Understanding the US wage gap

News coverage and analysis of pay inequality often focus on the difference between how much the boss makes and the salaries earned by the rank and file.

But Stanford economist Nicholas Bloom says the widening gap between highly paid executives and everyone else in the firm is not the biggest factor behind the rise in U.S. earnings inequality. An increasing spread in what different companies pay their employees has a lot more to do with why some workers are making so much more than others.

In other words, the bulk of the rise in inequality is being driven by some firms paying much higher wages than other firms — rather than by widening gaps between high- and low-paid workers within each company.

This means that focusing on CEO pay, which leads to policies like requiring some firms to publicly disclose the ratio of what the CEO makes with respect to the median employee, might be less effective in dealing with inequality than understanding why some firms are paying better than others.

Using data from the U.S. Social Security Administration based on W-2 forms filed between 1978 and 2012, Bloom – a senior fellow at the Stanford Institute for Economic Policy Research – found two very different patterns. A relatively small number of people, the richest 0.2 percent of earners, made dramatically more money than everyone else within the same firm.

"Particularly in firms with more than 10,000 employees (there are about 1,000 of them) the CEO has seen pay rise by an astounding 140 percent while the median worker has seen real pay fall by 5 percent over the last 35 years, so a huge gap has opened up," says Bloom.

But the focus on CEO pay obscures a more widespread trend. The distribution of wages within firms has remained fairly constant for the majority of workers in the U.S. The gap between what workers earn and the average wage at their firm has not changed much over the past 30 years.

Compare, for instance, the median worker in 1982 to the median worker in 2012. With 50 percent of full-time workers making less and the other 50 percent making more, the individual at the 50th percentile in 1982 was making about $27,700 per year at a company that paid about $33,600 on average. Three decades later the workers and firms are almost certainly different but the picture had not changed much: the worker at the 50th percentile in 2012 was making about 20 percent more in real terms – $33,200 – but still less than the $42,200 average wage at his firm.

Where you see the real increase in inequality is when you contrast what happened in the middle with the much more dramatic gains for workers higher up in the income distribution. The workers at the 99th percentile saw a much steeper increase of 67 percent in their earnings: going from $156,000 in 1982 to $260,000 in 2012. But, again, the rise in earnings for these top-earning employees happened in lockstep with the rise in average pay at their companies: The average wages at the companies that employed them were 87 percent higher in 2012 than in 1982, going from $56,000 to $105,000.

This means that the growing spread in wages paid by different firms does a better job accounting for the growth in inequality than changes in the pay structure within each of the firms for the bottom 99 percent of workers.

Bloom’s findings are in a working paper he co-authored with Jae Song, of the U.S. Social Security Administration; David Price, of Stanford; Fatih Guvenen, of the University of Minnesota and NBER; and Till von Wachter of the UCLA.

There are several possible factors contributing to the growth of pay inequality among firms. A trend toward outsourcing positions such as reception, janitorial, security and catering crews contributes to separating high- and low-wage workers into different firms and increasing differences in their pay.

But pay inequality has also increased between firms in the same industries. It could be that some firms are more productive and can pay higher wages. If you had two skilled engineers and one went to work for Apple while the other was hired by Motorola, it is likely that after 20 years the Apple engineer would be making much more.

But, given the high mobility in the U.S. labor market, it would also be likely that the skilled engineer at Motorola would have sought to switch to Apple way before then.

"It seems that a large part of what is driving this is basically a sorting effect where top individuals tend to go together into the same firm," says David Price a Stanford doctoral student and one of the paper’s co-authors.

So, for instance, higher-paid investment bankers would sort into the better paying investment banks as workers who are more skilled within their respective educational groups are increasingly more likely to work in the same firms.

Moreover, this seems to be a global issue. New research is finding growing pay inequality between firms in economies as diverse as Germany, Sweden, the United Kingdom and Brazil. The fact that it is happening in such different settings likely indicates that the main forces behind it are likely to be global in nature, like the growth of technology and globalization, rather than institutional factors such as unions, minimum wage laws, or social norms that are particular to each country.

Greater research is needed to determine how much of a role each of these factors is playing. What is clear though, is that changes in the nature and composition of firms have played a part in the rise in earnings inequality and that we need to take a closer look at the role of firms.

Miriam Wasserman is a freelance writer.