Testing a forecast model to predict movement of older Americans into retirement housing

Purpose – Expectations of an aging population increases the need for accurate models and reliable forecasts of housing demand in the later stages of life. This paper aims to examine whether the only published forecasting model predicts the actual movement of older Americans into retirement housing. Design/methodology/approach – A logistic regression model is used to test which older movers chose retirement housing using data from the 2002 and 2004 Health and Retirement Survey. Findings – Age is found to have a positive relationship with the choice to move into retirement housing, but the other socioeconomic variables are not significant and the model is not robust. Research limitations/implications – More investigation is necessary to identify the variables that will provide a reliable estimate of those choosing retirement housing. Segmented by housing type may be necessary rather than estimating total retirement housing demand. Originality/value – The paper tests whether a model previously used to forecast senior housing demand actually predicts housing choice. It demonstrates that additional research is needed to develop models that can more accurately estimate demand in this growing segment of the housing market in aging countries.