Unusual patterns in executive compensation

We extend prior studies by incorporating inferences from the tax, earnings management and psychology literatures to explore discontinuities in the distributions of salary compensation, bonus compensation, and option grants that are unrelated to determinants derived from traditional agency theory. We document discontinuities at amounts evenly divisible by 100,000 or 10,000. Empirical analyses provide evidence that the sensitivity of bonus compensation to performance is lower when bonus compensation is rounded. We find evidence that rounding of bonus compensation has behavioral effects that improve future operating performance, consistent with Akerlof’s (1982) theory of gift exchange. We also find evidence that rounding is more common in firms where agency conflicts are more prevalent