Linking the Regional Demographic Processes and the Small Area Housing Growth:

The population and demographic change of small areas is mostly driven by both the regional economic-demographic influence and small area demographic processes in a spatio-temporal context. The popular cohort-component method is not easily applicable to population projections of small area(e.g.,city or censustracts) for a couple of reasons: (1) the historical and current trends in vital statistics and migration of the small areas are not easily available; (2) it is difficult to independently develop reasonable migration assumptions of small areas. An alternative approach is to derive population projections of local communities through the enhanced link age of small area housing growth and regional demographic processes. This study presents a modeling approach toward developing the long term projection of total population and the key demographic characteristics(e.g.,age,race / ethnicity) at the transportation analysis zone level. A proposed small area modeling approach is as follows: (1)project the regional employment, population, and household growth using the employment, demographic, and household projections models; (2) allocate the regional housing and employment growth into the small area; (3) convert the small area housing growth into population growth using the housing unit method; (4) disaggregate the small area population into their demographic characteristics. This study focuses on the fourth stage of projecting the small area demographic characteristics and presents the multi-nomial logit regression method to project the small area demographic characteristics utilizing the past trend of those demographic components of population at the small area. The study finds the proposed method as reasonable, and suggests that the topic requires the further research.