A Note on Geographic Living Cost Differentials
利用佛罗里达州县级六类商品价格数据，检验城市规模与生活成本的关系，发现人口规模降低生活成本而人口密度推高成本，但模型对分项预算的解释力有限。
Recent papers have found that there is a systematic relationship between urban size and geographic living cost differentials (Cebula 1980 and 1984; Hogan 1984). These papers have analyzed Bureau of Labor Statistics' Urban Family Budgets for 37 metropolitan areas because they are one data source that provides required geographically comparable data. This paper provides further evidence on this subject by offering new empirical evidence using county-based cost of living data. A unique data set on prices of six commodity groups for Florida counties makes this analysis possible. Cebula (1980) predicts a negative relationship between population size and cost of living because of agglomeration economies that are associated with lower costs of production. Population density was assumed to exert upward pressure on cost of living because congestion would lead to greater transportation and marketing costs. Cebula includes in model SMSA per capita income, business property taxes, and a dummy variable indicating if SMSA was located in a right-towork state. Population size has predicted negative coefficient while that of population density was positive. Hogan (1984) questioned validity of these results because Cebula's model does not satisfactorily explain disaggregated budget categories. Hogan tested Cebula's model for twelve individual budget categories, concluding the shelter component was only major consumption category with relatively large differences for which Cebula's model was able to explain more than half variance. In his reply, Cebula (1984) argued