This study investigates whether community-based long-term care services can achieve net reductions in overall long-term care costs through offsetting nursing home use. While this issue is often seen as having been resolved in the negative by several major community care demonstrations that have failed to achieve net savings, such a verdict is both premature and incomplete. First, existing studies have considered only total or average effects on nursing home use and on costs, and have not expect in very ad hoc ways considered the issue of cost-savings at the level of the individual consumer of services. Second, effects of community service have been considered only as an aggregate treatment """"""""package"""""""", and the differential effects of different types of community service have been examined. In short, it has not been taken into account that different types and levels of community services may affect overall costs in different ways for individuals with different characteristics. Some services, in some combinations of levels, may be cost-effective for some, perhaps many, individuals even if aggregate costs rise. Our research will address these issues systematically, developing empirical models within which cost-effectiveness can be assessed at the individual level, and in the process hitting the """"""""service targeting"""""""" issue much more squarely on the head than has previous research. This will be done through discrete-time transition probability models which link projected nursing home use and overall long-term care costs for individuals to their use of community services. The coefficients from these models, will be sued to empirically assess total and marginal cost effects of different community services for a given individual. Then, for any individual (and hence for any set of individuals) the community service """"""""package"""""""" that minimizes their overall long-term care costs can be determined. This type of service analysis and planning at the individual client level, which is the essence of practical service targeting, is central to understanding both why community service demonstrations have not been more cost-effective, as well as how and to what extent their performance can be improved through more efficient management of resources. Because unconstrained cost-minimization may have adverse effects on the quality of life on individual clients, optimizations will be undertaken that pursue both unconstrained cost minimization, as well as ones which impose varying degrees of constraint on the optimization in terms of its impact on life quality. In this way the tradeoff between individual cost and life quality in optimally management long-term care systems can be made explicit.
