Informing businesses of policymakers' views on inflation uncertainty

How to cite

Meyer, Brent; Parker, Nicholas; Foster, Kevin. 2021. "Informing businesses of policymakers' views on inflation uncertainty." AEA RCT Registry. AEARCTR-0008741. https://doi.org/10.1257/rct.8741-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Brent Meyer* , Nicholas Parker , Kevin Foster

* Primary investigator

Registry ID
AEARCTR-0008741
DOI
https://doi.org/10.1257/rct.8741-1.0
First registered
Last updated
Status
Completed
Trial dates
2020-10-05 – 2020-11-13
Countries
United States of America (6th Federal Reserve District)
JEL codes
E6, E31, E52, L2
Registry record
http://www.socialscienceregistry.org/trials/8741

Abstract

In October 2020, we asked firms for their highest and lowest potential expectations for PCE inflation in 2021. In this experiment, we wanted to test whether giving business decision makers information on the uncertainty (in the form of confidence intervals) around monetary policymakers’ projections influenced their inflation projections in October, and to see if this information impacted firms’ own unit cost expectations over the year ahead a month later in November 2020. To make sure that the information we were supplying was about the second moment only, we provided both the control and the treatment groups with the median expectation for PCE inflation over calendar year 2021 from the FOMC’s Summary of Economic Projections on September 15, 2020. For the treatment group we provided the 70 percent confidence interval around those projections based on the historical forecast errors of professionals (excerpted from the minutes of the September 2020 FOMC meeting).

Intervention

Target group: C-suite executives and business owners (panelists) in the Atlanta Fed's Business Inflation Expectations Survey.

In October 2020, we asked firms for their highest and lowest potential expectations for PCE
inflation in 2021. In this experiment, we wanted to test whether giving business decision makers
information on the uncertainty (in the form of confidence intervals) around monetary policymakers’
projections influenced their inflation projections in October, and to see if this information impacted
firms’ own unit cost expectations over the year ahead a month later in November 2020. To make
sure that the information we were supplying was about the second moment only, we provided both
the control and the treatment groups with the median expectation for PCE inflation over calendar
year 2021 from the FOMC’s Summary of Economic Projections on September 15, 2020. For the
treatment group we provided the 70 percent confidence interval around those projections based on
the historical forecast errors of professionals (excerpted from the minutes of the September 2020
FOMC meeting).

Specifically, we asked
"The median expectation of monetary policymakers for inflation over calendar year 2021 was 1.7 percent (as of September 15th). [Treatment: Based on forecasts over the past
20 years, there is a 70 percent chance that actual inflation will be in the range of 0.7
percent to 2.7 percent over calendar year 2021.] What is your best estimate for the
highest and lowest potential rate of inflation over calendar year 2021?"

The treatment and supplied information came from the supporting documentation from the September 2020 FOMC meeting, which had concluded just weeks prior to the beginning of our survey period.

In the attached documentation, we show that firms' expectations and uncertainty for PCE inflation in 2021 and, importantly, for firms' unit cost expectations and uncertainty were largely unaffected.

Experimental design

The experimental design is a simple RCT. Respondents to the October 2020 BIE survey are chosen at random to receive treatment. There are only two groups: treatment and control.

The target sample group is between 200-225 respondents.

Design details

Randomization method
Randomization done in office by computer using Qualtrics survey software.
Randomization unit
Firm-level (respondents are c-suite execs, business owners, or financial professionals).
Clusters
Approximately 200-225 firms.
Observations
200-225
Arms
100-113 firms in both the control and treatment group
Primary outcomes
The key outcome variables of interest are firms' 1-year ahead unit cost expectations and uncertainty, and firms' expectations and confidence interval for year-ahead PCE inflation.

Cited by

No citation data available for this registration yet.

Keywords