Methodology
Empirical design and estimation strategy
A polished presentation of the core econometric conditions, identification strategy, and robustness checks behind the workforce ESG analysis.
Panel DataIndustry Fixed EffectsFirm Fixed EffectsClustered SEsLagged VariablesInteraction Models
Interactive methodology cards
Click each card for a plain-English explanation and econometric interpretation.
Residual financial slack
The residual slack construction separates discretionary cash from predictable cash flows, improving interpretability for the joint-condition test.
ROA
Firm Size
Leverage
Firm Age
First-Stage Regression
Cash-to-Assets regressed on firm fundamentals
Predicted Cash Holdings
Expected cash given firm profile
Actual Cash − Predicted Cash
Difference between observed and expected
Residual Financial Slack
Discretionary cash beyond what fundamentals predict
Sample construction
The study uses a ten-year EU listed-firm panel with industry and time controls to balance cross-sectional and within-firm evidence.
Starting file
Initial EU listed firm-year dataset
62,664
4,476 firms
Firm-year restructuring
Reshaped from long format into longitudinal panel structure
44,760
4,476 firms
Workforce ESG availability
Missing Workforce ESG values converted to missing without direct row deletion
44,760
4,476 firms
Outlier treatment
61 extreme financial observations set to missing using variable-specific plausibility bounds
44,760
4,476 firms
Minimum consecutive-year requirement
Minimum 3 consecutive years of complete data per firm
8,180
818 firms
Lagged-variable construction
First observation per firm removed after lagging variables
5,552
818 firms
Final analytical sample
21 countries · 63 industries · 2013–2022
5,552
818 firms
Workforce ESG unavailable and outlier treatment refer to observations converted to missing rather than direct row deletions. Final analytical sample consists of 5,552 firm-year observations across 818 firms.