The construction of U.S. consumption data: some facts and their implications for empirical work.
To highlight data flaws that may affect the outcome of empirical work, a paper examined the sources and methods used to construct aggregate data on U.S. consumer spending. Two flaws are identified: sampling error and compositional error. Sampling error reflects the fact that monthly retail sales...
| Publicado en: | American Economic Review Vol. 82; pp. 922 - 942 |
|---|---|
| Autor principal: | |
| Formato: | Artículo |
| Publicado: |
American Economic Association
September 1992
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| Sumario: | To highlight data flaws that may affect the outcome of empirical work, a paper examined the sources and methods used to construct aggregate data on U.S. consumer spending. Two flaws are identified: sampling error and compositional error. Sampling error reflects the fact that monthly retail sales estimates are derived from a sample rather than the complete set of stores. This measurement error is significant because retail sales are an important component of estimates on personal consumption expenditures. Compositional error occurs because monthly data on the product composition of retail sales do not exist: Information on the product composition of sales is gathered by survey only once every 5 years. Several examples illustrate the empirical effects of sampling and compositional errors in U.S. consumption data. To remedy these errors, guidelines for empirical practice are suggested. |
|---|