I wrote recently that my mind was changed by the evidence on how much underwater homes were causing a decrease in mobility which in turn was causing higher unemployment. I believe it does not explain much of the current unemployment we are seeing. A new paper defends the connection between lower equity and lower mobility. The paper, by Ferreira, Gyourko, and Tracy is an update on an earlier paper of theirs that includes 2009 American Housing Survey data and improves some coding and econometric issues highlighted by another recent paper by Schulhofer-Wohl that was critical of their work.
One criticism that FGT makes of Schulhofer-Wohl is that some observations which they code as moves are in fact temporary moves, and not permanent moves. It is strikes me as debatable as to which type of move is more relevant for labor markets, and the effects of both are worth knowing.
Another issue is that knowing who has moved today from AHS data is easier to know once future data arrives, and so you can be conservative and code censor observations where move status is unclear, waiting for future data to clarify the issue. Or you can can generate a more inclusive measure of moving and risk including false positives. As FGT state, these coding decisions are consequential for the results:
…it still is useful to understand that the potential fragility of our results (and, possibly, those who came before us) arises from the fact that it is difficult to properly measure mobility in a number of cases.
In the end, it seems likely that underwater homes are decreasing housing mobility defined as permanent moves. I also agree with FGT that the true extent of this won’t become clear until future data arrives.
However, this falls short of providing evidence that housing equity is affecting labor markets. Looking at other information from AHS data, FGT note that:
Most moves are for quality-of-life, personal/family and financial reasons, and do not appear to be primarily job-related. This is especially the case for local moves. In contrast, longer distance moves, particularly those that cross a state border tend to be job-related. One potential implication of these data is that financial frictions to household mobility are more likely to reduce local moves such as trade-up purchases that need not have any significant spillover effects for labor markets.
This, they point is, is consistent with other studies on the issue. I agree with the reasoning here, and so I think it remains safe to conclude that the evidence suggests housing equity led mobility declines are not a significant cause of unemployment.