Working Paper: NBER ID: w7650
Authors: Dora L. Costa
Abstract: I use micro data on food and recreation expenditures from 1888 to 1994 to provide the first estimates of overall CPI bias prior to the 1970s and new estimates of bias since the 1970s and to reassess long-run growth rates. I find that CPI bias was -0.1 percentage points per year between 1888 and 1919 and rose to 0.7 percentage points per year between 1919 and 1935. CPI bias was low in the 1950s and 0.3 percentage points per year in the 1960s and then rose to 2.7 percentage points per year between 1973 and 1982 before falling to 0.6 percentage points per year between 1983 and 1994. Inadequately accounting for the introduction of new consumer goods probably was the biggest source of bias between 1919 and 1935. Revised growth rates suggest that despite the Great Depression real per capita personal income was not falling but was rising by 0.5 percentage points per year between 1919 and 1935 and that growth rates were not stagnant in the 1970s but were almost as high as in the 1960s (4.0 and 3.2 in the 1960s and 1970s, respectively).
Keywords: No keywords provided
JEL Codes: D12; N11; N12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| CPI bias (C43) | real income growth (O49) |
| introduction of new consumer goods (F61) | CPI bias (C43) |
| real per capita personal income (R19) | CPI (E31) |
| CPI bias (C43) | growth rates in the 1970s (O49) |
| demographic shifts and changes in consumer behavior (R20) | identification of causal effects (C22) |