1
Does the Economy Affect Trends in Suicide?
Module 5 Group 4
Shelly Matushevski/mlm112
Karen Pavlisko/kmp43
Bennett Cowie/bc33
Matt Morrison/mwm16
University of Akron Fall 2009
Department of Economics
Suicide, unemployment, and gross domestic product data was retrieved from The Center for Disease Control and the Bureau of Labor Statistics, and complied together for analysis. The analysis of the data and the opinions of other economists, are used to conclude if a relationship between the suicide rate and the unemployment rate or changes in GDP exists. A regression analysis using SAS was used to test the data. A useful relationship was found between the suicide rate and the unemployment rate. However, the relationship between the suicide rate and changes in GDP was inconclusive.
I.INTRODUCTION
Can money buy happiness? The time old question appears as we try to answer another question: do times of economic hardship affect the rates of suicide in the United States? With a weakening economy, rising unemployment rates, and not-so-active changes in gross domestic product that define recessions, the average American likely faces economic hardship at home. Does this economic hardship spill over into a person’s mental health? Is there a direct relationship between the economy and suicide rates? These questions will be explored throughout the context of this paper.
II. THEORY
Does worry (about economic issues) lead to stress, then lead to depression, and then maybe even onto suicide? One might speculate that there is a relationship between an economy in recession and the deterioration of a person’s mental health.The deterioration would be spurred by a rising unemployment rate, falling profits for businesses, and fear of bankruptcy for an individual or business.Based on this theory one might conclude that constant worry can lead to extreme amounts of stress and drive an individual towards depression. In turn, one might wonder whether or not the suicide rate is in fact related to theeconomic business climate.
For the purposes of this analysis, we will be using the state of the economy as measured by the unemployment rate and changes in Gross Domestic Product (GDP). Although the only factors tested by our SAS analysisare unemployment and GDP, other factors (economic and noneconomic) are also discussed. The suicide rate will be measured as a percent of deaths per year. This paper will attempt to demonstrate that when unemployment rates rise and GDP falls, suicide rates increase; or, conversely, when unemployment falls and GDP rises, the suicide rate decreases, all other factors (economic or noneconomic) not being considered.
In general, many writings seemed to find links between suicide and unemployment, or suicide and economic variables. In fact, the economist Yong-HwanNoh states in his paper, “Does Unemployment Increase Suicide Rates? The OECD Panel Evidence (2009),”“the unemployment does significantly affect suicide rates.”He believes the unemployment rate “also potentially involves the loss of social networks and self-confidence.” This would seem to fit with the finding that, overall, as incomes rise, suicide rates decline (Hamermesh, 1974). Based on data gathered by the Bureau of Labor Statistics and Center for Disease Control, these writings will become even more valid, as direct relationships between suicides and economic conditions will be illustrated.
Throughout the research there were distinct differences between maleand female suicide rates. Men typically had higher rates of suicide; however, women were twice as likely to be suffering from depression (Oquendo, 2001). While this seems to disprove a relationship between depression and suicide, the authors of “Ethnic and Sex Differences in Suicide Rates Relative to Major Depression in the United States” (2001) found in a study that most suicide victims undergo major depression right before the time of death. Following the predicted trend, over the nineteenth century, peoples’ “happiness” has gradually gone up, along with the standard of living in the United States (Oswald, 1997) and the suicide rate has fallen.
Between the years 1985-1999 the suicide rate fell 13.5 percent. Overall, an association between the decline in suicide rates and the use of non-tricyclic antidepressants exists, as the use of these antidepressants increased over this time period (Grunebaum, 2004). An interesting point brought up in the article “Antidepressants and Suicide Risk in the United States, 1985-1999” touches on the relationship between unemployment rate and alcohol consumption. The article notes an overall decrease in alcohol consumption during that time period as well. Alcohol consumption is just one factor that affects the suicide rate although not thoroughly discussed in this analysis, it is interesting to note that unemployment affects both alcohol consumption and suicide rate, while alcohol consumption itself affects the suicide rate as well.
III. METHOD
For this analysis three sets of data were collected: suicide rate, unemployment rate, and percent change in GDP. All of these data factors were set against the variable “time” and flow over the time period from 1981 to 2006 to show current trends and patterns relating suicide to recessions. The unemployment rate and GDP data depicts the recessions in this analysis as both are economic indicators. Although unemployment is typically a lagging indicator (meaning unemployment rates usually increase after the beginning of a recession and don’t go back down until after a recession is over) it may play a role in the discussion of suicide. In fact, a paperby Oquendo, Ellis, Greenwald, Malone, Weissman, and Mann (2001) discussed three factors that they felt were key influences on a person’s depression. These three factors were: low income, unemployment, and disrupted marriages.
After the data was collected, SAS (Statistical Analysis Software) tests were run on the data to derive correlations between either the suicide rate and unemployment or the suicide rate and GDP. In both cases the suicide rate was the dependent variable. A regression was run on the data which can be found below in the “Results” section[1].
The data and SAS results, along with various economic writings (a listing of which is found in the Bibliography page) helped to arrive at the conclusion that the suicide rate is affected by unemployment rates.
IV. RESULTS
The following graph (Figure 1) shows suicide and unemployment rates in the United States from 1981 to 2006 and is a result of our merged data. In the general context of our data range several interesting trends reveal themselves. Most notably, suicide rates as a whole steadily decline for a 15 year period beginning in 1986. This decline clearly corresponds to the general decline in unemployment rates over our data range. In terms of a more specific approach, the recession of the early 1980’s is clearly shown by a sharp spike in the unemployment rate that peaks at 1983. Between 1981 and 1983 there exists no strong correlation with unemployment and the rate of suicide, however there appears to be a period of lag as unemployment then sharply and steadily declines. In fact, between 1983 and 1986, as unemployment fell, suicide actually rose. The peak of suicides during this time corresponds directly with a brief flattening-out of the unemployment rate that then continues to fall. The recession of the late 1980’s and early 1990’s is also clearly shown on Figure 1 with a well-rounded increase in unemployment. Again, there exists a slight lag between the data. Between 1990 to 1992 rates in suicide decreased while the unemployment rate increased severely. The two rates interestingly even-out from 1992 to 1993, and then continue to decline in unison roughly until the turn of the century. Therecession of the early 2000's, spurred mainly by the terrorist attacks of September 11, 2001, appears on the graph with a rather sharp increase in unemployment. This increase in unemployment directly corresponds with a significant increase in the suicide rate. The suicide rate also dips downward in relation to the leveling-out of the unemployment rate at the same time. After this point, however, the two rates being to split in opposite directions as suicide rates maintain a moderate rate of increase and unemployment rates steadily decline.
Figure 1
In the next part of our analysis we did separate regression analyses between suicide rate and unemployment rate, and between suicide rate and the changes in GDP. We used the method of hypothesis testing to determine if the relationship between the variables created or provided a useful model. A hypothesis test is a method of making a statistical decision using experimental data. A useful model is one that must have a p-value less than the generally accepted alpha value of 0.05. A p-value[2] represents the probability of error that is involved in accepting our observed result as valid. The value of alpha represents the probability of making a Type I error. A Type I error involves rejecting the null hypothesis when it is true. For purposes of our study the null hypothesis is that the model is useless, meaning the alternative hypothesis is that the model is useful.
In Figure 2, the regression line is shown graphing the suicide rate dependent upon changes in GDP. From the graph it can be seen that there seems to be a positive, weak correlation between GDP predicting the suicide rate. When the hypothesis test was run between these two variables, the p-value equaled .1350. This value is much larger than the alpha value of .05 which means that we cannot reject the null hypothesis which means that we cannot conclude that the model is useful. As Figure 2 shows, the R2 value[3] for this relationship is only .0907, which also indicates a very weak relationship.
Figure 2: Suicide Rate Dependent upon Changes in GDP
In Figure 3, the regression line is shown graphing suicide rate dependent upon the unemployment rate. In this graph there is a much more noticeable trend between the two variables and the results of the regression analysis agreed with this. When the hypothesis test was conducted between these two variables, the p-value was found to be .0007, which is less than the alpha value of .05. This means that we would reject the null hypothesis and conclude that the model is useful. The R2 value also shows the trend is much more significant at .3885. This shows that there is a useful positive relationship between the suicide rate and the unemployment rate, and this can then be taken to say that the unemployment rate can be a decent predictor for the suicide rate.
Figure 3: Suicide Rate Dependent upon Changes in Unemployment
V. CONCLUSION
Throughout the context of this paper it is shown that there is a relevant relationship between the unemployment rate and the suicide rate. There is a possibility of error in the analysis of the data. First, the data only spans twenty-six years. For a statistical analysis, this is very few observations. Also, it is stated that there should be a negative relationship between the suicide rate and changes in GDP to support the initial hypothesis that the suicide rate is inversely related to changes in GDP. A problem that presents itself with this is the time period that was chosen includes no negative changes in GDP; therefore, there was no chance that the analysis could have led to a negative correlation. This was because the suicide data that was collected limited the data to this time period. It should be noted that there are many other factors, besides those relating to the economy, that can and do affect a person’s decision to commit suicide.
Bibliography
Grunebaum, Michael, Steven Ellis, Shuhua Li, Maria Oquendo, and J. John Mann.
"Antidepressants and Suicide Risk in the United States, 1985-1999." Journal of Clinical Psychiatry. November 2004. (accessed November 5, 2009).
Hamermesh, Daniel, and Neal Soss. "An Economic Theory of Suicide." The Journal of Political
Economy. February 1974. (accessed November 5, 2009).
Noh, Yong-Hwan. "Does Unemployment Increase Suicide Rates? The OECD Panel Evidence."
Journal of Economic Psychology. August 2009. (accessed November 5, 2009).
Oquendo, Maria, Steven Ellis, Steven Greenwald, Kevin Malone, Myrna Weissman, and J. John
Mann. Ethnic and Sex Differences in Suicide Rates Relative to Major Depression in the United States. October 2001. (accessed November 5, 2009).
Oswald, Andrew. "Happiness and Economic Performance."The Economic Journal. 1997. (accessed November 5, 2009 ).
Ruhm, Christopher. "Are Recessions good for your Health?" The Quarterly Journal of
Economics. May 2000. (accessed November 5, 2009).
Yang, Bijou. “The economy and suicide: a time-series study of the U.S.A.” American Journal of
Economics and Sociology, volume 51, no. 1 January 1992 (pp. 87-99)
[1]Figures created by the information derived from SAS, and the data in a Microsoft Excel worksheet can also be found in “References.”
[2] This value will be obtained through a regression analysis in SAS.
[3] The R2-value represents how well the regression line is able to approximate the actual data.