Skip to content

DATA_INVENTORY.md

Latest commit

 

History

History
89 lines (70 loc) · 4.06 KB

DATA_INVENTORY.md

File metadata and controls

89 lines (70 loc) · 4.06 KB

Data Inventory and Replication Guide

This document links each section of the paper to the underlying data files and scripts.

Section 2.4 – Fixed Factors and Severance Tax

  • Data files:
    • analysis/2.4_fixed_factors_logit/output/Sev_early_switch.csv – cleaned dataset for regression
    • Generated from raw inputs (stored locally in this section):
      • analysis/2.4_fixed_factors_logit/data/filled_data_jmp.csv
      • analysis/2.4_fixed_factors_logit/data/Non_CIT_states_FRED_OH.csv
      • analysis/2.4_fixed_factors_logit/data/ssfa_data_jmp.xlsx
      • analysis/2.4_fixed_factors_logit/data/clean_rates_1976-2022.csv
      • analysis/2.4_fixed_factors_logit/data/rates_jmp.csv
      • analysis/2.4_fixed_factors_logit/data/elasticity_rates_jmp.csv
  • Scripts:
    • analysis/2.4_fixed_factors_logit/scripts/Severance2WFE.R
  • Results in paper:
    • Table 1 – Logistic regression (odds ratios) showing higher severance tax revenues reduce probability of SSFA adoption

Section 2.5 – Employment and Manufacturing

  • Data files:

    • analysis/2.5_employment_probit/data/Severance_Cap_mutate.csv – base panel of severance tax and adoption timing
    • analysis/2.5_employment_probit/data/BEA_GDPGrowth.csv – BEA GDP growth controls
    • analysis/2.5_employment_probit/data/CurrentEmploymentStatistics_National.csv – CES national employment
    • analysis/2.5_employment_probit/data/CurrentEmploymentStatistics_States.csv – CES state-level employment
    • analysis/2.5_employment_probit/data/LocalAreaUnemploymentStatistics_States.csv – LAUS unemployment rates
    • analysis/2.5_employment_probit/data/StatewideManufacturingEmployment_States.csv – BLS manufacturing employment

  • Scripts:

    • analysis/2.5_employment_probit/scripts/SSFA_Why_Switch_Script.R – builds hazard panel, cleans employment/manufacturing/unemployment datasets, runs logit and probit models

  • Results in paper:

    • Table 2 – Logit hazard model of SSFA adoption
    • Table 3 – Probit hazard model of SSFA adoption

Section 7.1 – Year-to-Year Changes

  • Data files:
    • analysis/7.1_yearly_changes/data/real_log_nci.csv – base dataset (real, log taxable corporate income by state/year)
  • Scripts:
    • analysis/7.1_yearly_changes/scripts/desc_log_nci.R
  • Outputs:
    • analysis/7.1_yearly_changes/output/sr_descriptive_log_nci.csv – computed year(-1), year(0), year(+1) logs and differences
  • Results in paper:
    • Table 7 – Yearly differences in taxable corporate income (real, log scale)
    • Figures 1–3 – descriptive plots by adoption timing

Section 7.2 – Truncated Sample

  • Data files:

    • analysis/7.2_truncated_sample/data/real_log_nci.csv
      State-level log non-corporate income dataset, used for TWFE and DiD analyses.

  • Scripts:

    • analysis/7.2_truncated_sample/scripts/TWFE_DiD_EventStudy_2007.R
      Estimates TWFE, simple DiD, and event-study style DiD for the 2007 adoption cohort.

  • Results in paper:

    • Tables 8–10 (TWFE, DiD, and Event-study estimates).
    • Figures 4 and 8 (event-study plots for truncated sample).

Section 7.3 – Synthetic DID

  • Data files:
    • analysis/7.3_synthetic_did/data/real_log_nci.csv – state-level log of real taxable corporate income (base dataset for all sDiD analyses)
  • Scripts: analysis/7.3_synthetic_did/scripts/log_nci.R – runs sequential cohort filtering, synthetic DID (short- and long-run), and generates per-state plots
  • Results in paper: Tables 11–13 (Synthetic DID estimates) Appendix: per-state short-run and long-run plots, Tables 17–21

Section 7.4 – Non-Corporate Income

  • Data files:
  • analysis/7.4_non_corporate_income/data/real_NCI_cap.csv – non-corporate income tax revenue per capita, derived from naive_ci.csv
  • Scripts:
    • analysis/7.4_non_corporate_income/scripts/pe_nonCI_SR_LR.R – constructs NCI per capita dataset, runs sDiD (short- and long-run), and saves estimates/plots
  • Results in paper:
    • Tables 22–24 (Synthetic DID estimates for non-corporate tax revenue)