You can divide up the economy into two parts: money that goes to workers, in the form of wages and benefits, and money that goes to owners of capital, in the form of corporate dividends, bond payments, rent to landlords, etc.
This isn't the only reason inequality is growing in the United States. Inequality in wages themselves — between highly paid executives and low-income workers — has grown a lot, as has inequality in capital income. But given that capital income has always been less equally distributed than wage income, if more money is going toward capital, that's going to exacerbate the inequality problem. And according to the Economic Policy Institute's Lawrence Mishel, the declining labor share was the single most important factor driving inequality growth from 2000 to 2011.
A declining labor share of income also makes it possible for labor productivity to rise without average wages rising in turn. That is, it means that workers can get more productive without seeing the fruits of their labor. That's a big deal, both for workers and for economics as a discipline. For decades, most economists have treated the existence of a stable labor share of income as something like a law of nature. It was one of six "stylized facts" the famous economist Nicholas Kaldor proposed in 1957. Those facts serve as foundations for a staggering number of influential macroeconomic models, including the ones that the Federal Reserve and federal budget officials rely upon. If the labor share is falling, things that most economists would have treated as impossible ten years ago start to look very troubling.
So what's behind the decline seen above? We really don't know yet. There's not nearly enough research for economists to come to a consensus. But here are six possible explanations.
1) The rise of the machines
Perhaps the most popular theory is that computers and other information technology have reduced employers' need for labor. Reduced demand for labor then translated into a lower price for labor — that is, lower wages and compensation. University of Chicago's Loukas Karabarbounis and Brent Neiman estimate that "a decrease in the relative price of investment goods" accounts for about half the decline in the labor's share of income globally. "Investment goods" are items firms purchase to aid in production: things like printing presses, or trucks, or computers. Computers, in particular, have gotten much, much cheaper, meaning investment goods as a whole got cheaper.
To most people, this explanation sounds intuitive. Of course automation hurts workers whose jobs are automated away! But until very, very recently, economists didn't take the idea of technological unemployment — or technological wage stagnation — seriously. There was a good reason for that. In the 19th century, Luddites anticipated mass unemployment would result from the adoption of the automated loom. What happened instead was that people moved from manual textile work to more labor-intensive parts of the economy. Employment itself didn't fall. That historical precedent led most economists to assume a similar adjustment would occur with computers and the internet. But recent trends, like the falling labor share of income, have made a lot of people reconsider technology's potential to boost unemployment or cut wages.
Karabarbounis and Neiman's numbers show a labor share decline in most countries. "Of the 59 countries with at least 15 years of data between 1975 and 2012," they write. "42 exhibited downward trends in their labor shares." A key advantage of their theory is that it can account for an international pattern like that. Computers have gotten cheaper everywhere, so if that brings the labor share of income down, you'd expect it to fall everywhere, as Karabarbounis and Neiman conclude it is; theories that attribute the US fall to changes in American public policy, for example, have a hard time accounting for the similar falls in Germany, China, and Japan.
Another prominent theory holds that the movement of manufacturing, textile, and other labor abroad is behind much or most of the fall in US labor share. It's easy to see how this would work. If a shoe manufacturer in North Carolina lays off 1,000 workers there and hires 1,000 more in Indonesia for a tenth of the cost, the share of their spending going to labor is going to fall, and the share going to American labor is going to fall even more.
But this too flies in the face of traditional economic theory. Everyone would concede that certain people are harmed by increased international trade, but economists advocating for expanded trade always emphasized that workers in importing countries would benefit overall. If trade is bringing down the labor share, though, then it's a net negative for workers. It's only US capital-owners and workers abroad who gain.
Edinburgh's Michael Elsby, the San Francisco Fed's Bart Hobijn, and the NY Fed's Aysegül Sahin estimate that US businesses' increased exposure to imported goods accounted for "3.3 percentage points of the 3.9 percentage point decline in the U.S. payroll share over the past quarter century."
This explanation has a harder time accounting for the international nature of the labor share decline than other theories, but the authors argue that it's possible that offshoring decreases the labor share in both the countries losing jobs and those gaining them: "it is possible that offshored production processes that are labor-intensive by U.S. standards also are capital-intensive relative to existing production in China."
3) Declining bargaining power and deunionization
Workers with less bargaining power have less ability to demand raises, which can depress the labor share of income over time. This is compatible with both the technological and globalization hypotheses; after all, technological change that reduces demand for labor reduces bargaining power, as does a decline in demand for US labor in sectors engaged in offshoring. But policy factors — like deunionization, or an eroding minimum wage — can also weaken bargaining power.
The University of Haifa's Tali Kristal has articulated a version of this hypothesis. She finds that labor's share of income declined "only in core, unionized industries," concluding, "organized labor’s decline is the main factor that led to the decline in labor’s share" in the US. In turn, she argues that computers helped undermine organized labor. Unionized manufacturing jobs were automated out of existence, for one thing, but Kristal also claims computers strengthened management's influence in workforces by making it easier to monitor worker activity, and that the "skill polarization" caused by the computer revolution undermined worker solidarity.
This explanation echoes a strain of Marx-influenced Keynesian economics, advocated for most prominently by the late Joan Robinson, that emphasizes the relative power of firms and workers in labor negotiations, and argues that the class struggle, rather than a mere market, determines workers' wages. If Kristal's right, then the most important fact underlining the labor share decline is that the proletariat has started to lose the class struggle really badly.
It's a bit tricky to define "financialization," but loosely put, it encompasses a number of changes in recent decades that have led the financial sector to take on a much larger role in the economy than it once did: finance income grew disproportionately fast, household debt increased, non-finance companies increased their financial investments, etc.
This can erode labor's share of income in a few ways. For one thing, when companies have more kinds of capital to invest in, using income to buy up capital rather than on wages becomes attractive. Second, shareholders have become more powerful, putting pressure on firms to pay out dividends and do share buybacks rather than raise wages. Kingston University's Engelbert Stockhammer, in a paper for the International Labor Organization, found that financialization was the main factor in the decline of labor shares in both developed and developing countries, swamping competing factors like globalization and technical progress.
5) Statistical quirks
Finally, some analysts think the fall in labor share is a statistical fluke. Once you adjust for other factors it goes away, the theory goes. The Bureau of Economic Analysis's Benjamin Bridgman has argued that once you adjust for certain taxes paid by corporations and for the fact that capital depreciates. When a firm buys a computer, the value of that purchase erodes with time. A $2,000 computer bought in 2012 is not worth $2,000 in 2014. Once that's taken into account, Bridgman claims, labor share is "near its historical highs" rather than falling.
The Manhattan Institute's Scott Winship cites another few factors, such as the fact that capital's share of income includes proprietors’ income, or income from one's own business, which is perhaps better thought of as wage income.
6) The recession
Another possibility is that the decline is more recent than we assume, and the result of slack in the labor market is due in large part to the recession. This theory is bolstered by the fact that labor's share of income went up significantly in the late 1990s, when unemployment was at an all-time low. Jared Bernstein of the Center for Budget and Policy Priorities built a model predicting labor share based on unemployment, and found that this alone explained the decline, at least post-2007.
Unlike a lot of other potential causes of labor share decline, Bernstein's is very easy to reconcile with a traditional economic framework. You don't need to reject the idea that the labor market adapts to new technology, or conclude that trade is a net negative for workers, or endorse a class struggle model. But if a robust decline continues even after the economy gets to full employment, then more heterodox explanations will start to look more plausible.