Shigeru Fujita, "Dynamics of Worker Flows and Vacancies: Evidence from 
the Sign Restriction Approach", Journal of Applied Econometrics, Vol. 
26, No. 1, 2011, pp. 89-121.

The data used in this paper are stored in three CSV files, which are ASCII
files in DOS format. These files are zipped in the file sf-data.zip.
Unix/Linux users should use "unzip -a".

1. benchmark.csv

This file contains the following 5 quarterly series, covering 1976Q1 through
2005Q4 (a total of 120 observations for each series). 
 
Column 1. Job separation hazard rate: Based on the CPS micro data.
          See Fujita and Ramey (2006) for data construction details. 
Column 2. Job finding hazard rate: Based on the CPS micro data.
          See Fujita and Ramey (2006) for data construction details.
Column 3. log (job vacancies): The index of help-wanted advertisements
          released by the Conference Board.
Column 4. Labor productivity growth: log changes in output per hour in the
          nonfarm business sector released by the BLS. 
Column 5. Inflation rate: annualized log changes in GDP deflator. 

2. disaggregate.csv

This file contains the following 13 quarterly series covering 1976Q1 through
2005Q4. The first 12 columns give the job separation hazard rates and job
finding hazard rates for 6 demographic groups, ordered as: 

Columns  1- 2: Young Male
Columns  2- 4: Young Female 
Columns  5- 6: Prime-age Male 
Columns  7- 8: Prime-age Female
Columns  9-10: Old Male
Columns 11-12: Old Female

The last (Column 13) is log(job vacancies) as mentioned above. Note that the
hazard rates in this file incorporate the out-of-the-labor-force flows as
explained in the paper. 

3. weights.csv

This file contains the labor force weights (6 numbers) used to aggregate
worker flows of 6 groups of workers. The order of the weights are the same
as the order used in the previous data file disaggregate.csv.