Gender difference in suicide, household production and unemployment

Abstract

This article aims to explain why men's suicide rate is generally higher than women's and why the former tends to fluctuate with unemployment. Adopting Hamermesh and Soss's suicide model (1974), with a two-period household production model, I argue that (1) the gender gap in suicide rate increases with the unemployment rate, because unemployed men suffer a larger ‘human capital loss’, due to the division of labour within their household and (2) men's suicide rate is generally higher than women's because of the shorter expected life of the former. Both international and US evidences support this hypothesis.

Acknowledgements

First of all, I wish to express my gratitude to Dr William Chan. This article would not be completed without his guidance and encouragement. I also want to thank Prof Wing Suen, Dr James Vere, Prof Daniel Hamermesh, Mr. Tom Lam, Mr. Julian Chan, Dr Hyman Cheung, Mr. Mole Hau and Dr Ho-yin Yick for helpful comments and suggestions. The usual disclaimer applies. The last but not the least, I would like to express my indebtedness to Jennifer, Alvin and Jon. It was Jennifer who encouraged and supported me to start from Diploma in Economics and continue my study after completing it. Alvin and Jon showed me their support through their smile and cry. This article is a modified version of my Master of Economics dissertation, ‘Gender Difference in Suicide, Household Production and Unemployment’ (2006) at the University of Hong Kong.

Notes

¹ In this article this definition is used mainly.

² My study focuses on analysing the completed suicide rate. I assume that people who commit suicide must kill themselves successfully. In other words, completed suicide and attempted suicide are considered as two different phenomenons. For a model of the dynamic of suicide attempted and completion of teenagers, see Cutler et al. (2000).

³ Note that all the Z's in this article, except those with time period indicators, are the total discounted value of the commodity in the remaining life-time.

⁴ The last assumption is not unrealistic, although the total number of women is more than that of men, in all the developed countries. This result is generally reversed if we consider only people aged below 50 (United Nations Population Division, 2004).

⁵ See Proposition 1.

⁶ In this article, for simplicity, without alternating the empirical implications, I assume that the number of employed men is larger than that of women. If this is not the case, some women would be forced to choose to remain married to their unemployed husband and not all couples with an unemployed husband would divorce.

⁷ Theorem 2.1 in Becker (1991) applies here.

⁸ Becker's Theorem 2.2 (1991) applies here.

⁹ The result does not change much even if the 15–24 years old age group is included in the data set.