This gem, for example, is delicious since it shows Krugman's ongoing intellectual dishonesty, as his explanations shift.
Critics were quick to point out that Krugman had his facts wrong.Rajan goes on to shred Krugman a half-dozen more ways from Sunday. The whole piece is well worth a read.
As Charles Calomiris and Peter Wallison of the American Enterprise Institute (Wallison is also a member of the financial crisis inquiry commission) explained, “Here Krugman demonstrates confusion about the law (which did not prohibit subprime lending by the GSEs), misunderstands the regulatory regime under which they operated (which did not have the capacity to control their risk-taking), and mismeasures their actual subprime exposures (which he wrongly states were zero).”
So, Krugman shifted his emphasis.
In his blog critique of a Financial Times op-ed I wrote in June 2010, Krugman no longer argued that Fannie and Freddie could not buy subprime mortgages. Instead, he emphasized the slightly falling share of Fannie and Freddie’s residential mortgage securitizations in the years 2004 to 2006 as the reason they were not responsible.
Here again he presents a misleading picture. Not only did Fannie and Freddie purchase whole subprime loans that were not securitized (and thus not counted in its share of securitizations), they also bought substantial amounts of private-label mortgage-backed securities issued by others. When taking these into account, Fannie and Freddie’s share of the subprime market financing did increase even in those years.
In the end, as the Grand Poobah says in The Mikado, it's nice to have my opinions confirmed by an expert.