Conclusion

The research described herein provides information on the persistence of the effects on government expenditures of an intensive program of nurse home visitation conducted in Elmira, New York. It also provides evidence about the effects of the same program in two higher risk populations in two disparate sites.

It is apparent that the beneficial effects on dependence on government programs and on labor force participation and income found in a 1993 study of the Elmira program have indeed persisted and in fact increased. These effects have been most pronounced in those who began the program with the lowest socio-economic status.

Replication of the program in Memphis and Denver has so far had more modest effects than did the original program in Elmira. In both Memphis and Denver, nurse-visited families were less dependent on welfare and contributed more in taxes than did the comparison families, but the impact on welfare was substantially less than we observed in Elmira. This is likely due to the impact of welfare reform, which placed five-year limits on life-time use and two-year limits on any one period of use. Moreover, the extremely favorable economic conditions of the mid-to- late 1990’s meant that many women, including those in the control groups, could find work more easily than did women in the period covered by the Elmira trial. In addition, as noted above, at least in the Memphis study, nurse-visited women indeed made greater use of subsidized child care, which enabled them to work more, but which incurred government costs, which offset savings in other areas.

Overall, the difference government expenditures for Memphis and Denver is not very large in comparison to control-group expenditures: 9.6% less in net government expenditures for nurse-visited families in Memphis and 8.2% less in Denver than for comparison group families. But the evidence provided by the Elmira study that differences found when the study child was age four persisted and increased over the next 11 years suggests that the differences between nurse-visited and comparison group families in both Denver and Memphis will continue to increase and that program costs may be fully recovered over time.

We were surprised that the Elmira program produced no effects on education costs. Recent analyses of children’s school readiness in Memphis and Denver suggest that there is some likelihood of program impact on education costs in those sites. Subsequent follow-up studies of the Memphis and Denver samples as the children enter school will provide estimates of additional economic benefits to government.