In the previous chapter we studied the excess-suicides that arise when a substantial fraction of a population cannot get married. In this and the next chapters we study other mechanisms through which the number of unmarried people is inflated, for instance:

• An increase in the mean age at marriage which results in a greater proportion of unmarried men and women, especially in the 20–30 age group.

• An increase in the rates of divorce or widowhood which results in more people living without a partner.

In sociological studies postponed marriage, divorce and widowhood are usually considered separately. However, seen from a network perspective these are three facets of the same phenomenon, namely the non-establishment or severance of marital bonds. If m, d, w denote the marriage, divorce and widowhood rates respectively, it is possible to define a generalized marriage rate as mg = m – d – w; this rate encapsulates both the establishment of new links and their dissolution through divorce or widowhood. Our objective is to see to what extent these rates are connected with suicide rates. In accordance with the extreme value technique defined in earlier methodological guidelines we first examine three fairly extreme situations: (i) the suicide of young widowers; (ii) the effect of a 200% decline in marriages rates; (iii) the effect of a huge and sudden increase in marriage rate.

Suicide rate of young widowers

It has been known since Emile Durkheim that widows and widowers have a higher suicide rate than married people. Similar observations have been made in many time intervals and countries.