Conclusion

These results show that most migrants in the population aged 65 years and over are women and that women contribute nearly four-fifths of migrants in the 75 years and over age group. Most of this ‘excess’ is demographically determined and simply reflects the preponderance of women in these age groups. Rates of migration, however, were also slightly higher among women than men in both 1971-81 and 1981-91. This seems to reflect gender differences in household circumstances, household change-including transitions to

institutions-and health, rather than any more specific difference by gender in propensity to move. Family/household change is a particularly important factor in this age group and rates of migration among those experiencing such changes were much higher than among those remaining in the same family/ household type. In the second decade considered here, rates of transition to institutions were much higher, and rates of transition to complex households much lower, than in 1971-81 and, as a result, the relative balance of moves to institutions and moves to other types of private household changed. However, the overall volume of migration in the two decades considered was very similar. Finally, the analyses presented here are necessarily constrained by the data, which include no directly gathered information on motivations for moving. We are unable to say whether women and men have similar or different frameworks in which they reach decisions about moving. Research on this question would seem an important topic for the future.