase; the share of the top 0.1 percent increased to more than 2.5 times.) After 1992 the distributional shares stabilise again (subject to various caveats including on measurement and tax avoidance; read them in the paper).

Since after-tax (or disposable) incomes are before-tax incomes with income tax deducted and social security transfers added, the change that took place between 1988 and 1992 has to be the result of income tax cuts, because there was no change in before-tax incomes. (Except for New Zealand Superannuation, there are no social security transfers to those at the top.) The key changes were the major reduction in top income tax rates and the change in the way that company incomes were taxed which was very favourable to those who received dividends – mainly those in top income brackets. It follows that the big increases in the gains at the top – a major contributor to the increase in income inequality – was that taxes on them were reduced.

We were promised that reduced taxes on the rich would result in them investing more (from their higher incomes) and innovating more (because lower taxes would be less penalising to risk). The idea was that this would result in faster economic growth, and that while they would benefit most (and therefore their income share would grow faster) there would be a ‘trickle-down’ effect so that (just about) everyone would benefit.

There is no evidence this happened. Once the Rogernomics Recession came to an end, the economy grew at much the same rate as it had in the past. There is no evidence that the Rogernomic measures lifted the long-term growth rate.

What happened? My guess is that the rich did not save much of their extra income but spent it, and that they were not particularly innovative or that, possibly, in total they made a lot of bad investments. The second paper goes on to discuss how the higher income share at the top led to an increase in conspicuous consumption by the wealthy and of the power of those at the top.

You may be surprised that New Zealand’s rich do not appear to follow the international pattern which Thomas Piketty found of their increasing income share in recent decades. I suggested two reasons. First, we do not have the hyper-financial sector which has driven the Piketty identified changes. Second, our rich are hardly so, compared to the rich that Picketty is tracking. Were they so rich, the New Zealand market would be too small for them, and they would diversify offshore, at which point they could live only half a year in New Zealand (at most) and report only their New Zealand income for tax purposes. The three richest ‘New Zealanders’ in the Fortune Rich List all live offshore.

That is where the empirical research ends. Except I also detail the measurement difficulties. As best as I can judge, none of them, if totally resolved, would markedly change the conclusion about trends. Opinion is cheap, but if anyone can do the hard empirical research which would modify my conclusions, I would be delighted to acknowledge it and incorporate into my analysis.