Maik H. Wolters, "How the Baby Boomers' Retirement Wave Distorts
Model-Based Output Gap Estimates", Journal of Applied Econometrics,
Vol. 33, No. 5, 2018, pp. 680-689.

The directory contains the online appendix and two datasets.

Each of the datasets comes in two versions, .xlsx and .csv. The .csv
files are zipped in mw-csv-files.zip, and the .xlsx files are not
zipped because they are already in a compressed format. Since the .csv
files are in DOS format, Unix/Linux users should use "unzip -a".


1.) Appendix-MW.pdf: This is the online appendix of the paper.

Appendix A: Data Sources
Appendix B: Model Equations
Appendix C: Estimated Parameters
Appendix D: Business Cycle Moments of hours Series


2.) DataDSGE: Contains all data used for the estimation of the DSGE
model in the paper.

a) csv-Version: DataDSGE.csv (contains the observables used in the
measurement equations)

b) Excel-Version: DataDSGE.xlsx (contains the raw data, the data
transformations and the observables used in the measurement equations)

The observables are:

* delta_y_t_obs: annualized quarterly real per capita output growth
* delta_c_t_obs: annualized quarterly real per capita personal consumption
                 spending growth	
* delta_i_t_obs: annualized quarterly real per capita investment growth	
* infl_t_obs:    annualized quarterly GDP-deflator inflation
* wage_t_obs:    annualized quarterly real wage growth
* int_t_obs:     federal funds rate
* spread_t_obs:  BAA-10yr-TBill spread
* hours1_t_obs:  average weekly hours in the nonfarm business sector (as in 
                 Smets/Wouters, 2007)
* hours2_t_obs:  demeaned hours per capita in the nonfarm business sector
                 (not used in the paper)
* hours3_t_obs:  demeaned hours per capita in all sectors
* hours4_t_obs:  demeaned hours per capita in all sectors with demographical
                 adjustment (recommended hours measure)
* hours5_t_obs:  demeaned hours per capita in the private business sector
* intexp1_t_obs: interest rate expectations one quarter ahead
* intexp2_t_obs: interest rate expectations two quarters ahead
* intexp3_t_obs: interest rate expectations three quarters ahead
* intexp4_t_obs: interest rate expectations four quarters ahead


3.) Francis-Ramey_JMCB_Data_09_ReplicationUpdate: replicates and
updates the demographic adjustment procedure of total hours of
Francis and Ramey (2009, JMCB). 

a) csv-Version: Francis-Ramey_JMCB_Data_09_ReplicationUpdate.csv

b) Excel-Version: Francis-Ramey_JMCB_Data_09_ReplicationUpdate.xlsx

* column A: date.
* column B: average weekly hours purged of demographics. This series equals
            the one by Francis and Ramey until 2007 (column D) and the update
	    from 2007 onwards (column E). Both series are connected via a
	    level adjustment factor of the updated series.
* column C: average weekly hours unadjusted. Total hours worked (column G)
            divided by the noninstitutional population 16+ (column F) times
	    1/52.
* column D: Original Series from Francis and Ramey. Contains the original
            demographically adjusted hours series from Francis and Ramey
	    (2009, JMCB).
* column E: Replication of the work of Francis and Ramey. Contains the
            replication of the series by Francis and Ramey (2009) and an
	    update after 2007. This is computed by using the unadjusted
	    series (column C) and subtracting the cumulated demographic
	    adjustment factor (column I) and a constant to achieve the same
	    mean as in the unadjusted series.
* column F: Noninstitutional population 16+.
* column G: Total hours worked.
* column H: Chain weighted change in average weekly hours due to demographics.
            This is the demographic adjustment factor as shown in equation (7)
	    in the paper based on population shares by age group (columns J-Q)
	    and average hours by age group (columns R-Y).
* column I: Cumulation of previous column.
* columns J-Q: frpop16_17, frpop18_21, ... , frpop65. Population shares
               by age group.
* columns R-Y: hr16_17, hr18_21, ... , hr65. Average hours worked by age group.


Maik Wolters
University of Jena
maik.wolters@uni-jena.de