By way of Mark Thoma and Simon Johnson I was led to this set of slides by Daron Acemoglu, from a presentation at the ASSA meetings in Denver.
The slides are titled "Thoughts on Inequality and the Financial Crisis." There is some hard evidence in there, but Acemoglu is mainly throwing out some ideas for discussion. A key aim of his is to debunk some ideas that you can find in Rajan's book, and in this "pre-report" from the apparently-dissenting Republican members of the Financial Crisis Inquiry Commission (FCIC). Those people think that an important cause of the financial crisis was the proclivity of the government to do things for the poor by way of credit market intervention, in particular through the Community Reinvestment Act of 1977 (CRA), and the GSEs (Fannie Mae and Freddie Mac). I discussed some of these issues here (learned some things too - see some of the exchange in the comments section and a later post on CMHC).
As Acemoglu points out, it's hard to find evidence that the US government has been particularly favorably disposed to poor people during the last few decades. As most economists are well aware, the dispersion in income in the US population has increased, and Acemoglu summarizes the research results as telling us that this is mainly due to "supply, technology, and trade." Here, I think he means scarcity of particular skills, technological change that has increased the skill premium, and import competition in goods that are low-skill intensive. Changes in some government policies may indeed have favored rich people, particularly 1990s welfare reform and changes in income tax rates in the second Bush administration, but for the most part the changes in the income distribution appear to be driven by the effects Acemoglu mentions.
Now, what bothers Acemoglu are some ideas he has about what is going on at the very top (i.e. above the 99th percentile) of the income distribution. Essentially the idea is that, even in a democracy, money is power, as money buys politicians. Too much wealth at the very top only perpetuates itself, as the very wealthy have too much power to relax regulatory constraints in ways that allow them to prosper at the expense of everyone else. The hypothesis here is that what is going on at the very top of the income distribution is related to deregulation (in particular of the financial industry). This deregulation is seen as essentially issuing licenses to steal. For example, the regulatory loophole that permitted mortgage originators to make loans to low-income people who were incapable of ultimately meeting the loan payments, with those originators then selling the loans and making off with the profits was, in this view, essentially a social waste. Income was transferred from poor to rich, more houses were built than should have been, and the borrowers and the ultimate holders of the mortgages were left to sort out the losses and pay the legal fees.
Acemoglu certainly has a point. While Rajan has some interesting things to say about Fannie Mae, Freddie Mac, and the CRA, for example, it seems his political economy story is wrongheaded. However, Acemoglu seems to want to claim that the thrust of federal government policy was to "marginalize" the GSEs, and I don't think that is correct. Indeed, Fannie and Freddie are part of the phenomenon he is discussing. Fannie and Freddie were granted tax advantages and implicit (now explicit) government guarantees that gave them a cost advantage over private sector competitors. In fact, one could argue that it was the Fannie/Freddie cost advantage that drove the private sector to the only profitable game in town - risky mortgage lending. Further, the huge growth in Fannie and Freddie, given their implicit subsidy, created institutions with a huge amount of power, and they used that power to lobby Congress for favors. We can think of Fannie and Freddie as being run by another set of rich people at the top of the income distribution.