A review of Dai, Y., P. R. Rau, A. Stouraitis, and W. Tan, 2020, “All Ill Wind? Terrorist Attacks and CEO Compensation,” Journal of Financial Economics 135, 379-398.
I. Research question and its importance
Drawing on the executive compensation literature, this paper estimates the causal impact of terrorist attack on executive compensation. The authors postulate and test two channels through which executives require higher compensation subsequent to a terror attack: rational (risk averse) and behavioural (personal wellbeing). Moreover, they examine the compensation structure change after the incident, such as cash- and equity-based compensations. Finally, they investigate the differential impacts amongst executives based on their levels of bargaining power. This research question is important because few prior study investigates the impact of terror attack at micro level, i.e., firm and CEO’s reaction. It also contributes to CEO compensation literature where academic findings are mix when discussing the effect of nonmonetary factor, such as quality of life.
II. Method and finding
The most important independent variable, attack proximity, is a dummy measured by whether headquarter of a firm locates within 100 miles of the incident, and “no other attack occurred within 100 miles of the same firm over the prior three years.” Specifically, they employ Global Terrorism Database for 569 observations of U.S. terror attack from 1992 to 2013. Their specification relies on difference-in-difference models with firm and year fix effects and many controls. Propensity score matching was adopted, and parallel trend assumption was tested. Seven yearly observation windows were set up, ranging from three years before the event to three years after it. For CEO bargaining power measurement, they follow convention from the corporate governance literature to adopt standard measures. Lastly, they conduct robustness checks by using alternative measures of CEO compensation as well as attack proximities.
The authors show that terror attack leads to compensation increase for CEO, but not other executives or staffs. And the increase is mainly driven by cash-based compensation, which altogether with the negative cumulative abnormal returns after the attack, support the view of behavioural channel where CEO requires premium for personal wellbeing following the attack. Yet, the rational channel didn’t get support. Corporate financial policy remains identical one year after the incident. In other words, CEO on average becomes more conservative, but the firm doesn’t.
III. Limitation
I find the definition of terror attack questionable. Initially, the authors follow START where their source originates. However, after checking the observations from the same database, I doubt if these events really represented threat to CEO wellbeing. For example,
Shots were fired overnight at a Coast Guard recruiting station in Woodbridge, Virginia, United States. No one was injured and the strip mall, where the station was located, suffered only minor damage.
(GTD ID: 201011000001)
Event like this is unlikely to affect either firm’s compensation scheme or financial policy. Future research may impose a fatality rate as threshold.