Table of Contents | Previous | Next |
Conceptual Framework and Rationale for Expecting Long-Term Economic Impacts on Government Costs
Figure 1 illustrates the conceptual underpinnings of the NFP and the way we have hypothesized that it reduces government expenditures. It provides the framework for our assessment plan.
Cost Outcome Domains
As illustrated in Figure 1, our findings from the Elmira cost study through the first four years of the children's lives suggested that the NFP would reduce government costs in five domains through a number of common pathways (Olds et al., 1993), which have formed the basis for our hypothesizing corresponding effects in each of the replication sites. First, by improving maternal health, quality of parental caregiving and child health, and by reducing subsequent pregnancies, we hypothesized that the NFP would reduce health-care costs (Medicaid costs from the standpoint of government). Second, by improving parental caregiving and reducing subsequent pregnancies, we hypothesized that the NFP would reduce costs associated with child abuse and neglect (investigation, on-going follow-up of cases, and foster care). Third, by improving parental caregiving and child development and by reducing the number of subsequent pregnancies (and children), we hypothesized that the NFP would reduce education costs. Fourth, by reducing the number of subsequent pregnancies and children and by increasing workforce participation, we expected the program to reduce public assistance costs. Fifth, by reducing the number of subsequent pregnancies and children, and increasing parental workforce participation, we hypothesized that the NFP would increase family income and tax revenues. The anticipated effects of home visitation in these five domains are discussed below.
Our analysis of the 15-year follow-up in Elmira suggested that we will find treatment effects in a sixth domain, criminal justice costs, but this domain is not shown in Figure 1 as the children in the Memphis and Denver trials are not yet old enough to incur these costs.
Reductions in Health Care Expenditures. It is important to note that our experience in the Elmira and Memphis studies indicated that total health-care encounters are not likely to be reduced substantially. This is because home-visited women and children are likely to make a larger number of visits due to increased detection of health problems (such as UTI among women during pregnancy) and to have fewer visits for prevented health problems (such as injuries and ingestions on the part of children). Total health-care expenditures are likely to be reduced, nevertheless, because of the influence of the program on the rate of subsequent pregnancy (which will affect pregnancy and delivery-related health-care costs). This will be important in those communities in which Medicaid costs are covered on a fee-for-service basis (in this study, in New York State only). In Tennessee and Colorado, most Medicaid patients were enrolled in managed care organizations.
[D] |
From the perspective of reducing government expenditures, we expected to find reductions in Medicaid and other government health care expenditures from the prevention of health problems but increases in Medicaid costs to the extent that the nurses found maternal and child health problems that otherwise would have gone undetected or treated. Total government-sponsored health care costs for the home-visited families were expected to be lower than for the comparison group, nevertheless, because families receiving home visitation would have fewer subsequent pregnancies. That portion of Medicaid cost reduction that results from reduced participation in the Medicaid program is discussed under "Reductions in Public Assistance Costs," because, like other reductions in public assistance, it results from increased workforce participation and corresponding reduced eligibility.
Reductions in the Costs of Child Abuse and Neglect. We also hypothesized that home visitation reduces government expenditures associated with child abuse and neglect, including the cost of programs that investigate, monitor, and treat abuse and neglect. This reduction will occur as a result of the program's directly influencing the quality of parental care as well as of the visited families' having fewer subsequent children (Olds et al, 1988, 1997; Kitzman et al., 1997, 2000). It is important to note that, in spite of increased detection of child maltreatment due to the involvement of the nurse (Olds et al., 1995), nurse-visited families in the Elmira trial had substantially fewer verified cases of child abuse or neglect during the first 15 years after delivery of the index child. This will be reflected in fewer government expenditures on investigation, on-going services, and foster care. Some of these effects on state-verified reports of child abuse and neglect will only begin to emerge as the children become older, however, as surveillance bias will increase the detection of low-severity cases while the program is in operation and in the first few years after it ends (Olds et al., 1995).
Reductions in Education Costs. We also expected the NFP to reduce the costs of children's education. To the extent that home-visited children had fewer neurodevelopmental disorders (impaired IQ, attention deficit, hyperactivity) resulting from improved health-related behaviors during pregnancy and improved parental caregiving in the early years of the child's life (as the Denver year-4 results now suggest), we hypothesized fewer placements of home-visited children in special education classes, fewer grade retentions, and fewer behavioral problems leading to school suspensions. Education costs also are likely to be reduced as a result of the home-visited parents having fewer subsequent children, some of whom will require special education.
Reductions in Public Assistance Costs. We expected to find that home visitation reduces public assistance costs. In the Elmira study, nurse-visited at-risk women participated in the work force to a greater extent and had fewer subsequent pregnancies during the first four years after delivery than did their counterparts in the comparison group. This led to substantial reductions in public assistance costs for low-income families. The 15-year follow-up data show even stronger effects on completed family size and months on AFDC (Olds et al., 1997). In the Memphis study, we found reductions in subsequent pregnancy, longer intervals between the birth of the first and second child, and reductions in AFDC and food stamps use among nurse-visited women by the time the first-born children were four years old (Kitzman et al., 2000). These findings led us to hypothesize that public assistance costs (AFDC/TANF, food stamps, Medicaid, WIC, subsidized child care, housing subsidies, energy assistance) eventually would decrease among the nurse-visited women in Denver, and that these effects would increase over time.
Because the NFP produced effects on employment, we expected that subsidized childcare expenses might increase for the nurse-visited group, at least in the short run. Especially since welfare reform was enacted in 1996, an increasing public investment has been made in government subsidized childcare so women could enter the work force, so we expected that subsidized childcare expenses would be increased for the NFP in the Memphis and Denver sites. We therefore examined subsidized childcare as part of public assistance costs.
Another government investment aimed at helping women gain economic self-sufficiency is job training, which the nurses helped families access. Job training is thus another government benefit that was factored into our estimate of government costs, given the likelihood that nurse-visited women would make greater use of it in their efforts to become financially independent.
Increases in Family Income and Tax Revenues. We also expected the home-visitation programs to increase family income and tax revenues. In the Elmira study, nurse-visited at-risk women participated in the work force to a greater extent than did their counterparts in the comparison group (Olds et al., 1988). Corresponding effects have emerged in the Memphis and Denver trials through child age two (Olds, 2002). These findings led us to hypothesize that earnings would increase among nurse-visited low-income women in the Memphis and Denver trials. These increases in family income would lead not only reductions in public benefits, but increased tax revenues, including increases in income and social security taxes. While we considered analyzing the effect of the program on sales and consumption taxes, our preliminary estimates of the influence of the program on these forms of revenue revealed that these are likely to be so small as to be inconsequential in light of the larger set of government costs and revenues.
| Table of Contents | Previous | Next |