As gushed forth by Poinçon and Wattman

There was some media coverage last week over an analysis which showed a causal relation between writing, and in particular using gloomy adjectives, and economic recession. The media gave it an airing because it was one more way to report on the financial crisis that has been a slow curse for over five years. The analysis is quite technical and it’s not right to discuss the validity of the correlations here. If you’ve the time and the inclination it’s worth reading as it’s not too long. Producing positive correlations is not very difficult and no one is going to publish without ensuring they have a few to demonstrate the strength of their analysis.

The objective of the analysis was to investigate how and why art comes about; what are the influences and themes of a particular time that writer may produce in their work.

At least initially this paper looks plausible but then like Lucky’s speech it is obvious that several questions make it dubious.

As a literary exercise the authors counted sad adjectives in English and German literature and correlated them to a time lag series after a critical financial event. That might work in a simple accounting method but when dealing with real books, written for other humans, this technique is too simplistic. The incidence of depressing words does not make a cogent case for some deprivation during a recession which has found expression over a decade later. The causal connection is too fraught with other factors. This is almost like astrology in assigning individual character and destiny to a distant star.

Where the assumptions and method with the literary analysis are unconvincing, the financial terms of the analysis are naïve. The authors used the inflation and unemployment rates to create a misery index. Economic misery coincides with WW1 (1918), the aftermath of the Great Depression (1935) and the energy crisis (1975). The Great Depression did not end in 1935, and in fact unemployment rose from 1937 in the US. The energy crisis and years of 1970s stagflation are very different to mass unemployment, in excess of 20% in the US, during the 1930s.

The most obvious objection is that inflation and unemployment are counted differently and have been for over 30 years as governments have changed the terms of reference. Since the 2008 crisis deflation, not inflation, has been the concern of central bankers and that has led to asset price inflation through low interest rates which in turn has seen a transfer of wealth from the poor and middle classes to the rich. These are different financial crises which affect people in very different ways and the application of a misery index based on unemployment and inflation is not credible.

In a wider sense the analysis has a bigger problem in that this address
got many economists talking about ‘secular stagnation’; that is low growth and periodic bubbles as the norm since the mid 1980s. If this argument is valid, and its one that many non-orthodox economists have held for a long time, it means the assumption of a crisis as posited in this literary causal analysis is not valid at all.

For years efficient market hypothesis was an idea that held minds until the center of the financial system melted. Its assumptions and method could not explain, nor act as a guide for players in the markets. Any forward looking statements, as they say in finance, on the correlation between recessions and sad literary works should be treated with extreme caution. The financial industry, despite its many overt failings, is quite good at setting the correct price on assets right through to junk.

Guy Cranswick
12th January 2014