tempagency-JOLE

This repository provides the codes to replicate Xinwei Dong, Dean Hyslop and Daiji Kawaguchi "Skill, Productivity and Wages: Direct Evidence from a Temporary Help Agency"

About data

The analysis uses confidential microdata provided by an anonymous Japanese temporary help agency. We are not permitted to share the firm-sourced data, including the original worker information datasets and sales records (.xlsx files started with "【提示用】") and the processed data (price.dta, info.dta, entry.dta and aptitude.dta). Researchers who are interested in replicating any of our results should take contact with Daiji Kawaguchi (kawaguchi@e.u-tokyo.ac.jp). The company agrees to give data access given the researcher is included in the joint research agreement between the company and the University of Tokyo and the researcher signs NDA with the company. Upon the approval of the company and the University of Tokyo, the researcher who are included in the joint research agreement can access to the data set from the designated office at the University of Tokyo only for the purpose of replicating the results reported in the paper.

The unemployment rate dataset regionunemp.dta is based on monthly Labor Force Survey: https://www.stat.go.jp/data/roudou/longtime/03roudou.html#hyo_1. The dataset on 2015-base monthly CPI for metropolitan areas regioncpi.dta is based on data provided by the Statistics Bureau of Japan: https://www.e-stat.go.jp/stat-search/files?page=1&layout=datalist&toukei=00200573&tstat=000001084976&cycle=0&tclass1=000001085995&tclass2=000001085937&tclass3val=0.

To run the code

1. IMPORT AND BASIC CLEANING

Put all the raw excel sheets provided by the firm in the current working directory, and run prepare.do to get 4 dta files.

2. CREATE THE SAMPLE

Put

• price.dta, info.dta, entry.dta, aptitude.dta and

• regionalunemp.dta and regionalcpi.dta

in the current working directory, and run clean.do in Stata to get the "full sample" and the "analysis sample".

3. TABLES AND FIGURES

After achieving the full sample and the analysis sample,

• Table 1-9 and Figure 1-5 can be generated by running tabfigs.do in Stata.

• Table 10 can be generated by running IRR_calculation.R after achieving fittedfw.dta from running tabfigs.do in R.