Replicate data for "Digital economic development and graduate labor demand: Evidence from online job postings in China"
Description
This study tests whether the development of the digital economy improves employment quality by increasing posted wages and the demand for high-skilled workers. The data combine 1,093,948 online job postings with city-level socioeconomic indicators and official statistical yearbook data for 294 Chinese cities from 2016 to 2022. Each job-posting observation contains wage information, employer and job characteristics, education and experience requirements, and occupation classifications based on job titles and descriptions. The empirical results indicate that digital-economy development is associated with higher posted wages and a greater share of high-skilled job opportunities. The study further examines instrumental-variable estimates, alternative specifications, education-related heterogeneity, skill-requirement mechanisms, external consistency with official indicators, and pandemic-period heterogeneity. The replication package contains the source data and Stata scripts required to construct the analysis sample and reproduce Figures 1–2 and Tables 1–12. Users should first set the project path in 0maindo.do, run 1data_cleaning.do, and then execute the remaining scripts in the order documented in the README. The data should be interpreted primarily at the job-posting level, while digital-economy and other socioeconomic measures vary at the city-year level.
Files
Steps to reproduce
1. Download and extract the replication package while preserving its directory structure. The package contains online job-posting data, city-year socioeconomic indicators, and official statistical yearbook data. The job-posting source data are divided into ten sequential files, raw_sample_part1.dta through raw_sample_part10.dta, to facilitate uploading and downloading. All ten files must remain in the rawdata folder with their original filenames. The workdata folder is initially empty and is populated by the Stata scripts. 2. Use a recent 64-bit version of Stata. Install the required community-contributed packages if they are not already available: ssc install estout ssc install winsor2 ssc install ftools ssc install reghdfe ssc install ivreg2 ssc install ranktest ssc install ivreghdfe 3. Open dofile/0maindo.do and replace global path with the absolute path to the extracted replicate data folder. Run 0maindo.do to define the paths, control variables, and random-number seed. Keep the same Stata session open for all subsequent steps. 4. Construct the analysis sample by running: do "$dofile\1data_cleaning.do" The script loads raw_sample_part1.dta, appends raw_sample_part2.dta through raw_sample_part10.dta in sequence, and verifies that the combined data contain 1,093,948 observations. It then winsorizes wages at the 1st and 99th percentiles, converts nominal wages into real wages using the CPI, constructs log-wage and high-skill-job variables, applies the classification-confidence threshold, and creates workdata/main_sample.dta. 5. Run the remaining scripts in the following order: do "$dofile\figure1.do" do "$dofile\2table1_summary.do" do "$dofile\3table2_base_wage.do" do "$dofile\4table3_base_hs.do" do "$dofile\5table4_robust.do" do "$dofile\6table5_robust.do" do "$dofile\7table6_iv.do" do "$dofile\8table7_iv_exclusion.do" do "$dofile\9table8_heterogeneity.do" do "$dofile\10table9_heterogeneity_checks.do" do "$dofile\11_1skill_count.do" do "$dofile\11_2table10_mechanism.do" do "$dofile\12_1_external_consistency.do" do "$dofile\12_2_figure2.do" do "$dofile\12_3_table11.do" do "$dofile\13table12.do" 11_1skill_count.do must precede 11_2table10_mechanism.do. Similarly, 12_1_external_consistency.do must precede the Figure 2 and Table 11 scripts. 6. All reproduced tables and figures are saved in the output folder. Intermediate datasets are saved in workdata. Existing files with identical names are overwritten. Additional information is provided in README.md.
Institutions
- Southwestern University of Finance and EconomicsSichuan, Chengdu