This is the summary of an article by Errol A. Henderson in the Journal of Politics, 60(2), May 1998, pp. 503-520. It uses the data for the u.S. from 1959 to 1992.

The main conclusions are as follows:

1. Increased peacetime military spending is positively correlated with poverty.

2. Increased wartime military spending is negatively correlated with poverty (i.e. we get richer when MS is up).

The explanation is the increase in military personnel in wartime, which absorbs labour, and the decrease in social programs if military spending goes up in peacetime.

The following are the alternatives for anti-poverty policy:

A. Increase social welfare spending. (Henderson finds this unacceptable in the present conservative political climate.)

B. Mobilize for war. (This is untenable as a social policy – I would hope.)

C. Convert defence industries to peacetime production in a way that is sensitive to poverty (i.e. maintains employment).

Apparently we should opt for C. according to Henderson, but I find A. acceptable as well.