This paper uses a clever check for reverse causality. The authors find that for firms whose incorporating state adopts anti-takeover laws in the preceding twelve months, wages rise and productivity falls. The reverse causality story is that rising wages (perhaps due to union pressure) and falling productivity threaten firms in a certain state, and therefore make that state more likely to enact anti-takeover laws. The authors’ story depends on causality in the right direction.
The paper’s main tests use a binary variable BC to indicate whether a firm’s incorporating state enacted anti-takeover laws in the prior 12 months. The coefficients on BC in the several regressions are the paper’s main findings.
In the reverse-causality check, the authors replace BC with four dummy variables:
- Before-1 for firms incorporated in states that had not yet passed legislation but would do so in the next 12 months
- Before0 for firms incorporated in states that had passed legislation in the prior 12 months
- After 1 for firms incorporated in states that had passed legislation between 12 and 24 months prior
- After2+ for firms incorporated in states that had passed legislation two or more years previously
Significant coefficients on Before-1 would have suggested that legislation was passed in response to an already-changing business environment.
Before0 was the variable of primary interest.
Small and Insignificant coefficients on After 1 and After2+ would have suggested that the effects of legislation were short-lived and therefore economically much less interesting.
In this case, the coefficients on Before-1 were small and statistically insignificant, ruling out the reverse causality. The coefficients on After 1 and After2+ were significant and even larger than the coefficients on Before0, strengthening support for the authors’ story since their reported effects of anti-takeover laws continued to grow in the longer term.