Why the Control Group Matters for Displaced Workers

"high-tenure workers who separate from distressed firms" — Louis Jacobson, Robert Lalonde, and Daniel G. Sullivan: The administrative-data identification establishes this worker-firm separation criterion, but the supplied text does not specify the control group’s composition, matching, timing, or identification assumptions.[[cite:1]]
"25 percent per year" — Louis Jacobson, Robert Lalonde, and Daniel G. Sullivan: The evidence establishes substantial long-term earnings losses for the identified displaced workers, but does not show how a control group was constructed to estimate the counterfactual.[[cite:2]]
"not confined to a few industrial sectors" — Louis Jacobson, Robert Lalonde, and Daniel G. Sullivan: The reported losses extend beyond a narrow set of industries, while the supplied text remains silent about whether controls were selected across industries or within them.[[cite:3]]
"some unidentified attribute of the employment relationship" — Louis Jacobson, Robert Lalonde, and Daniel G. Sullivan: The finding highlights that an important lost employment-relationship attribute is not identified; the supplied text does not specify how control-group design separates that attribute from other differences between displaced and comparison workers.[[cite:4]]

Why the Control Group Matters for Displaced Workers

Explore related topics