It would be easy to report my latest findings on the income distribution with grandiose hysteria. But I am a social statistician and readers deserve a sober assessment. The conclusion that high incomes are rising faster than the rest is powerful enough on its own.

Dealing with the long-run income distribution is difficult. Often the data is incomplete and subject to statistical quirks. Here is my summary.

Census data suggests that, broadly, personal income inequality was stable from 1926 – when households incomes were first asked – to about the 1960s. From the 1960s to the 1980s there was some reduction in inequality mainly, I think, because mothers began to get paid jobs.

After that we can switch to data from the household survey, which means we can adjust for household composition (the number of adults and children) and the impact of taxes and social security transfers. Not every year was surveyed until 2007. This series is the one which most people use when discussing income inequality.

To simplify, I focus on the top 10 percent of people – ‘the top decile’. That is where the really big changes are. (The survey sample is not big enough to enable us to track the top 1 percent.)

The evidence from the data series is that inequality was stable in the early to mid-80s. But between 1988 and 1993 there was a huge increase in inequality with the share of the top decile increasing from about 20 percent of total income to about 25 percent; their incomes leaped a quarter relative compared to the average..

During the period of the neoliberal ascendancy, the government gave substantial tax cuts to those at the top which were paid for by (surreptitious) tax hikes on those on middle incomes and benefit cuts to those lower down (together with cuts to government expenditure which tend to favour those nearer the bottom). I have been unable to find evidence that the changes were caused by the market income distribution – that is, wages, salaries, interest and dividends and the like. (I did pick up an effect from cyclical unemployment.)

For almost fifteen years after about 1995 the income distribution seems stable. But from 2007 a sample of households was surveyed every year (hooray!). About that time, perhaps a few years later – turning points are always difficult to pick up – the trend share of the top decile starts rising again.

Compared to the period of the neoliberal ascendancy the income increase is not great – about 8.5 percent relative to the average over 11 years compared to 25 percent over the 5 years of the neoliberal ascendancy.

Had inequality remained at the same level it had been a decade earlier, a beneficiary with a child would have had about an extra $24 a week in hand, and the minimum wage would have been, roughly, $1 an hour higher.

Why? Perhaps it is too soon to tell; the data base is fragmentary. Some other work suggests the remuneration of those at the top has been increasing faster than other incomes. One might guess that this reflects their superior productivity growth. Perhaps our top salaries are being pushed up by international trends. Cynics point out that top rates of remuneration are not set by the market but that they set them themselves. Tax changes may not have had a major impact.

Because the changes are small, trends in other deciles are difficult to identify. Not surprisingly, the second to top decile is only slightly losing its relativity. On the other hand it has been clear for some time that the share of those at the bottom has been falling, presumably because benefits were indexed only to prices not to wages (like New Zealand Superannuation is). The rising tide has not been lifting all boats.

The period of rising inequality was under the Key-English Government. It is too early to tell the trend under the Ardern-Peters Government. They have made some changes such as indexing benefits to wages rather than just prices and their 2017 tax package favoured those nearer the bottom; minimum wages have been increased. We shall have to wait until after the 2020 election for the data to tell us with this government has changed the trend.

In the interim, it looks pretty certain that household income inequality has been increasing over the last decade.