In the popular imagination, the US economy is certainly rich in total. But most people who think about the matter probably assume the US partly secures those overall riches by tolerating much higher inequality than is normal elsewhere in the developed world (certainly than in Europe). They assume the US has much higher income and wealth inequality than is allowed in other developed countries, with that occurring partly because transfers to those lower down in the income distribution are much more restricted than those in Europe and partly because those at the very top of the income distribution are allowed to pay lower taxes than is permitted in other countries — indeed, those at the very top paying lower tax rates than those in the middle. They probably also think that inequality in the US is rising over time.
Some people (e.g. myself) might say inequality is of no moral importance. Others may even see the US as an example to follow — if one wants a prosperous society one ought not to try hard to limit inequality. Others may feel high American inequality is a moral and social indictment of capitalism. But whichever of these responses one picks, at least there is agreement that US economic inequality is high and rising.
However, in their book, The Myth of American Inequality, Phil Gramm, Robert Ekelund, and John Early argue that this picture is wrong in almost every respect. Crucially, the vast majority of their case is not about alternative judgements or complex additional factors the standard analysis does not take into consideration. Rather, they argue that the picture is wrong because obvious and in many cases officially acknowledged statistical or methodological errors are not reflected in the relevant official statistics that academic or popular analyses of US inequality rely upon.
A case in point is the income of the poor. US official income statistics simply exclude a large portion of the monies those on lower incomes receive as government transfers. Monies given to those on lower income in the form of food stamps, Medicare, Medicaid, and the reimbursable Earned Income Tax Credit and Child Tax Credit are not counted. That means that only 32 per cent of transfers appear as income. For poor families only 12 per cent of transfers are counted.
Adjusting for this, the authors argue that the top 20% of households have an income only four times that of the bottom 20% (not 16.7 times as per the official statistics). This ratio has declined over time (not risen) – e.g. it was 5.6 to 1 in 1947. Instead of 12.3% of the population meeting the official US poverty definition, only 2.5% do. Perhaps most strikingly, once the effects of taxes are included, the authors argue that the bottom 60% of the population all have essentially the same income.
The authors frequently appeal to independent ways to think about the factors they are considering. For example, when thinking about how poor low-income US households really are they note that the average ‘poor’ American family lives in a house larger than the average middle-income French, German or British family do.
Some of the factors they highlight are extraordinary. The US census-based measure of inequality changed its basis between 1992 and 1993 by increasing the maximum level of income considered (from just under $250k to just under $1m), then in 1999 removed it. That factor alone greatly increased the aggregate measured income at the top even though it involved no actual increase in inequality. Similarly, in 2013 additional categories of income were asked about, again boosting the total without any actual change in inequality. Yet there was no adjustment to the back-series, so official measures of inequality proceed as if the extra income ‘discovered’ by these methodological changes were actual extra income, creating a completely artificial impression of increasing inequality driven by rapid rises at the top.
These two factors alone boosted official statistics on a rise in inequality by 42.1%. The authors argue that income inequality after taxes and transfers actually fell. They find that as of 2017 inequality in disposable income is less in the United States than in the UK or Japan, only slightly above Australia and Canada, and slightly more above that in France and Germany.
Even in terms of taxes, the authors argue that the US imposes the most ‘progressive’ top-end taxes amongst developed countries. The top 10% of US households each 33.5% of income and pay 45.1% of taxes – a ratio of 1.35. The ratio in Germany is 1.07 and in France 1.1.
Tax changes also changed the definition of personal income versus company income, because a 1986 law tax-favoured income over company profits. This led to wholesale corporate structure change, with a flourishing of partnerships and other corporate forms tending to make income more personal. This again boosted measured income at the top without changing actual income.
Academic studies and books such as those of Thomas Piketty that claim the very rich pay lower taxes than average income or low income Americans embody the above errors, but perhaps even more significantly they don’t measure incomes at all. Instead they take the appreciation of asset wealth of the rich and treat it, quite erroneously, as if it were income. Think about that. If you do a degree then your human capital increases — you have some extra skills that you could one day turn into money. But you haven’t done so yet. Your degree cost you money. You didn’t make money by studying because your (human capital) assets had appreciated! If and when the rich spend the increased value attributable to their assets they will pay taxes on that.
Earned income inequality in the US has indeed risen, but that is heavily driven by a large rise in another form of inequality: working hours inequality. The average working hours of the bottom 20% of US households have risen by much less than those in higher quartiles. The authors argue that the reason is that large transfer payments, along with increased ease of receiving them and of receiving them indefinitely (rather than using them for a short period before becoming self-supporting), have reduced the incentives for those at the bottom end of the income distribution to work. Other factors explaining increasing earned income inequality include education and female labour force participation. As the authors put it: ‘Not surprisingly, increased equality of opportunity and the attendant expansion of effort to succeed often generates more earned-income inequality.’
This is not a book for the faint-hearted. It contains extensive statistics and involved reasoning. Some of the points it makes are quite well-known — e.g. the incomplete reflection of increased welfare in income statistics because the value to consumers (what economists call ‘consumer surplus’) of consumer products has greatly increased over time; or the fact that some inflation measures don’t take account of the way consumers can substitute away from products that rise more rapidly in price to slower price-rising products. Some points are contentious — e.g. it is not so obviously a mistake for the benefits paid to those on low incomes to rise in line with inflation metrics that do not assume substitution, because those consuming only necessities may not easily be able to substitute for them.
Overall, though, this is a book that may make you think differently about the inequality (or otherwise) of the American economy — which is exactly what the authors hoped to achieve.
‘The Myth of American Inequality: How Government Biases Policy Debate’ by Phil Gramm, Robert Ekelund and John Early was published with a new preface in 2024 by Lexington Books / Bloomsbury (ISBN: 978-1-538-19013-5). 274pp.
Andrew Lilico is an economist and a writer on politics and philosophy. As managing director of an economics consultancy firm he works extensively with governments and bodies such as the European Commission, especially on technical matters relating to price caps for utilities, auction design and economic analysis of emissions trading systems, and on competition cases (especially cartels). His doctorate is in Game Theory.