Working Paper: NBER ID: w26424
Authors: Victor Chernozhukov; Jerry A. Hausman; Whitney K. Newey
Abstract: From its inception, demand estimation has faced the problem of "many prices." This paper provides estimators of average demand and associated bounds on exact consumer surplus when there are many prices in cross-section or panel data. For cross-section data we provide a debiased machine learner of consumer surplus bounds that allows for general heterogeneity and solves the "zeros problem" of demand. For panel data we provide bias corrected, ridge regularized estimators of average coefficients and consumer surplus bounds. In scanner data we find smaller panel elasticities than cross-section and that soda price increases are regressive.
Keywords: Demand Estimation; Consumer Surplus; Machine Learning; Panel Data; Econometrics
JEL Codes: C13; C14; C21; C23; C55; D12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| presence of many prices (E30) | omitted variable problem (C20) |
| omitted variable problem (C20) | biased estimates of demand systems (C51) |
| Hicks-Leontief composite commodity theorem (F11) | treat groups of commodities as a single good (Q02) |
| machine learning methods (C45) | address dimensionality problem (C23) |
| machine learning methods (C45) | estimate average demand (C51) |
| double-debiased machine learning (DML) approach (C51) | root-consistent estimates of consumer surplus bounds (D11) |
| double-debiased machine learning (DML) approach (C51) | approximately unbiased and normally distributed estimates (C13) |
| methodology (B41) | more accurate estimation of average welfare effects from price changes (D69) |
| panel elasticities (H30) | smaller than cross-section estimates (C21) |
| inclusion of zero expenditure values (P44) | more comprehensive understanding of consumer behavior (D19) |