Abstract
This paper examines the accuracy of the home purchase component of the Consumer Price Index. Alternative monthly index series are simulated over the period 1973–1978 using FHA Master Statistical File data. Hedonic indexes derived from monthly regressions are compared to series constructed by the current CPI index method, which utilizes only age and size of house in correcting for intertemporal quality change. Of particular interest is the BLS practice of “linking out” price changes occurring when FHA raises its ceiling on the size of insurable mortgages. This procedure may have led to a significant downward bias in the official index.

This publication has 8 references indexed in Scilit: