Complicated Southern European recessions

On both sides of the Atlantic, the last recession was unconventional. This has meant that the mainstream business cycle models needed to be rethought to make good sense of what is happening. By that I mean, they need to be augmented or altered, not thrown out completely, as some have claimed. In doing so, one needs to identify new channels for the transmission of shocks, and possibly new shocks as well. And because this is unconventional, it is sometimes difficult to wrap one's head around some of those models.

One good example of that is the paper by Zhen Huo and José-Víctor Ríos-Rull that tries to understand the last Southern European recessions. They are looking for a model that would explain simultaneously an increase in savings while wealth and employment are decreasing. To make it work, they need frictions in the reallocation of resources across sectors, frictions on the labor market, and some shock that increases the savings rate. The latter generates then a paradox of thrift: even though savings are up, wealth is down. This comes by in the following way: as savings go up, consumption is down in a way that reduces the number of varieties of goods that are demanded. This leads to excess capacity, an apparent decrease in total factor productivity and thus the value of firms and capital decreases.

Now, the shocks triggering this are shocks to patience. Alternatively, it works as well with shocks to financial costs. Yet, I have a hard time believing that these were the triggers of the last recessions in Southern Europe. It seems to me that wealth decreased before the savings rate went up. And the reason of the former was a sudden recall of debt by Northern savers (in particular banks) that needed to cover losses in US mortgage instruments. In other words, it may have been as simple as a negative wealth shock triggering a standard decrease in consumption that gets the ball rolling. The financial costs came after. The labor market frictions and the reallocation frictions should also be enough to prevent labor to increase.