The authors propose an alternative robust technique to test for discontinuities in distributions and provide consistent evidence of discontinuities around zero for both scaled and unscaled earnings levels and changes. The advantage of the proposed test is that it does not rely on arbitrary choice of bin width choices.
To evaluate the power of the test, the authors examine the density function of non-discretionary earnings and detect no evidence of discontinuities around zero in levels and changes of these non-discretionary earnings. As robustness, the authors use pre-managed earnings excluding accrual and real manipulation and find similar evidence.
The finding using our technique support the Burgstahler and Dichev (1997) interpretation on earnings management, even for smaller sample sizes and reject the theory that discontinuities arise from scaling and sampling methods.
The study provides an overview of those studies that support and those that oppose using “testing for discontinuities” as a way to examine earnings management. The authors advance the literature by providing an alternative methodology supporting the view that the kink in the distribution represents earnings management.
