EVERYONE knows that economic inequality has increased dramatically since the 1970s, and this has created a new cottage industry: dissecting "the top one per cent". We now have a study from three economists that broadens what we know about these top earners. The study's biggest news: economic inequality is becoming more gender neutral.
Go back to the early 1980s, and almost no women were in the top one per cent. Now there are lots, says the study, though women's representation still remains well below their overall share in the workforce, about half in 2012.
The study was conducted by Fatih Guvenen of the University of Minnesota, Greg Kaplan of Princeton University and Jae Song of the Social Security Administration.
The study divides the top one per cent of earners into two parts: the top 0.1 per cent (meaning that 99.9 per cent of earners are lower) and the second 0.9 per cent. From 1981 to 1985, women averaged about 2 per cent of the top 0.1 per cent; now, they're about 11 per cent. Among the second 0.9 per cent of earners, women have gone from 3 per cent in the early 1980s to 17 per cent now. Gender representation is still wildly skewed - but less so.
The odds of women staying in the top one per cent have also improved. In the 1980s, women who made it to the top "were much more likely than men to drop out" within a year, the study says. The economists call this a "paper floor" through which women fell, implying that their high earnings reflected one-time bonuses or special circumstances.
By contrast, women now in the top one per cent are about as likely to stay there as men.
The study doesn't explain these trends but speculates that the continuing gender gap may reflect "career interruptions", presumably for child-rearing. The study also rejects the theory that women's earnings increased because their jobs clustered in a few high-paying industries. In reality, top earning women are spread across many industries. The main cause of their higher earnings is probably that women's careers have become more like men's.
Although the study focuses on gender, it also unearths other intriguing facts about the top one per cent. Consider:
- It's a lot easier to get into the top one per cent than into the top 0.1 per cent. In 2012, the annual earnings threshold for the top one per cent - the minimum amount to qualify - was US$291,000. By contrast, the threshold for the top 0.1 per cent (what divides it from the other 99.9 per cent of earners) was US$1,018,000.
- Adjusted for inflation, earnings of the top one per cent have stagnated since 2000 after two decades of big increases. From 1981 to 2000, average earnings of the top 0.1 per cent quadrupled to US$2.841 million; in 2012, the average was lower, US$2.368 million. For the next 0.9 per cent, the average almost doubled from 1981 to 2000 to US$451,944. In 2012, average income was nearly the same, US$455,172. (All figures are adjusted for inflation and are expressed in "2012 dollars".)
- The finance industry (bankers, traders and brokers) has replaced healthcare (doctors) as the largest source of earners in the top one per cent. High-level consultants and accountants have also surged. In 2012, 29 per cent of earners in the top 0.1 per cent came from finance and another 10 per cent from consulting and accounting. Healthcare - which was almost 20 per cent in 1981 - was only 7 per cent in 2012.
To put this study in perspective, remember two things.
First, we think of the top one per cent as a small, exclusive group, but it actually includes lots of people. Payroll employment is now almost 140 million. One per cent of that is 1.4 million workers.
Second, this study doesn't encompass all aspects of inequality. It relied on social security wage records, with individuals' names blocked. This means it captured all wage and salary income, but not capital gains (profits from the sale of assets) or income from interest, dividends and privately held businesses. These would undoubtedly alter the picture somewhat, but not enough to change the study's main message: This is not your father's inequality.