Income inequality measures leave a lot to be desired.
First off, inequality measures based on pre-tax income get used to argue for ever-greater redistribution. It’s a bit of a nonsense because the tax-and-transfer system is already highly redistributive. If after-tax-and-transfer income inequality worsened substantially, then there could be something to talk about. But before-tax-and-transfer figures, used in policy arguments about how much redistribution should be undertaken, are really misleading – and often deliberately so.
But even if we use after-tax-and-transfer income measures, we still have a problem. People who are relatively asset rich can draw down from savings to smooth consumption from one period to the next if they have volatile earnings. Think of the contractor who does really well one year then decides to take a lot of holiday the next year because works is slow anyway. An inequality measure based on income would not tell us as much about real inequality. So a measure of consumption inequality probably tells us more about real experienced differences in the variable people really care about.
And so we come to some excellent new work by Chris Ball and John Creedy at Treasury. They recalibrate inequality measures for income and consumption. And here are the main results.
The dashed “Market” line traces Gini inequality in market earnings over the period – that’s before taxes and transfers. That series rose from the late 1980s through about 1994, then levelled off before easing back to early 1990s levels.
The solid “Disposable” line tracks Gini inequality in disposable incomes – that’s after tax and transfer. This measure rose from 1988 through to about 1994 then was basically flat. Note that the spike at 2001, and again around 2010, coincide with tax changes that encouraged income shifting from one year to another, generating the hump.
The dashed “Consumption” line is the one that’s particularly interesting. It measures inequality in real consumption. That measure rose a bit from the late 80s, plateaued through the mid-90s, and has eased off since then. Current inequality in consumption is lower than it was before the 80s reforms.
I also find it really interesting that pre- and post- tax-and-transfer income measures track very similarly after Working for Families.
And now, for the most fun of all, let’s combine all of this with Bryce Edwards’s tabulation of New Zealand Herald stories about inequality from the 2000s onwards. If I had Ball and Creedy’s underlying data, I would overlay the two; you’ll have to use your imagination.
Not for the first time, I wonder whether our media, and pundits, have confused their monsterometer with their frog-exaggerator. None of the inequality measures Ball and Creedy produce above are currently higher than they were in 2001. But crises sell papers and policies, regardless of whether there’s any real basis for them.