Article Text

Download PDFPDF

Political regime and suicide: some relevant variables to be considered
  1. S Stack
  1. Wayne State University, USA
  1. Correspondence to:
 Dr S Stack;

Statistics from

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.

Future work needs to include unemployment, alcohol consumption, and strikes to the analysis

The study by Page et al assesses the impact of type of political regime (conservative compared with social democratic) on suicide rates over time in Australia. There are comparatively few studies on the politics of suicide and this is apparently the first on Australia. It is careful to control for some of the possible confounding variables such as GDP growth and world wars, that might render the relation between political regime and suicide spurious. Australia is known to have relatively sound historical data on suicide. These are among the reasons that the study makes a path-breaking contribution for the field of the politics of suicide.

There are, however, some limitations of the investigation that are apparently not fully addressed by the authors. These are largely in two broad areas: model specification, and review of the previous related literatures on the issues addressed. For a summary of the key works on the politics of suicide see the review by Stack.1

Firstly, the issues of model specification include the omission of standard variables in time series analyses of national suicide rates: unemployment, divorce, and religiosity.1 The omission of major sociological correlates of national suicide trends makes the findings somewhat problematic. If these variables were included, the results may have been the same anyway, but we cannot be absolutely certain until they are incorporated into future work. Perhaps adequate data were not available for the entire time period on the variables of interest.

Most important, the investigation does not include an exploration of the possible relation between type of political regime and the rate of unemployment. Previous works, uncited in the study at hand, have often found that while political integration is related to suicide risk, that once a control for unemployment is introduced the relation becomes insignificant. For example, Wasserman’s 2 work on the monthly suicide rate and presidential elections finds a dip in suicide during the month of an election, but this is, in fact, attributable to a corresponding tendency for unemployment to dip during presidential elections. Turning back to the Australian study, if political regime is associated with unemployment trends, a control for unemployment trends might render the reported relation between regime and suicide spurious, or the relation might weaken. Of course, this might also provide a key rationale for why political regime might be related to suicide. Regimes often lower or increase employment through such tools as manipulation of fiscal and monetary policy.

Some key works on the impact of war on suicide would also suggest that additional controls should be added to the model, if possible, in future work. For example, Marshall 3 found that once a control for unemployment was introduced into a time series equation of suicide in the US, that the relation between war and suicide vanished. That is, the real reason wars may reduce suicide is that they tend to reduce unemployment. Wasserman 4,5 has isolated another possible reason that wars reduce suicide. Alcohol availability tends to decrease because of the need for alcohol related products for the war effort. With alcohol production down, alcohol misuse is reduced. As alcohol availability is a key predictor of national suicide rates, the impact of war on suicide may be attributable to the reduction in alcohol misuse, not political integration in itself. Hence, the Australian study could be strengthened by a control for unemployment and indicators of alcohol consumption. It is possible that the observed relations between war and suicide in Australia would weaken under such controls.

Finally, future work might incorporate an index of industrial strikes as a possible mediating variable between political regime and suicide. Labor’s strikes against management may reduce suicide by uniting the labour movement and increasing its partisan spirit against a common enemy. A cross national investigation of 31 nations confirmed this position.6 However, a time series analysis of the US, where labour unionisation is at a low level by international standards, did not.7 If type of political regime in Australia is associated with strike activity, this could change the reported results.

The study is a fine contribution to the politics of national suicide rates. It should stimulate a considerable amount of work in the future in this neglected realm of suicide studies.

Future work needs to include unemployment, alcohol consumption, and strikes to the analysis