The measurement choice is the part worth studying. Rather than asking executives how confident they feel, the team behind the paper, published in the Journal of Banking & Finance [13], watched what CEOs did with their own money: a chief executive who keeps holding profitable options is leaving personal wealth tied to one company's fortunes, where a more cautious peer would have exercised and diversified [3]. Revealed preference beats a questionnaire, and it is noisier than one, because holding can reflect a vesting schedule or a tax position as easily as the inflated belief in one's own judgment that the authors are trying to capture [9].
The panel is wide rather than deep, and the arithmetic is worth doing before reading "2002 to 2023" as two decades of continuous coverage. That span contains 88 quarters [1]. Spread 37,000 firm-quarter observations across 1,369 firms and the average firm contributes about 27 of them [2], near 31% of the window [3]. A good share of the estimate therefore comes from comparing firms with each other rather than from watching one firm change course after its CEO changed.
A percentage with no level attached is hard to convert into money. The reported gap is around 24% less per quarter [4], and no dollar base is given for it, so whether that is a rounding error or a real withdrawal of political cover is not something this summary settles. The direction does get support from the other margin: these firms were also likelier to do no lobbying at all [5], which is the cleaner behavioural signal, because a zero is a decision rather than a budget trim.
The financial-crisis comparison is the closest thing here to a control condition. When uncertainty rose and many firms increased lobbying, firms led by overconfident CEOs reduced theirs [7], and Lin reports the same countercyclical pattern in quarters of higher political risk generally, when lobbying would be most useful as a risk-management tool [6]. That helps rule out the dull explanation, which is that these firms simply operate where lobbying has little value. If that were the story, they should have tracked the crowd upward from a lower starting point.
The thing this does not tell you is whether skipping the premium cost anything. Lin's account of the mechanism is that lobbying builds political connections and mitigates risk from regulatory change, and that CEOs usually direct the decision [8], which is how a personal trait becomes visible in a spending line. But the exposure in the Queensland framing is described as potential [1]. What is measured is a spending gap, not an enforcement action or a rule that landed badly. My read is that this is useful as diligence rather than as a risk score: a board that already tracks its CEO's option holdings for alignment can read the same data a second way and check it against a lobbying total it already files [11]. Converting 24% less lobbying into 24% more regulatory risk needs an outcome test this design does not attempt.
