Divorce and Remarriage: Occupational Hazards

The number of one-parent families has risen by 60% during the past 15 years. 70% of this increase is attributable to the growth in the number of divorced or separated mothers. At a lunchtime meeting on 23 January, John Ermisch reported the results of new research into the social and economic factors that influence divorce and remarriage, based on the 1980 Women and Employment Survey. His analysis, reported in CEPR Discussion Paper No. 140 , is the first UK study of the influence of women's work experience and earning capacity on the probability of divorce and remarriage. Ermisch found that women with more work experience and earning capacity face a greater risk of divorce, but they are also more likely to remarry. Current patterns of divorce and remarriage in the United Kingdom mean that divorced women who remain heads of one-parent families tend to be older, to have less earning capacity and to have larger families; their economic situation is therefore precarious.

John Ermisch is Research Officer at the National Institute of Economic and Social Research and a Research Fellow in CEPR's Human Resources since 1900 programme. He has published widely on the economic causes and consequences of demographic change, particularly in relation to women. The meeting at which he spoke was one of a series organized jointly with the British Society for Population Studies (BSPS), designed to present the results and policy implications of recent demographic research.

Ermisch first discussed the problems involved in measuring the expected economic consequences of divorce, which he defined as the change in the expected value of income flows (adjusted by needs) associated with divorce. The 'equivalent income' of a recently divorced woman in the United Kingdom is 40% less than that of a married woman. After taking account of the pattern of remarriage probabilities, however, Ermisch found that divorce reduces the expected average annual income over the 15 years following divorce by only 20%, not 40%.

Ermisch's research focused on the factors which affect the risk of divorce in a woman's first marriage and the speed of remarriage. His research confirmed some earlier findings: at all durations of marriage, the risk of divorce declined with age at marriage. Ermisch speculated that women who married younger might have spent less time searching for a suitable partner. The pattern of childbearing also affects the risk of marital dissolution. Women who experienced a pre-marital pregnancy faced a higher risk of divorce or separation, probably because their pregnancy had discouraged further search for a suitable partner. Early childbearing within marriage also increases the risk of marital breakdown at marriage durations of five to ten years, as CEPR Research Fellow Michael Murphy has discovered. Ermisch therefore concluded that there was no evidence that early childbearing makes marriages more stable, nor that larger families are less likely to split up. But couples who were childless beyond their fifth wedding anniversary did face a higher risk of marital breakdown. This may be because the presence of children discourages divorce; alternatively, childlessness may be associated with unanticipated fertility impairment or may facilitate the search for another partner.

Ermisch noted recent research by CEPR Programme Director Heather Joshi, which confirmed that more work experience increases a woman's earning capacity. Economic theory suggests that the advantage to a woman of the traditional marital division of labour is reduced when her earning capacity is closer to her husband's. The WES data made it possible for the first time to measure how women's participation in paid employment affects the risk of marital dissolution. Ermisch's analysis showed that during the first 10 years of marriage, an additional year of paid employment increased the odds of divorce by 6% among the women in the survey. Ermisch distinguished between work experience before and after the first child and found that work experience after the first birth had a larger positive effect on the risk of marital breakdown.

Ermisch warned that these effects could represent 'reverse causation'. Women who believe they face a high risk of divorce may participate more in paid employment in order to insure themselves against divorce, but he thought it unlikely that this could fully account for his findings.

These influences mean that newly divorced women tend to be to have started childbearing earlier or be childless, and to have higher earning capacity, than women who remain married. But the average characteristics of families headed by women also depend on who remarries and how quickly. Ermisch therefore analysed the effects of work experience, age and childbearing on the probability of remarriage within three years of divorce or separation. He found that this probability increased with women's work experience. Ermisch conjectured that women with more work experience during their first marriage may have better opportunities for receiving a better marriage offer, thereby precipitating divorce and rapid remarriage. In addition, work experience in the first marriage is strongly related to participation in paid employment after the end of that marriage, which is likely to provide better opportunities for finding a new mate.

Ermisch's results suggested that the presence of children in small numbers was not an important obstacle to rapid remarriage. Women with only one, two or three children appear as likely as childless women to remarry within three years. Mothers of four or more children are, however, about 65% less likely to remarry at any given interval after divorce or separation. Another strong influence on the probability of remarriage is a woman's age at the time of divorce or separation, possibly reflecting the steep decline with age in the ratio of unmarried men to women.

The characteristics of one-parent families reflect the factors influencing not only the likelihood of divorce but also the propensity to remarry. Thus, while the median duration of lone parenthood among all women whose first marriage ended is almost five years, as CEPR Research Fellow Kathleen Kiernan has discovered, the expected duration for a particular woman depends on her age, the size of her family, and her work experience. The tendency for older women, women with lower earning capacity and women with large families to remain unmarried clearly affects the economic characteristics of divorced women and one-parent families.

Many of those present at the talk wanted more detail on the issues Ermisch had raised. Had the rise in divorce rates led to any change in the relationships between risk of divorce and age at marriage, childbearing or work experience? There is evidence that the relative risks associated with age at marriage are widening, said Ermisch. Does the type of work experience alter its effect on the risk of divorce? Apparently not, according to Ermisch, but full-time employment has a much more powerful impact than part-time employment.

Economists had concentrated on the role of labour supply as an influence on the income and welfare of households but, Ermisch argued, the composition of households had a far stronger effect, which had so far received little attention from economists. In future research, Ermisch will combine his models of divorce and remarriage to simulate their joint effect on the characteristics of one-parent families. In addition, the factors which influence divorce and remarriage are themselves affected by demographic phenomena. Ermisch's current research with Heather Joshi aims to analyse more comprehensively the interrelations among work experience, age at marriage, childbearing, and household formation and dissolution.