Skip to content
Replication Archive for "The Impact of Unemployment Insurance on Job Search: Evidence from Google Search Data"
TeX Stata R
Branch: master
Clone or download

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.


Type Name Latest commit message Commit time
Failed to load latest commit information.

Replication Repository: "The Impact of Unemployment Insurance on Job Search: Evidence from Google Search Data"

This file contains information regarding scripts, paper results, software versions, and data files for the paper:

The Impact of Unemployment Insurance on Job Search: Evidence from Google Search Data


  • Main Analysis Code:
    • analysis/, Clean Texas Unemployment by MSA Data.
    • analysis/, Combine raw Google daily data for Texas.
    • analysis/, Merge Files Necessary for Non-Linear Least Squares Analysis of Texas Data.
    • analysis/, Run non-linear least squares analysis.
    • analysis/, create national Google series from raw files.
    • analysis/, add CPS data to Google search data.
    • analysis/, run national ols regressions.
    • analysis/4d_new_panel_restat.R, run impulse function panel regressions.
    • news_analysis/, create news figures.
    • comScore_analysis/, comscore search correlations.
    • atus_analysis/, compute atus correlations, Table A3.
    • atus_analysis/, generates numbers for Table A1.
    • atus_analysis/, generates estimates for Table 4, Figure A1.
    • google_analysis/, generates estimate for Table 5.
    • latex/*, generate paper and appendix.


  • Tables:

    • Table 2: Correlation of Google Search to Online Job Search Time - comScore Data

      • File Name: GooglecomScore.tex
      • Script: comscore_analysis/
    • Table 3: ATUS Search Time Correlation

      • File Name: ATUSCorr.tex
      • Script:
    • Table 4: Day of Week Fixed Effects for Google, comScore, and ATUS

      • File Name: holiday_dow.tex
      • Script:
    • Table 5: Empirical Tests of Google Job Search Measure

      • File Name: Table1.tex
      • Script:
    • Table 6: Effect of UI Status and Composition on Job Search (Non-linear Least Squares)

      • File Name: nllsregs_current_law.tex
      • Script:
    • Table 7: Effects of UI Expansions and Composition by State

      • File Name: national_regs_new_restat.tex
      • Script:
    • Table 8: Event Study Regression Results.

      • File Name: all_impulse.tex
      • Script: 4d_new_panel_restat.R
    • Table A1: Google Search Term Correlations

      • File Name: manually inputted into paper latex file
      • Script:
    • Table A2: Change in log(GJSI) Following Expansions: Robustness

      • File Name: robust_expansion_restat.tex
      • Script:
    • Table A3: ATUS Summary Statistics

      • File Name: Table created manually following stata output.
      • Script:
  • Figures:

    • Figure 1: Day-of-week Fixed Effects and Placebos
      • File Name: Day_of_Week_Effects.xlsx, dayofweekgraph.png, dayofweekgraph_placebo.png
      • Script:, paper figure made in excel
    • Figure 2: Average Weeks Left by Type
      • File Name: weeks_left_by_type.png
      • Script:
    • Figure 3: News Articles Regarding EUC
      • File Name: national_euc.png
      • Script: news_analysis/
    • Figure 4: Dynamic Response to Changes in UI Generosity
      • File Name: jump_shock_pre_2012.pdf, drop_shock_post_2011.pdf
      • Script: 4d_new_panel_restat.R

Software Versions:

  • R: Revo.version _

platform x86_64-apple-darwin15.6.0 arch x86_64 os darwin15.6.0 system x86_64, darwin15.6.0 status major 3 minor 3.2 year 2016 month 10 day 31 language R version.string Microsoft R Open version 3.3.2 (2016-11-18 17:31:12 UTC)

  • Stata: version 11.2

Data Files:

  • National
    • final_merged_data.dta (file used for national level regressions of job search on UI variables.)
    • rothstein_data.dta (CPS data from Rothstein BPEA, 2011)
    • Folder: Google_Data/StateMonth_2004_2016: Most up to date Google Data in flat files for state by month.
    • Folder: Google_Data/StateMont2004_2013_1: Data used in old working paper versions of the paper. Only goes through 2013.
    • Google_Data/*: Contains aggregations of Google data, either at a daily, weekly, or monthly level.
  • ComScore:
    • comScore_analysis/all2007_final.dta: Data on job search time by user in comScore
    • comScore_analysis/idtostate.dta: map comScore to state
    • comScore_analysis/state_population.dta: state population
    • comScore_analysis/comscorestateday.dta: state by day comscore
  • American Time Use Survey (ATUS):
    • Gooogle_Data/all_aux_term_monthly_new.dta, jobs search for ATUS analysis.
    • ATUS/AllCPSATUS.dta ATUS and CPS data.
    • ATUS/statejobssearch, state level trends jobs
    • ATUS/statelocalsearch, state level trends local jobs
    • ATUS/stateweathersearch, state level trends weather
    • ATUS/StateUnemployment.dta, state unemployment
    • ATUS/gestcenstate.dta, GestfipsGestCen.dta, state and fips codes for merging.
  • Google News:
    • Google_Data/all_news_search_data: Data on news articles from Google.
  • Texas Designated Metropolitan Area (DMA)
    • tx_msa_unemp.dta (TX Unemployment)
    • clean_tx_msa_unemp.dta (TX Unemployment - Clean)
    • final_weekly_search_data (Combined Texas Google Search Data)
    • number_affected_data.dta (For each expansion / dma, how many people affected by getting more weeks?)
    • soph_weeks_left_data.dta (For each dma / week, how many individuals with specified weeks left of UI (current law)?)
    • number_weeks_on_data.dta (For each dma / week, how many individuals with specified weeks on UI?)
    • tx_msa_pop.dta (Population of texas msas.)
    • soph_readyforweeklyregs.dta (Combined all relevant weekly data for Texas NLLS analysis)
    • TX_Weeks_Left_Data_Alt.dta (Weeks left maximum by week in Texas by UI program)
    • Google_Data/TexasDailyData_* : Folders with csv and dta files of Google Data through 2013.
You can’t perform that action at this time.