After cancer and heart disease, suicide accounts for more years of life lost than any other cause of death, both in the United States and in Europe. In 2013 there were 41,149 suicides (12.6 every 100,000 inhabitants) in the US. To contextualize this number just think that the number of motor vehicle deaths was, in the same year, around 32,719 (10.3 every 100,000 inhabitants). Non-fatal injuries due to self-harm cost an estimated 2 billion dollars annually for medical care. Another 4.3 billion dollars is spent for indirect costs, such as lost wages and productivity. In the EU (at 27 member countries) statistics are very similar to those of the US: in 2010 there were around 60,000 suicides (12.3 every 100,000 inhabitants) and 38,119 motor vehicle deaths (7.8 every 100,000 inhabitants).
Since Durkheim’s famous study (Suicides, 1897) there has been extensive quantitative evidence in sociology, psychology, and psychiatry, as well as economics on the personal, social, and economic conditions associated with suicides.
The modern approach to suicide stresses its immorality and its connection with mental illness. This approach is in contrast with Greeks and Romans as well as Asian cultures (especially Japanese and Indian) that consider suicide to be a rational response to illness, disgrace, and other pain and suffering. David Hume, one of the most prominent British empiricists, and Arthur Schopenhauer, the famous German philosopher, looked upon suicide as a perfectly reasonable response to bad circumstances. Hume (Of Suicide, 1783) writes: “I believe that no man ever threw away life, while it was worth keeping” while Schopenhauer (Studies in Pessimism,1891) says: “It will generally be found that, as soon as the terrors of life reach the point at which they outweigh the terrors of death, a man will put an end to his life.”
Economists have tried to rationalize the decision to commit suicides, looking at the correlation with contemporaneous socio-economic factors being unable to identify a causal link so far. In particular, they have analyzed social contagion (an emulation of someone else suicide), unemployment, income, economic recession, education (as an important determinant of income), age, gender, monetary incentives (e.g. insurance coverage), female labor force participation, alcohol consumption, divorce and marriage rates, social isolation, population density and trends, health and health care, cultural factors/norms (such as ethnicity, religion, social capital, or the social stigma that suicide may carry), homicide rates, geographical and climatic conditions, lifestyle, civil liberties and quality of government. Typically they have found a positive correlation between suicides and adverse economic conditions.
One of the main obstacles to identify a causal relationship is the difficulty to measure expectations about one’s future wellbeing.
expectations and uncertainty about future conditions play an essential role in the decision whether to commit suicide in prison.
Starting from the Becker and Posner (Suicide: An Economic Approach, 2004) definition of rationality “people maximize their utility in a forward looking fashion taking account of the uncertainty of future events and the consequences of their action,” we tested whether there is a rational component in the decision to commit suicides. It is very hard to measure suicides and expectations because there are sparse data on suicides and there are no data on expectations. We mimicked a controlled experiment, looking at prisons, which are closed and isolated environments, where expectations can arguably be measured.
It is fairly easy to determine the inmates’ present discounted utility because it depends on how soon they will get released. To get out of jail is their main concern. We measured inmates’ changes in expectations using the timing of proposals of an unusual policy: nationwide large sentence reductions (collective pardons or amnesties). They must be proposed by a member of the Parliament and issued by the legislator (both the parliamentary chambers) with an absolute majority requirement of 2/3 (simple majority before 1992). When collective pardons and amnesties are granted, inmates with a residual sentence length of a given number of years, usually 2 or 3, are altogether released within a few weeks. In the last pardon, beginning of August 2006, more than 20,000 inmates were released within 1 week. Very few criminals are excluded from these sentence reductions (sexual offenses, mafia related crimes, terrorist attacks, and kidnapping). What we find is that suicide rates fall, on average, when new pardons are proposed. This means that suicide rates in Italian prisons particularly, do respond to changes in expectations about the time of release. Pardon proposals represent “good news” for inmates because they change the expectations of reduction in their sentence length.
We provided evidence that expectations and uncertainty about future conditions play an essential role in the decision whether to commit suicide in prison. Criminals who are spending time in prison exhibit forward looking behavior that proves that suicides have a rationalisable component. Furthermore we showed that criminals, who are often viewed as boundedly-rational or even irrational individuals are, indeed, rational.
Our results suggest that providing inmates with hope (perhaps through programs of “early release for good behavior”) can effectively reduce the risk of suicide in prison. Whether this can be generalized to the rest of the population remains an open question though it is highly likely that even there the decision to commit suicide contains a rational component. If this is the case those treatments that improves someone’s present wellbeing or future expected wellbeing could be effective in, to some extent, discouraging people from making the irreversible choice of killing themselves.
Featured image credit: silhouette of a man in window by Donald Tong. Public Domain via Pexels.