Some Direct Evidence on the Dividend Clientele Phenomenon

TLDR: In this paper, the authors address the question as to whether a tax-induced "clientele" effect seems present in the equities marketplace, and provide evidence which bears on the potential role of one of the key such imperfections commonly cited: differences in the rates of tax paid by different investors on their dividend receipts.

Abstract: dividends-irrelevance proposition of Miller and Modigliani [28], and the more recent empirical investigations by Friend and Puckett [14], Diamond [7], and Black and Scholes [1], the dispute as to a dividend effect has continued. Although the issue would appear even yet not to be fully resolved, the character of the debate was modified substantially by the Miller-Modigliani demonstration that, in a perfect capital market, dividend policy cannot matter-given an established investment policy. Since then, the emphasis of conjectures that dividends nonetheless may still matter in practice has shifted to a concern primarily with market imperfections as an underlying cause. We offer here some evidence which bears on the potential role of one of the key such imperfections commonly cited: differences in the rates of tax paid by different investors on their dividend receipts. In particular, we shall address the question as to whether a tax-induced "'clientele" effect seems present in fact in the equities marketplace. Since treatments of the other side of the dividend coin-the acceptability criteria for investment projects financed by retained earnings-have also generally contended that investor tax rates bear on the problem [3][12][23], there are a number of ramifications for managerial decisions if a clientele array can be discerned.