TY - JOUR AU - Griffith,Rachel AU - Leibtag,Ephraim AU - Leicester,Andrew AU - Nevo,Aviv TI - Timing and Quantity of Consumer Purchases and the Consumer Price Index JF - National Bureau of Economic Research Working Paper Series VL - No. 14433 PY - 2008 Y2 - October 2008 UR - http://www.nber.org/papers/w14433 L1 - http://www.nber.org/papers/w14433.pdf N1 - Author contact info: Rachel Griffith IFS 7 Ridgmont Street London WC1E 7AE UK E-Mail: rgriffith@ifs.org.uk Ephraim Leibtag U.S. Department of Agriculture FRED-ERS-USDA 1800 M Street, NW Rm. N2137 Washingron, DC 20036 Tel: 202-694-5349 Fax: 202-694-5662 E-Mail: eleibtag@ers.usda.gov Andrew Leicester Institute for Fiscal Studies 7 Ridgmount Steet London WC1E 7AE, UK E-Mail: andrew_l@ifs.org.uk Aviv Nevo Department of Economics Northwestern University 2001 Sheridan Road Evanston, IL 60208-2600 Tel: 847/491-8212 Fax: 847/491-7001 E-Mail: nevo@northwestern.edu AB - A common approach to measuring price changes is to look at the change of the expenditure needed to purchase a fixed basket of goods. It is well-known that this approach suffers from problems and creates several biases in the measurement of price changes faced by consumers. Substitution and outlet bias, two commonly studied concerns, are both driven by consumer choices of what and where to buy. However, consumers also make other choices, including how much and when to buy. We discuss the implications of consumers' timing and quantity decisions have on standard practices of computing of computing a price index. We use household-level data on quantities purchased and prices paid to construct a measure of the savings made by consumers' optimizing behaviour in the purchase of food. In particular, we compare the prices actually paid by the consumers to various alternatives that do not allow for substitution. Our analysis suggests that the average consumer makes significant, and comparable in magnitude, savings from the four dimensions of choice that we study. Furthermore, our data suggests significant heterogeneity in consumer behavior, and that this behavior is correlated with demographics. Our findings suggest that ignoring timing and quantity decisions, when computing a price index, can generate biases on the order of magnitude of substitution and outlet biases. ER -