(1/N) Thanks to those liking my @BrookingsInst #BPEA discussion, but I need to clarify one point, and I screwed up in the chat so also will be clarifying, explaining, and giving apologies to @profsufi and @AtifRMian below.
(2/N) First, clarifying (for @Claudia_Sahm and others) the general point: while providing data for policy guidance mid-crisis is very useful, we should not regard the answers in today’s papers as the definitive answers. Many examples of how our understanding has changed . . .
(3/N) A decade later, we are still not sure about a lot of the house price bubble, financial crisis and Great Recession. For example, ongoing work on relative roles for expectations vs. credit conditions: violante.mycpanel.princeton.edu/Workingpapers/… and dlgreenwald.com/uploads/4/5/2/…
(4/N) And the example I cited in my discussion was "subprime lending caused the financial crisis" -- the evidence now shows that it did not. Subprime defaults came first, but vast majority of $ losses did not come from subprime, & lending expanded across the income distribution.
(5/N) That is, the financial crisis was not caused because we lent to poor people, maybe pushed by the Community Reinvestment Act. Yet, I still see papers with this view, expressed here in the minority FCIC report in 2011: fcic-static.law.stanford.edu/cdn_media/fcic…
(6/N) Now, what I screwed up. . . People still sometimes write that subprime lending caused the crisis & cite Mian Sufi (2009) & I wrote in that chat that the paper was wrong, when I meant "subprime caused the crisis was wrong." I apologize to @profsufi and @AtifRMian
(7/N) I still want to kill off "subprime caused the crisis" but Mian Sufi (2009) spotted what I would now phrase as payment to income requirements were relaxed, not just collateral requirement. This contribution is a big part of current analysis, e.g. dlgreenwald.com/uploads/4/5/2/…
(8/N) Not *all* of Mian Sufi (2009) has stood the test of time (and Figure 7 is misleading), see the argument mitsloan.mit.edu/shared/ods/doc… and the response papers.ssrn.com/sol3/papers.cf…
(9/9) But not bad for a 2009 paper with a huge amount of focus on it. I am not sure if Mian & Sufi agree w/ all of above/my reading of the literature, but they are on Twitter & are deeper in this literature than I am. Given my public chat screw-up, this is my public fix.
(10/9) always adding one more Tweet: My 1st point - take all early conclusions with a grain of salt - partly covering my own ass. I drew a big implication: early consumption declines unrelated to income declines & causation was from consumption to income not the reverse.
Ooops, I think its Figure 3 not 7 (looks like apples to apples but compares multi-year changes to single year changes due to data availability).
