Socioeconomic factors influencing longevity in Japan.(ANTHOLOGY)

Although there are no obvious reasons for this phenomenon, it is sometimes stated that, in Japan, after they are separated from their partners (due to death, divorce, etc), old men arc prone to dying earlier whereas women tend to live longer lives. The purpose of this paper is to examine if this empirical observation can be confirmed by the socioeconomic data of the entire nation at the local government level. Although, available data arc limited, our results may suggest socioeconomic policies that may extend the average life expectancy by improving living conditions of both old men and women.

The explained variables are average life expectancies of men and women (Lm, Lw), and are provided by the Ministry of Health, Labor and Welfare (http://www.mhlw. g0.jp/toukei/saikin/hw/life/ckts05/index.html). The explanatory variables are taken from the data provided by the Ministry of Internal Affairs and Communications in Japan (http://www.stat.g0.jp/data/ssds/5b.htm) with some modification, as and when necessary. These are divorce rate (Dr), marriage rate (M), percentage of elderly couple households (Ec), percentage of elderly single households (Es), birth rate (B), death rate (D), income (/), unemployment rate (U), number of restaurants per capita (R), number of parks per capita (P), number of doctors per capita (Do), number of nursing homes per capita (N), number of traffic accidents per capita (T), and crime rate (C). Most of the data are those for the year 2005; the numbers have been adjusted as and when necessary.

We constructed a full model that included all the variables and applied ordinary least square methods. We deleted the explanatory variables one by one until all the p-values were less than 10%. The final models were as follows.

In the above equations, * and ** denote significance at the 10% and 5% levels, respectively. Otherwise, the variables are significant at the 1% level. The signs of all the parameters in the full models are preserved in the final model. The number of observations for the final models Lm and Lw are 1914 and 1920, respectively.

When we compared the Lm and Lw models, we found that most of the parameters had opposite signs. The average life expectancy of men decreased while that of women increased as Es and U increased. The opposite was true as Ec and I increased. These results suggest the following: First, men need women but not vice versa. Second, men tend to live longer if the employment situation and per capita income of the region are favorable. However, in the case of women, the average life expectancy decreased under such conditions. The signs of the parameters R, N, and P suggest that men who are introverts and women who are extroverts tend to live longer. This suggests that men prefer to stay at home while women prefer to go out and interact with others. The parameter Dr had a negative impact on men, while it may not have a significant impact on women. All of the above results suggest that during post-separation, men are prone to dying early while women live longer.

With regard to the other parameter signs, Lm and Lw (as explanatory variables) had positive values, which suggests regional difference in average life expectancy. This is reasonable given the context that if men live long in a certain region, the women in the region will obviously live long as well. The positive values of Do in both the cases are consistent with expectations.

Finally, we provide some implications of our results. Although most of the parameter values have opposite signs, there are some ways in which the situation of both men and women can be improved. For example, increasing employment opportunities is beneficial for men and increasing the number of public squares helps women, both of which may contribute to improving the average life expectancies of both men and women without causing any significant contradiction. Thus, our results imply that mere fiscal or monetary policies are not enough. Socioeconomic policies that improve citizens' social welfare are also required.

JEL J10-J12

Published online: 8 October 2008

Y. Kawata ([mail])

Department of Animal and Food Hygiene, Obihiro University of Agriculture and Veterinary Medicine, Inada-cho, Obihiro, Hokkaido 080-8555, Japan

e-mail: ykawata@obihir0.ac.jp

Print Get the Mag Weekly Updates