Article Text

other Versions

Bank failures, mortality and the Great Depression
  1. José A Tapia Granados
  1. Correspondence to Dr José A Tapia Granados, Social Environment & Health Program/Survey Research Center, Institute for Social Research, University of Michigan, 426 Thompson Street, P.O. Box 1248, Ann Arbor, MI 48106-1248, USA; jatapia{at}

Statistics from

I was quite surprised when I found my name in the acknowledgements of the paper by Stuckler et al.1 Since nobody requested my permission to put my name there, I would like to provide some comments now.

Though only data for 1929–1937 are considered in the paper, quite bold statements are made on changes in these years being consistent with longer-term patterns commonly referred to as epidemiologic transition. Consistent with this transition ‘chronic non-communicable diseases and injuries tended to rise during the Great Depression.’ However, as data in the same paper show, in 1929–1931, the worst years of the Depression, heart disease and cancer mortality deviated downward from its rising trend. Traffic mortality also had a major drop in the early 1930s. Since these deviations from trend during the Great Depression are downward, and in the epidemiological transition these rates of death tend to increase, the former cannot be due to the latter.

Changes in infectious disease and chronic disease mortality during the Great Depression are said to be ‘independent of bank suspensions’ so that ‘these changes cannot clearly be linked to the Great Depression.’ That statement is based in the dubious assumption that bank suspensions are a proxy for economic conditions. The peak of the banking crisis occurred in 1932–1933 well after the real economy had started to deteriorate in 1929 (table 1). While in 1929–1930 the unemployment rate tripled, bank assets increased by 2.7%; only in the next year they dropped by 8%. As explained by an author who studied the Great Depression in the USA, it would be expected that the states that had the most severe banking panics should have had the worst depressions, but data show there is ‘no correlation whatsoever between state income and the severity of the banking panic’.4 While the economy seriously deteriorated in 1929–1933, life expectancy at birth increased.

Table 1

Changes in national unemployment rate (UR), life expectancy at birth (LEB) and the banking system in the United States, 1921–1939

The paper sometimes compares the US Great Depression of the early 1930s—when mortality clearly decreased and life expectancy at birth rose—with the 1990s in Russia—when the Soviet regime broke down and mortality skyrocketed. This seems a most inappropriate comparison in the context of which the paper refers to increases in pneumonia mortality as if they perhaps may be ‘due to increases in hazardous drinking, as was seen in Russia in the 1990s, or associated with a deterioration in housing conditions’. But table 1 in the paper shows that pneumonia mortality in the USA dropped drastically from 1929 to 1933.

The paper concludes that except rising suicides and falling traffic fatalities, changes in mortality during the Great Depression cannot be attributed to the economy. In a different paper, Stuckler et al5 concluded that only increases in suicides and drops in traffic mortality are to be expected from recessions. I believe that this conclusion is wrong. As research by Ruhm,6 7 myself8 9 and others10 11 has shown, using data from a variety of market economies, economic expansions raise total mortality and cardiovascular deaths which, conversely, fall in recessions. Applying Ruhm estimates6 7 to the 2.5 million annual deaths observed in recent years in the USA, an economic recovery (ie, an expansion) reducing the unemployment rate from, say, 9% to 4%, is expected to be associated with some 60 000 extra deaths, of which almost a third would be heart attacks. In spite of the conclusions reached by Stuckler, this is what the historical experience suggests.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles