Torquing the numbers doesn’t help

The Oxfam report on global wealth inequality did what so many similar reports in the past have done: it further reduced public understanding about the nature and extent of economic inequality. The factoid designed to make headlines — “eight men now own the same amount of wealth as the poorest half of the world” — did its job, but it told people very little about economic inequality.

When it comes to economic inequality, what really matters is economic welfare, or well-being. An economy in which everyone is equally happy is not necessarily one in which everyone is earning the same incomes or consuming the same combination of goods and services. People have different ways of finding happiness: economic equality is not the same thing as uniformity.

Of course, we don’t directly observe economic well-being; we have to infer it from available data. And the sort of data used by Oxfam is among the least useful for evaluating welfare.

The bulk of empirical work in economic inequality is done using income data, mainly because they are more readily available, and because the link between incomes and standards of living is fairly straightforward. The recent trends in income inequality have been discussed extensively elsewhere, so I’ll pass over them quickly. After a surge in the 1980s and 1990s, what I call “first-order” inequality — the sort that is reflected in a divergence between growth rates of median and average incomes — has remained relatively stable for the last 20 years.

“Top-end” inequality — the concentration of income among a high-earning elite — continued to increase until 2006 in Canada, and has declined since. But even after this decline, Canadians earners in the top one per cent of the income distribution still have a larger share of income than they did 30 years ago: just over 10 per cent. This makes for headlines less eyepopping than in the Oxfam press release, but it’s also a more realistic characterization of economic inequality.

The link between consumption expenditures and welfare is tighter than it is between income and welfare, but the obstacle here is data availability. Instead of tax files and census data, consumer spending data are largely surveybased. Available evidence suggests that the increase in inequality in consumption has increased slightly less than inequality in income.

The connection between the wealth measure used by Oxfam and economic welfare is tenuous, and it’s pretty much nonexistent for people with low incomes. It’s certainly the case that a large number of low-income households have little in the way of net worth, but this does not justify a conclusion to the effect that income redistribution programs have failed. If anything, the fact that low-income households have not saved enough to accumulate significantly positive net worth shows that existing income support programs are working as intended.

Assets generate revenue, and the reason people buy assets is to finance future consumption. For most people, the principle motivation for saving and accumulating assets is to finance spending during retirement. Even if everyone had exactly identical wage profiles throughout their lives, there would still be inequality in net worth: older people would have saved for longer and would have had the time to accumulate more.

But if low-income households can count on a public pension in retirement, there’s little reason for them to save. And since lowincome households don’t spend very much, there’s even less reason to force them to sacrifice some of their current spending in order to accumulate more net worth.

Of course, one way to increase net worth for low-income households is to replace public pensions with privately-managed social security accounts, but I’m pretty sure that’s not what most progressiveminded people have in mind when they deplore low net worth in the lower end of the income distribution.

So the fact that low-income households have little net worth is exactly the outcome that policymakers concerned with their welfare would want: they don’t save for retirement, because they don’t have to.

It’s also worth mentioning that what Oxfam and others who use these data call “wealth” is only a small subset of total wealth. Wealth generates income and human wealth — or human capital — generates employment income. Wages and salaries are a much larger share of total income than the interest income generated by financial and tangible assets.

In a 1997 study published by former Bank of Canada deputy governor Tiff Macklem, Canadian estimates for human wealth were found to be about four times larger then estimates for non-human wealth. Since the distribution of human wealth corresponds closely to employment income, the distribution of total wealth is much less concentrated than what the numbers for nonhuman wealth suggest.

I’m reminded of the egregiously mendacious attack ad run by Engage Canada, an astroturf group — hilariously-billed as “non-partisan” — that emerged to run anti-CPC ads just before the last election, and which quickly disappeared once the writ was dropped. Even though measures of income inequality had declined or remained constant while the Conservatives were in power, it ran ads saying that income inequality had “skyrocketed” under Harper’s government, using numbers that had been torqued beyond all meaning. The campaign succeeded on its own terms — the Conservatives lost the election — but it also degraded Canadians’ understanding of the issue. Progressives and civil society groups that claim to be concerned with income inequality should probably start toning down the volume, and stop overselling and distorting the evidence on economic inequality. It’s best for everyone concerned if excessive inequality is viewed a problem to be understood and solved, and not as a cudgel to be wielded in a partisan cause.