Previous studies have documented that an announcement of dividend initiation and resumption is associated with an increase in stock price, while Boehme and Sorescu (J Finance 47:871–900, 2002) argue that the dividend anomaly only occurs by chance. However, their sample contains firms listed within 3 and/or 5 years of their respective initial public offering (IPO) dates, as well as regulated firms. We conjecture that the confounding effects of IPOs and regulated firms may interfere with the increase in stock prices due to dividend initiations and resumptions and bias their results. We thus reexamine the long-term stock performance following dividend initiations and resumptions by excluding newly IPO firms and regulated firms. We find no evidence that the non-robust positive price drifts for firms, which initiate or resume cash dividends, is due to the confounding effects of IPOs and regulated firms. Therefore the price drifts after dividend initiation and resumption announcements may be a sample-specific result of chance, even after controlling for possible sample selection biases.