Full report ( 2,07 MB )

 

  The US recession of 2008 and the first half of 2009 has been the deepest and longest lasting since 1945. In the two-year period of 2008-2009, 8.75 million jobs were lost, durable consumption plummeted (as shown by the 24.4% drop in auto sales throughout the second half of 2007 and the whole of 2008), doubtful loans reached record highs and house prices have fallen by 31.7% since April 2006. In addition to this severe recession, the recovery is also turning out to be much more problematic and feeble than those following previous crises. Of these four indicators, only durable consumption seems to have recovered, thanks largely to massive government aid. On the whole, private consumption is currently at the level of the end of 2007.

  Finding out why the recession has been so severe and the recovery so slow is the million dollar question. One hypothesis blames the situation on the excessive indebtedness of US households and its impact on net financial and non-financial wealth when the value of assets falls. As explained by Fisher in his classic article on debt and deflation in 1933, the greater the leveraging, the larger the fall in net wealth caused by loss in asset value and, consequently, the greater the deterioration in the net wealth-income ratio and the sharper the drop in consumption. In the years prior to the crisis, the gross debt of households went from 90.1% of disposable income in the first quarter of 2000 to a peak of 130.2% in the third quarter of 2007. Similarly, the savings rate fell to an annual minimum of 1.4% of the disposable income of households in 2005, when the average for 1990-1995 had been over 6%. This increasing indebtedness occurred within a context of sharp rises in asset values, especially real estate, obscuring the risks involved. As a matter of fact, the relationship between indebtedness and prices was more causal than casual. The consequence was that, from the end of 2005 to the beginning of 2009 and with the fall in asset prices, net household wealth fell from a level equivalent to 448% of GDP to 352% of GDP.

  However, in addition to indebtedness, there are other factors that might also have contributed to making the recession worse. Firstly, the lower availability of credit on the part of the financial sector, hard hit by the Lehman Brothers going bankrupt and the difficulties encountered by many other institutions. We also need to take into account the relative weakness of investment in production goods, which in the pre-recession years was already growing less than in previous expansionary periods.

  One way of isolating the pre-crisis rise in indebtedness from other factors underlying the crisis is by using data broken down by county. Among the 3,143 counties of the United States, there are significant differences regarding the growth in leveraging in the years before the crisis, while the availability of credit from institutions is much more homogeneous, as national financial institutions are present in most counties. If those counties with the highest levels of indebtedness have been harder hit by the crisis, this might suggest that debt has really been a key factor.

  This suspicion has been confirmed. As shown by Mian and Sufi (2009),(1) two economists from the San Francisco Fed, pre-crisis household leveraging is an accurate indicator of the variation in the recession's severity among counties. Mian and Sufi's sample is made up of the 450 counties with the highest number of households, accounting for more than 70% of the population, divided into deciles based on the growth in indebtedness compared with income. The top decile is made up of the 10% of counties where growth in the indebtedness-income ratio between 2002 and 2006 was the highest. The bottom decile corresponds to the 10% of counties where the indebtedness-income ratio grew the least. They then compare these top and bottom deciles, noting the variables that indicate the severity of the recession, such as unemployment, durable consumption (for which they use auto sales), house prices, building permits and the default rate.

  The findings by Mian and Sufi differentiate between two periods. In the first phase of the recession, which runs from the last quarter of 2006 to the second of 2008, the counties in the decile with the highest growth in indebtedness were much harder hit by the crisis than the national average, while the counties with a low rise in indebtedness were almost unaffected. Counties in the indebted decile suffered a 40% drop in house prices between the second quarter of 2006 and the second of 2009, their default rate rose by 12 percentage points and auto sales fell by 40% compared with 2005. On the other hand, the counties in the decile where indebtedness had grown the least saw a 10% rise in house prices, while their default rate rose by a mere 3 percentage points and auto sales in the third quarter of 2008 were 20% above the level of 2005. Similarly, between the end of 2005 and the third quarter of 2008, the unemployment rate of high debt counties grew by 2.5 percentage points while that of low debt counties scarcely changed.

  In a second phase, as from the third quarter of 2008, the severity of the recession starts to be felt in all counties. This is explained, on the one hand, by the weakness that had affected those counties most in debt being transmitted to other counties. Moreover, of course, the bankruptcy of the Lehman Brothers was a shock that affected the economy as a whole, paralyzing the flow of credit overall.

  The conclusion is that the recession started sooner and was more severe in those counties where, in the years prior to the crisis, indebtedness had risen the most. In the second half of the recession, the effects of economic decline, added to the impact of the Lehman Brothers crisis, made the economic recession more widespread. In the future, the strength of the recovery will depend very much on the behaviour of household debt that is still above its long-term trend, although it had fallen to 114.1% of disposable income at the beginning of 2011 from a peak of 130.2% in the third quarter of 2007.

  (1) Atil, Mian, and Amir Sufi. 2010. 'Household Leverage and the Recesion of 2007 to 2009,' IMF Economic Review 58, pp. 74-117.

  This box was prepared by Jordi Singla

  International Unit, Research Department, "la Caixa"

---

All documents are in Adobe Acrobat format (PDF).
To view a document in PDF format you need the Adobe Acrobat Reader. If you don’t have it already loaded on your computer, you can donwload it now.