Impact of the coronavirus (COVID-19) pandemic on the Producer Price Index data for September 2020

The Producer Price Index (PPI) is a measure of the average change over time in the prices domestic producers receive for the sale of their products. PPI estimates for September 2020 were published on October 14, 2020. The pricing date was Tuesday, September 15, 2020, which means that most producers, but not all, provided prices they received on that date. The summary statistics in tables 1, 2, 3 and 4 indicate the impact of the COVID-19 pandemic on the PPI data collection, index estimation, and the BLS ability to publish indexes was minor for the release of preliminary September data. No changes in estimation procedures were necessary. 

This page provides further information on the effects of the COVID-19 pandemic for the September 2020 Producer Price Indexes news release.

Collection mode

The PPI collects nearly all of its repricing data by web collection. There were no changes to collection mode in response to the pandemic. 

Response rates

Table 1 provides the September 2020 response rate by industry sector, along with comparison periods of August 2020, September 2019, and a 12-month average for March 2019 to February 2020. The September 2020 response rate for items requested shows a similar level of response for the vast majority of industry sectors. While the response rate for entertainment, accommodation, and food services sector remains down when compared to historical comparison periods, both the September and August 2020 response rates reflect an uptick in response as businesses begin to reopen. 

It is important to note that typically, the PPI survey’s monthly price collection response rate averages 76 percent by the end of its 4-month index revision period. All PPI indexes are recalculated 4 months after publication of preliminary data. Thus, September indexes may be revised when BLS publishes the PPI for January on February 17, 2021. All summary statistics in the tables below reflect comparisons to PPI preliminary data. 

Imputation

When a respondent does not provide a price for an item, BLS employs an imputation method of replacing that missing price with a statistically-estimated value. Missing prices are imputed by the prices of similar items that have been collected. Essentially, the price movement of items that are not collected are estimated to be the same as those in their general index category. Table 2 indicates that the September 2020 item imputation rate of 39 percent reflects a slight decrease from the preceding month and is in-line with the comparison periods. 

Cell, series, index, or observation suppression

BLS calculates over 10,000 PPIs for individual products and groups of products for publication during the release of preliminary data. The first row of Table 3 shows the September 2020 publication rate of 89 percent is consistent with the comparison periods. BLS suppresses indexes when they fail to meet publication criteria. Because of additional collection efforts along with late-arriving data, these suppressed indexes often will meet publication criteria at the end of the 4-month index revision period, at which time they are published. (Upon request, BLS will provide a list of indexes that did not meet publication criteria for the release of preliminary PPI data. Please email requests to ppi-info@bls.gov.) 

Further illustrating the nominal impact of the COVID-19 pandemic, the second row of table 3 shows the percent of industry indexes that BLS was unable to calculate because BLS did not receive any prices for that index. This proportion of not calculated indexes is similar to the normal result. 

Table 4 takes a closer look at the not calculated PPI indexes by industry and reflects that across all industry sectors there is no discernable difference between September’s result when compared with earlier periods. 

Last Modified Date: October 14, 2020