Association for Public Policy Analysis & Management

Prior studies have documented that single adults born between 1955 and 1964--now aged 54 to 63--have consistently faced a disproportionate risk of homelessness over the past 25 years. Indeed, persons aged 55 and above now comprise nearly one third of all homeless adults nationwide and there is likely to be substantial growth over the next decade in the number of older adults experiencing homelessness. Older homeless adults have medical ages that typically exceed their biological ages by more than 20 years, meaning that they are likely to be heavy users of health care services, and long-term care services, such as nursing homes, in particular. Absent an expansion of new housing opportunities for this cohort, the effects of premature morbidity and increases in the size of the older adult homeless population will result in pronounced increases in health care costs much of which will be borne by state Medicaid programs.

To understand the scope of this problem and explore policy solutions, in the present study, we link records from the emergency shelter system in a major northeastern city from the period from 2009-2015 for all 4,333 individuals shelter using adults aged 55 and above with state Medicaid claims data to: 1) Forecast both the expected future trajectory in the size of the older homeless adult population and their expected health care and emergency shelter costs; and 2) Model the potential cost offsets associated with policy interventions that would target a range of expanded housing options to this population. We use the cohort-component method to forecast changes in the size and age composition of the sheltered homeless population over the period from 2016 to 2030. We then apply age-group specific estimates of annual Medicaid and emergency shelter costs to our forecasts to arrive at population-level forecasts of health care and shelter costs for persons in this cohort for the period from 2016 to 2030, absent any intervention. We use cluster analysis procedures to identify housing and health care need subgroups in this population and estimate the impact of providing housing and supportive service interventions calibrated to the needs of these subgroups under three different cost reduction scenarios.

We find that the numbers of individuals aged 55 and 65 and above using emergency shelter on an annual basis are expected to increase by 45% and 230%, respectively, between 2016 and 2030. Total health care and shelter cost increases will be roughly proportional to the growth in the population of older homeless adults, although nursing home costs will rise by between 99% and 260%. We estimate that, depending on the scenario, expanding housing and supportive service opportunities to house all older homeless adults would result in between a 25% and 46% reduction in future health care and shelter costs for this population, enough to offset between 45% and 85% of the cost of these interventions themselves. Our findings highlight the potential benefit of developing and expanding new strategies for using Medicaid funds to help address the housing needs of older homeless adults.