« Back to Results

Recent Research on Price Index Topics

Paper Session

Sunday, Jan. 7, 2024 8:00 AM - 10:00 AM (CST)

Marriott Rivercenter, Conference Room 5
Hosted By: Society of Government Economists
  • Chair: David M. Byrne, Federal Reserve Board

Alternative Data Sources and Hedonic Price Indexes for High-Tech Products in the U.S. Consumer Price Index

Craig Brown
,
U.S. Bureau of Labor Statistics
Jeremy Smucker
,
U.S. Bureau of Labor Statistics

Abstract

The Bureau of Labor Statistics has a goal of transitioning a portion of the U.S. Consumer Price Index market basket from traditional data collection to non-traditional or alternative data sources and collection modes. This paper explores the use of alternative data sources and methods for adjusting price indexes for quality change. We combine web-collected and household survey data to estimate hedonic regression models. We examine price estimation methods that consider the valuation of both observable and unobservable characteristics at the time of product turnover. A superlative index formula is used in the estimation of experimental price indexes for wireless telephone services and televisions.

Extending the Range of Robust PCE Inflation Measures

Dominic A. Smith
,
U.S. Bureau of Labor Statistics
Sergio Ocampo
,
Western University
Raphael Schoenle
,
Brandeis University

Abstract

We evaluate the forecasting performance of a wide set of robust inflation measures between 1960 and 2022, including official median and trimmed-mean personal-consumption-expenditure inflation. When trimming out different expenditure categories with the highest and lowest inflation rates, we find that the optimal trim points vary widely across time and also depend on the choice of target; optimal trims are higher when targeting future trend inflation or for a 1970s-1980s subsample. Surprisingly, there are no grounds to select a single series on the basis of forecasting performance. A wide range of trims-including those of the official robust measures-have an average prediction error that makes them statistically indistinguishable from the best-performing trim. Despite indistinguishable average errors, these trims imply different predictions for trend inflation in any given month, within a range of 0.5 to 1 percentage points, suggesting the use of a set of near-optimal trims.

Using Rolling-Window Multilateral Price Indexes to Track Food Costs across Space and over Time

Chen Zhen
,
University of Georgia
Mary Muth
,
Research Triangle Institute
Megan Sweitzer
,
USDA Economic Research Service
Abigail Okrent
,
USDA Economic Research Service
Anne Byrne
,
USDA Economic Research Service

Abstract

Food price data are instrumental for studying food markets, the consumer food environment, and consumer welfare. The BLS Consumer Price Index (CPI) has been the gold standard for price information but has limited use in food and nutrition policy because (1) the CPI only compares price variation across time and many analyses require price variation across time and geographies; (2) foods in the food at home (FAH) CPI are aggregated into categories that are often too coarse for nutrition analyses. We describe a new data product called the Food-at-Home Monthly Area Prices data product (F-MAP), which provides price information that varies temporally and spatially for foods categorized into 90 categories that closely align with the Dietary Guidelines for Americans. The F-MAP includes two classes of price indices: four bilateral price indexes (Laspeyres, Paasche, Törnqvist, and Fisher Ideal) and two multilateral price indexes (GEKS and CCD). We compare the short-run relationships between the F-MAP to CPI FAH indexes at the national level, by Census Region, and for select major FAH subcategories. We find that using retail scanner data in constructing price indices puts downward pressure on measurement of food price inflation compared to the CPI, though effects vary by region and FAH subcategory.

Using Constant-Quality Price Indexes to Deflate Spending in the National Accounts

Ana Aizcorbe
,
U.S. Bureau of Economic Analysis
Daniel Ripperger-Suhler
,
U.S. Bureau of Economic Analysis

Abstract

The fundamental role of price indexes in national accounts is to decompose changes in nominal spending into a piece that measures changes in prices—called inflation—and another that measures the growth in “quantities” or real Gross Domestic Product. In a static economy with a fixed set of goods whose quality remains constant, the inflation piece of this decomposition would measure changes in the prices of identical goods and the implied real GDP component would measure only changes in the number of goods produced. However, product innovation that brings new and better goods to market is the rule rather than the exception and properly accounting for these changes in the “quality” of goods when measuring inflation remains a first-order challenge in price measurement.

We contribute to the vast literature on quality change by looking at this issue from a different lens: while existing studies use the cost-of-living index paradigm tied to utility maximization as their benchmark, our criteria is the ability of different methods to provide a clean decompositions of spending: Does a price index method yield an inflation measure that holds quality constant and an implied measure of real GDP that holds prices constant? Looking at both sides of the GDP decomposition allows us to (1) provide an algebraic expression for the elusive concept of “quality change” that we call a quality index, and (2) use properties of those quality indexes as a tool to assess which price index methods are best suited for the national accounting decomposition. Along the way, we also provide an intuitive explanation of the chain drift problem and an assessment of the potential usefulness of methods that were designed to ameliorate that problem to serve as deflators.

We use scanner data on consumer electronics to draw conclusions about

Discussant(s)
David M. Byrne
,
Federal Reserve Board
Adam Hale Shapiro
,
Federal Reserve Bank of San Francisco
Metin Cakir
,
University of Minnesota
Bart Hobijn
,
Federal Reserve Bank of Chicago
JEL Classifications
  • E3 - Prices, Business Fluctuations, and Cycles
  • C4 - Econometric and Statistical Methods: Special Topics