Data Set and Online Appendix for: Josefine Quast and Maik H. Wolters, "The Federal Reserve’s Output Gap: The Unreliability of Real-Time Reliability Tests",
Journal of Applied Econometrics, forthcoming.
The directory contains the online appendix and 7 data sets.
1.) Online Appendix
File Name: Appendix-QW.pdf
This is the online appendix associated with the paper.
Appendix A: Replication in a Narrow Sense
Appendix B: Details on Drawing Inference
Appendix C: Summary Statistics for Longer and Updated Samples
Appendix D: Updated NSRs from Statistical Output Gap Estimates
Appendix E: Results for Different Output Gap Projection Horizons
Appendix F: Inference and Different Block Length Settings
2.) Datasets
2a) File name: EdgeRudd2016_gbgap_all.xlsx.
This is the data set on Greenbook/Tealbook output gap vintages used in Edge, R. M., & Rudd, J. B. (2016). Real-time properties of the Federal Reserve's output gap. Review of Economics and Statistics, 98(4), 785-791.
* column A: date.
* column B: end-point series.
* column C - column BB: GB_GAP_GDP94Q3 - GB_GAP_GDP08Q3 -- Greenbook/Tealbook output gap vintages from 1994Q3 to 2008Q3.
For Edge and Rudd's (2016) replication material, see here: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/LQGQAE.
2b) File name: b_vintages.xlsx.
This is an updated version of Edge and Rudd's (2016) original data set in which we maintain the same structure as in their original data file.
Note that the info sheet in the Excel file provides extra information on how we spline data based on Edge and Rudd's (2016) file for some vintages.
* column A: date.
* column B: end-point series.
* column C: Greenbook output gap vintage as of 1994Q3.
* column D - column CI: GBgap_960516 - GBgap_161207 -- Greenbook/Tealbook output gap vintages from 1996Q1 to 2016Q4.
We obtain the updated series from the "Tealbook (formerly Greenbook) Data Sets", specifically "Output Gap and Financial Assumptions from the Board of Governors":
https://www.philadelphiafed.org/surveys-and-data/real-time-data-research/gap-and-financial-data-set
Note that the Tealbook Data Set contains more output gap vintages than used in Edge and Rudd (2016) and therewith Quast and Wolters, forthcoming since #FOMC meetings per year > #yearly quarters.
We follow Edge and Rudd's (2016) logic to expand the data set. Note also that we maintain the Philadelphia Fed's variable labeling to increase tractability for future updates.
2c) File Name: Greenbook_Output_Gap_DH_Web.xlsx.
For transparent documentation, we include the Tealbook Data Set obtained from the Philadelphia Fed's real-time data sets we used as of May 2022.
* column A: date.
* column B - column FI: GBgap_960321 - GBgap_161207 -- Greenbook/Tealbook output gap vintages.
Source: https://www.philadelphiafed.org/surveys-and-data/real-time-data-research/gap-and-financial-data-set.
2d) File Name: MixedFinalVintage.xlsx.
This sheet contains the series used in section 2.3 "Which Vintage To Use as the Final Output Gap Estimate?" of the paper.
* column A: date.
* column B: Mixed final vintage.
We compose the mixed final vintage as follows:
The sample from 1980Q1 - 1992Q4 is assessed against the GB_GAP_GDP94Q3 Greenbook output gap vintage.
The sample from 1994Q1 - 2001Q4 is assessed against the August 2002 Greenbook output gap vintage.
The sample from 2002Q1 - 2009Q3 is assessed against the March 2012 Tealbook output gap vintage.
The sample from 2009Q4 - 2015Q3 is assessed against the September 2015 Tealbook output gap vintage.
2e) File Name: EdgeRudd2016_gdp6679.xlsx.
Data sets on real output (in levels) used for the statistical trend-cycle decompositions. For completeness, we include Edge and Rudd's (2016) data sets for the real GDP vintages between 1966 and 1979, which we used as well.
EdgeRudd2016_gdp6679
* column A: date.
* column B - column BD: GDP66Q1 - GDP79Q3 -- real GDP vintages.
For Edge and Rudd's (2016) replication material, see here: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/LQGQAE.
2f) File Name: EdgeRudd2016_gdp8094.xlsx.
Data sets on real output (in levels) used for the statistical trend-cycle decompositions. For completeness, we include Edge and Rudd's (2016) data sets for the real GDP vintages between 1979 and 1994, which we used as well.
* column A: date.
* column B - column BJ: GDP79Q4 - GDP94Q4 -- real GDP vintages.
For Edge and Rudd's (2016) replication material, see here: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/LQGQAE.
2g) File Name: gdp_allvintages.xlsx.
Real GDP vintages starting since 1991 are available from Alfred, which we use to update the data set.
The data set used in the paper that combines all three sources is gdp_allvintages.
* column A: date.
* column B - column CY: GDP66Q1 - GDP91Q2 -- real GDP vintages from Edge and Rudd's (2016) material.
* column CZ - column HI: GDPC1_19911220 - GDPC1_20200326 -- real GDP vintages from Alfred.
Source: https://alfred.stlouisfed.org/series?seid=GDPC1.
Note that an info sheet in the Excel file contains further information on how we spline the series for particular vintages.
