Presented by: Alexis Antoniades, Georgetown University in Qatar
with Robert Feenstra, University of California
National Institute of Economic and Social Research, 2 Dean Trench Street, Smith Square, London SW1P 3HE
Online price data provide a new and rich source of information. But in the absence of information on quantities or on expenditure, researchers average prices and price behaviour across all products within a product group. In doing so, they introduce significant measurement error and potential bias. In this paper we address this challenge by presenting a simple methodological innovation that allows researchers to obtain information on expenditure when expenditure is not known. Specifically, we argue that measures of retail distribution – which can be computed solely from prices – provide a good proxy for expenditure.
Through a series of simulations that use scanner-level price and quantity information on about 85% of the Fast Moving Consumer Goods (FMCGs) sold in the GCC countries, we show that treating all products equally introduces substantial measurement error and bias in the calculation of price stickiness, of inflation, and of international price differences. But we also show that adding information on retail distribution reduces measurement error by 71%, 73%, and 75%, in the above estimations, respectively. Our findings also have important implications for the work of the International Comparison Program (ICP).
Alexis Antoniades is an Associate Professor and Director of International Economics at Georgetown University in Qatar. His work focuses on international economics and micro-level data. A Fulbright/CASP scholar, he holds a Ph.D. in Economics from Columbia University. Between 2012 and 2013 he was a Niehaus Fellow at Princeton University and between 2001 and 2002, an Assistant Economist at the Federal Reserve Bank of New York.