Thursday 16 October 2014

Inequality narrative-buster

It's been pretty clear that income inequality in New Zealand has been stagnant for at least the last decade or two. It rose in the late 80s and early 90s, but nothing much has happened since then.

And whenever I point it out on Twitter, against assertions from others that income inequality is worsening, I get replies that the problem isn't income inequality but wealth inequality. I hadn't seen great data on wealth inequality, so I couldn't much comment. But we have some data on wealth growth and median wealth growth now from Credit Suisse. Everyone should read the report. Here are the important bits for the New Zealand debates.

First, New Zealand is far wealthier now than it was in 2000. Per-adult wealth is up by over 300% (a 100% increase is a doubling, a 200% increase is a tripling, a 300% increase is a quadrupling) over the period, when we evaluate wealth at current exchange rates: a high dollar means we are wealthier, no matter how much some groups want to make us poorer by devaluing the dollar. When we use constant exchange rates instead, wealth has more than doubled (the yellowish line).

Per-adult measures, which I interpret as meaning an average, can be a problem though: what if all of that is because of massive gains at the top? We need medians. Medians are below.

This is the one using current exchange rates rather than constant ones, so growth is overstated relative to constant exchange rates. Note that here the base year is 100, so our being nearly at 500 means a near quintupling of median wealth. Now it would be lower were we using constant dollars, but we can compare the current dollar mean rise with the current dollar median rise. If gains were accruing at the top, we'd see larger increases in means than in medians. If they were accruing at the bottom, we'd see bigger gains in medians than in means. And it looks rather like larger gains in medians. 

This is then confirmed when we get out to Table 1, which lists New Zealand as a "Medium wealth inequality" country, along with Australia, Canada, Finland, France, Greece, Ireland, Italy, the Netherlands, Portugal, Singapore, Spain and the UK. I note that Denmark, Norway and Sweden rank as having higher wealth inequality than New Zealand.

The report then discusses trends in wealth shares of the top decile, by country, from 2000 through 2014. Their data doesn't go back to the 80s and 90s, unfortunately. But over this period, New Zealand ranks as one of the countries experiencing a significant reduction in wealth inequality over the period. Their measure is the share of national wealth held by the top 10%. In 2000, the top 10% owned 62.3% of the wealth in New Zealand. This placed New Zealand as 19th out of 46: the top 10% had a smaller share of national wealth than was seen in the median country in the survey (19th lowest). 

In 2007, the top decile owned 61.2% of the wealth; wealth inequality dropped slightly from 2000 to 2007. In that year, New Zealand had the 22nd lowest wealth inequality - about the median. In 2014, the top 10% owned only 57% of the wealth in New Zealand: a substantial fall. In 2014, we had the 12th lowest wealth inequality. 

So, according to the Credit Suisse report:
  • Wealth inequality in New Zealand is lower than in most other countries in their survey;
  • Wealth inequality in New Zealand has been falling from 2000 through 2014, with the biggest drop concentrated in the period from 2007 to 2014;
  • Low wealth inequality isn't necessarily a great thing: while Australia, Belgium, and the Netherlands had less wealth inequality than did New Zealand in 2014, so too did Spain, Greece and Italy. 
Please update your narratives accordingly. 


  1. Nice post. What about human capital, which is 70%+ of all wealth. Gibson and Li estmated nz human capital at $800b

  2. I'd be a bit more cautious, personally.

    First, when you claim "nothing much has happened" with income inequality since the early 90s, you don't just get people on Twitter saying we should talk about wealth. You also get me pointing out that, firstly, that's not true: income inequality continued to rise until the early 2000s, whereupon it fell slightly and thanks only to deliberate actions like Working for Families, so plenty "has happened", in both directions; and, secondly, to concentrate your narrative on what happened *after* a period in NZ that represented the developed world's biggest increase in income gaps gives a very partial picture, to say the least, of how the country's changed.

    Second, the Credit Suisse report seems to be based on Reserve Bank household wealth data. I've never seen anyone use that to estimate wealth shares before, and it's not clear from the report (or accompanying databook) how they've arrived at their estimates. Nor do their figures match up with what little we know about wealth inequality from other sources. So I certainly won't be "updating my narrative" until I've had a careful, rigorous look around the data - as I'd expect others to do.

  3. Agree that it's worth checking the data sources on the new report.

    I've focused on the post 80s changes because much of the current inequality narrative has presented the increase from the 80s to present as an ongoing trend rather than a sharp rise followed by a flatlining. A lot of people seemed to be deliberately mischaracterising the data to make the case that inequality was worsening, rather than that it did worsen a couple of decades ago. Again, I will quote from the MSD report: "Overall, there is no evidence of any sustained rise or fall in inequality in the last two decades."

  4. Might stencil this URL to my forehead and walk around in public with it for a while.

    (If I had a point otherwise it would be to say, that income/wealth equality per se isn't a problem when living standards are rising for all.)

  5. And rising housing costs are in large part due to over-regulation limiting land supply, and increasing cost of build. (For holiday house I've just paid consent work of $20,000 for a poxy renovation. When I got to $700 consent to simply replace 'existing' log burner, decided the old one will do well enough.)

    So freeing up land by deregulation, and returning us to capitalism the best solution for that Max.

  6. Jim has this right, I think. Looking at inequality in finanical wealth but ignoring the capital that generates most of income (i.e. human capital), makes as much sense as looking at inequality in, say, cash balances, and ignoring bonds and shares.

    Also, I would guess that the financial wealth stocks include housing, and much of the changes in the distribution of wealth reflect house price inflation and the differences in the proportion of households' portfolios that come from owner-occupied houses. The trouble is that, while financial assets return a monetary income that can be spent on anything, housing returns a flow of in-kind services. Equivalently, owning a house returns a financial flow that has to be spent on housing, and so different portfolios have to be deflated by different price indices that reflect the relative role of housing in the portfolio.

    Best to not even suggest that wealth inequality is relevant at all. Anyone got any good data on consumption inequality?

  7. That's BHC and AHC income measures; query was on wealth.