I don't always agree with Kevin Drum, but usually I find that he has good arguments. In this passage, however, I think he is overreaching:
Still, I don't think that a plausible story of causation [the gains of the wealthy to the stagnation of other incomes] is really all that hard. First, take a look at middle class income stagnation. What caused that? Matt already pointed to one cause: monetary policy since the late 70s that's kept inflation low at the cost of keeping labor markets persistently loose. To that, I'd add several other trends that have marked the past three decades: trade policies that accelerated the decline of U.S. manufacturing; domestic deregulation policies that squeezed workers; stagnation in the minimum wage; immigration policies that reduced wages at the low end; and a 30-year war against labor that devastated unions and reduced the bargaining power of the working class.
First, money matters in the short-run most of all. Tight money (unless maybe it is radical deflation, but even then the U.S. resumed growth out of the GD fairly quickly and furthermore the median worker was not unemployed) is not a plausible cause of median income stagnation over decades. The link between trade and wage stagnation does not find support in the data. Deregulation hurt the wages of air traffic controllers, but how many other groups? Enough to shift the median? Stagnation in the real minimum wage shouldn't much hurt median wage growth over a forty year period. Unions fell mostly because of the shift to services, and furthermore the "union wage premium" in the data is a one-time ten to fifteen percent gap, not an ongoing change in rates over decades, and it is reaped by some not all workers.
In my view the gains of the top one percent and the stagnation at the median are largely separate phenomena; note that the top gains so dramatically only in the Anglosphere, whereas growth stagnation affects most of the OECD after the early 1970s.
How then might rent-seeking in the top income brackets damage ordinary Americans, as I have myself suggested? The most plausible mechanism is this: gains of the wealthy through capital markets come at the expense of other individuals in capital markets. Some kinds of capital — for instance finance capital – had been profiting more and implicitly taxing other kinds of capital, often through "trading," broadly not narrowly construed. The taxed capital "retreats" and this has adverse affects on labor, much as an increase in the corporate income tax falls partially on labor and partially on consumers (which also shows up as lower real wages).
The labor which is "subsidized" by this change in capital returns is smaller in number and more concentrated, sectorally, than the labor which is taxed.
I wouldn't say there is massively firm evidence for this hypothesis, but rather it is the major contender by process of elimination. If you are wondering, many of the years of wage stagnation have shown relatively low shares of investment in gdp; see the new McKinsey report "Farewell to cheap capital?".
So there we have one factor connecting the gains at the top to losses at the median but still I do not think it is the major factor behind median stagnation. I will be writing more on this.