Dividend Policy And Earnings Management Across Countries
University of New South Wales (UNSW) – School of Accounting
York University – Schulich School of Business
Suffolk University Sawyer Business School
University of New South Wales (UNSW) – School of Banking and Finance; Financial Research Network (FIRN)
November 1, 2015
This paper examines whether dividend policy is associated with earnings management and whether the relationship varies across countries with wide-ranging degrees of institutional strength and transparency. Based on a sample of 23,429 corporations from 29 countries, we show that dividend payers manage earnings less than dividend non-payers, and that this evidence is stronger in countries with weak investor protection and high opacity. Further, we find that dividend payers manage earnings less when they issue equity following dividend payments, and that this result is more pronounced in countries with weak institutions and low transparency. Overall, our evidence suggests that firms may employ dividend policies associated with less earnings manipulation to mitigate agency concerns and to establish credible reputation, thereby facilitating access to external funds.
Dividend Policy And Earnings Management Across Countries – Introduction
Paying dividends limits private control benefits available to insiders as cash paid out provides fewer opportunities to consume these benefits (Pinkowitz et al., 2006). Prior studies, however, show that paying dividends does not prevent firms from committing accounting fraud,1 thereby suggesting that dividend paying firms may not necessarily act in line with shareholder interests. The goal of this study is to examine whether there is any association between dividend paying status and earnings manipulation across firms from a broad spectrum of markets with wide-ranging degrees of investor protection and transparency. Such a pursuit is particularly relevant as extant studies report that non-U.S. firms may experience less pressure to maintain their dividend policies than their U.S. counterparts.
Prior literature finds that earnings management is less prevalent when private control benefits are limited (e.g., Leuz et al., 2003; Gopalan and Jayaraman, 2012). Thus, firm managers who decide to pay dividends may have fewer private control benefits to consume and conceal and hence, could be less likely to fabricate accounting information. Furthermore, both theory and empirical evidence indicate that the U.S. firms are unwilling to cut dividends, and that they engage in dividend smoothing to maintain a constant stream of dividends (e.g., Lintner, 1956; Brav et al., 2005; Skinner and Soltes, 2011). However, non-U.S. firms appear to be less concerned about dividend cuts and thus alter their dividend policies more frequently (e.g., Dewenter and Warther, 1998; Chemmanur et al., 2010; Hail, Tahoun, and Wang, 2013). Such evidence implies that the relationship between dividend paying status and earnings management may not persist or may be weaker in an international setting. Alternatively, it is plausible that the link between dividend policy and earnings management is stronger in a cross-country examination. This may be due to a large variation in country-level investor protection measures (e.g., Jensen, 1986; La Porta et al., 2000). Specifically, such differences may result in a potentially stronger need of foreign firms to convey their governance quality via alternative credible signals, compared to their U.S. counterparts.3 We therefore conjecture that dividend paying firms manipulate earnings less than their non- paying counterparts and that the strength of the relationship may vary with country-level investor protection and transparency.
Our analyses focus on the following three closely related issues. First, we examine the relationship between dividend policy and earnings management. We gauge a firm’s earnings management by the magnitude of its abnormal accruals with a larger magnitude of abnormal accruals indicating more aggressive earnings management. We test the relationship on a large cross-section of 23,429 rms, whose dividend policies vary substantially across 29 countries, over a 21-year period from 1990 to 2010. Second, our analyses exploit the richness of our data by examining whether and how the relationship between dividend policy and earnings management differs across countries with varying degrees of investor protection and transparency. Third, our study attempts to provide further insights regarding the observed variation of the dividends-earnings management relationship with investor protection. We do so by examining the established link in the context of subsequent access to external financing.
We find that dividend payers have smaller abnormal accruals than dividend non-payers, suggesting that the former are less likely to engage in aggressive accruals management to conceal firm performance. These results are economically significant, even after controlling for firm-specific variables that have previously been shown to affect earnings management. For example, our model estimates indicate that the size of abnormal accruals for dividend payers is 7% smaller than that for non-payers. Our findings are also robust to sub-samples of U.S. and non-U.S. firms, sub sample periods of 1990s and 2000s, and various alternative measures of payouts and of earnings management. We also conduct several additional robustness tests and confirm that our results persist. Overall, we offer strong evidence that paying dividends is associated with and leads to lower earnings management.
See full PDF below.