Building statistics from the ground up
Traditional methods of collecting data from businesses and households face increasing challenges. These include declining response rates to surveys, increasing costs to traditional modes of data collection, and the difficulty of keeping pace with rapid changes in the economy. The digitization of virtually all market transactions offers the potential for re-engineering key national economic indicators. The challenge for the statistical system is how to operate in this data-rich environment. This paper focuses on the opportunities for collecting item-level data at the source and constructing key indicators using measurement methods consistent with such a data infrastructure. Ubiquitous digitization of transactions allows price and quantity be collected or aggregated simultaneously at the source. This new architecture for economic statistics creates challenges arising from the rapid change in items sold. The paper explores some recently proposed techniques for estimating price and quantity indices in large-scale item-level data.
Re-engineering Key National Economic Indicators
Authors: Gabriel Ehrlich, John Haltiwanger, Ron Jarmin, David Johnson and Matthew D. Shapiro
From: University of Michigan, University of Maryland, U.S. Census Bureau
Seasonally adjusting scanner data
Large volumes of highly heterogeneous data are increasingly available, but they are often not immediately useful for economic analysis without removing some nuisance variations and performing some form of aggregation. In this paper, the nuisance variations in question are the seasonal and holiday effects. As they cannot be adequately removed by conventional procedures, the adjusted data continue to exhibit seasonal patterns when aggregated over counties. Guha and Ng propose to augment univariate seasonal adjustments with a machine-learning step that pools information across US counties.
A Machine Learning Analysis of Seasonal and Cyclical Sales in Weekly Scanner Data
Authors: Rishab Guha, Serena Ng
From: Harvard University, Columbia University
New goods, scanner data and price indexes
Determining how to incorporate new goods into the calculation of price indexes is an important, unresolved issue for statistical agencies. That issue becomes particularly important with the increased availability of scanner data to measure prices and quantities, because new and disappearing products at the barcode level occur frequently in such data. In this paper Diewert and Feenstra compare several empirical methods to deal with new and disappearing products, and illustrate their results using the barcode data for frozen juice from one grocery store.
Estimating the Benefits of New Products
Authors: W. Erwin Diewert, Robert C. Feenstra
From: University of British Columbia, University of California, Davis