Longitudinal firm panels
Follow the same firms across periods
Longitudinal firm panels link repeated observations on firms or establishments over time. They support analysis of change within entities alongside differences between them. Corporate identifiers, ownership changes, entry, exit and accounting conventions are substantive measurement issues. Fixed effects, random effects and dynamic panel estimators answer different questions and require assumptions beyond simply having repeated firm records.
Choose when the question concerns organisational change, investment or performance over time and stable entity linkage is possible. Confirm whether the unit is a firm, subsidiary or establishment and audit coverage changes.
Strengths
- Controls some stable entity differences with suitable models
- Reveals trajectories and responses over time
Limitations
- Survival and ownership changes complicate comparability
- Time-varying confounding and measurement error remain
Know the boundary
A long firm panel is a data structure, not automatic evidence of causal effects.