The graph below, based on the work of economists Gabriel Zucman, Thomas Piketty and Emmanuel Saez, has been receiving considerable attention since it appeared in the New York Times last week. It shows the rate of annual income growth for adults at each percentile in the income distribution — from those who have the lowest incomes to those who have the highest incomes — over two time periods: the mid-1940s to 1980 and 1980 to 2014. Over the first period, post-tax income growth was fastest at the bottom, about 2 percent per year for the “middle class” (the 40th to the 80th percentiles), and a little slower among the wealthy.

The growth pattern over the second 34-year period looks very different: The richer you were, the faster you got ahead. Incomes grew less than 1 percent for the bottom 50 percent and less than 2 percent for the next 45 percent. They then took off for the richest Americans, with the growth rate for the richest adults ending up about six times that of those in the middle.

The chart is a clear, intuitive way to show the increase in income inequality over the past few decades, and an important reminder that growth for the rich cannot be expected to trickle down to everyone else.

But it doesn’t show why inequality has grown. What explains this portentous change, one that has had profound effects on our society, our living standards, and our politics?

In fact, there are many perps, each of which is captured in the “Inequality’s Causes” slide below. They do, however, share a theme: Many of the factors that enforced a more equitable distribution of growth in the earlier period have been eroded. Moreover, that erosion is neither an accident nor the benign outcome of natural economic evolution. It is often the result of policies that have reduced workers’ bargaining power and supported the upward redistribution of growth.

Starting from the upper-left-hand corner of the slide, the first two icons stand for many economists’ most common explanations: globalization and technology. Although increased trade has delivered benefits as its expanded supply chains have lowered prices, it has also lowered wages for workers in trade-exposed industries, a phenomenon that has been exacerbated by persistent trade imbalances. Between 1946 (the year Zucman, Piketty and Saez’s new data start) and 1980, the United States averaged a trade surplus of 0.5 percent as a share of GDP; since then, we’ve averaged a 2.7 percent of GDP trade deficit.

The technology explanation hinges on the premium earned by college-educated relative to high-school-educated workers, which stands at historically high levels. The idea is that employers’ skill demands related to computerization have increased the return to education, especially in the 1980s. Yet the college wage premium hasn’t gone up much in recent years, meaning tech-driven skill demands can’t explain the changes shown in the top figure. Something else is going on when low-wage workers are more educated than they’ve ever been, and, as Paul Krugman likes to note, there’s a big distance between the paychecks of schoolteachers and investment bankers, despite both groups having college degrees.

Although the next graphs show only correlations, they’re illustrative of that critical “something else”: the link between diminished bargaining power for most Americans and their diminished slice of the economic pie.

Note the phrase “full employment” on the man’s placard in the “Inequality’s Causes” slide. In very tight labor markets, workers with otherwise low bargaining power have a much better chance of claiming their fair share of the growth they’re helping to produce. Yet as the graph below shows, we’ve spent increasingly less time at full employment since 1980, a trend that quite closely tracks the declining median income growth that Zucman, Piketty and Saez have documented.

There’s also a racial justice component to this explanation. This marcher is from the 1963 “March on Washington for Jobs and Freedom” organized by the Rev. Martin Luther King Jr. The civil rights leader recognized this critical bargaining power role of full employment in pushing back on racism in the workplace, and, in fact, research confirms that tight labor markets reduce (although do not come close to eradicating) racial disparities in unemployment. The absence of full employment has thus both contributed to inequality and gaps in racial opportunity for much of the period that income inequality has grown.

The rightmost icon in the top row of the “Inequality’s Causes” slide represents both unions and labor standards. The decline of union coverage has taken an obvious toll on worker bargaining power. The figure below shows another measure of inequality — the share of income going to the top 10 percent of adults each year — against union membership rates. The fact that private sector union membership is now less than 6.5 percent is one reason, even as the economy closes in on full employment, wage growth has accelerated more slowly than expected.

The erosion of key labor standards has also helped hold wages down at both the middle and bottom of the income scale. Policymakers haven’t raised the federal minimum wage for the past eight years, for example, and the minimum wage’s value is 24 percent lower than where it was in 1968. At the same time, the share of income going to the bottom 20 percent has fallen sharply over the past 50 years, as the next graph shows.

Similarly, the salary threshold below which all workers must receive overtime pay if they work more than 40 hours a week has eroded, having lost more than half its value since 1975. As a result, it no longer provides the income support to middle-class workers — a group whose income share also has declined a great deal since the mid-1970s — that it once did.

Moving to the bottom row of the “Inequality’s Causes” slide, the monopoly man represents new research showing that the concentration of monopoly power and profitability is squeezing the share of labor income. Inequality between both wage earners and those making business profits has increased since 1980, but profits are far more skewed toward the wealthy than wages are. So, as numerous academic papers have shown, when more income flows to profits and less to wages, inequality rises.

Then there’s tax policy. As the next graph shows, the top marginal rate has fallen over time, contributing not to faster growth, as phony trickle-down arguments maintain, but (as one might expect) to more money for rich people. Rate cuts for the wealthy and the many loopholes that privilege high incomes both exacerbate inequality and deprive the government of revenue needed to offset inequality’s damaging effects through public investments.

Yet another factor is deregulation of the finance industry, which helped inflate the bubble that led to the Great Recession. The largely unchecked financialization of the economy over a longer period has contributed to the economic shampoo cycle — bubble, bust, repeat — that wreaks the most havoc on people at the bottom of the income scale.

Finally, the last icon in the bottom right-hand corner of the “Inequality’s Causes” slide stands for the political dimension of the inequality problem. Inequality is partly sustained by a feedback loop, as wealth concentration interacts with our heavily moneyed political system to produce policies that redistribute money upward and block policies that push in the other direction. Our politics are demonstrably more responsive to the interests of the wealthy than to the interests of the average American.

Given this multitude of causes for increased inequality (our list isn’t even exhaustive), reversing the decades-long pattern shown in the top figure may seem daunting. But remember, much of what has occurred is the result of conscious policy choices that have transferred power — and thus income and wealth — away from folks at the bottom and middle and toward people at the very top. Beneath all the data is thus a message that is simple and clear: It’s time we started making different choices.