In a well done and interesting work, DeAngelo, DeAngelo, and Stulz tie dividend policy and agency costs (particularly the free cash flow problem) together. Their main point is that if firms did not pay dividends, managers would have too much cash at their disposal.
The authors begin by asking the question "why do firms pay dividends." To answer the question they examine what would happen if firms didn't pay dividends. Specifically they "conservatively estimate that, had the 25 largest long-standing dividend-paying industrial firms in 2002 not paid dividends, they would have cash holdings of $1.8 trillion (51% of total assets), up from $160 billion (6% of assets), and $1.2 trillion in excess of their
collective $600 billion in long term debt. Absent dividends , these firms would have huge cash balances
and little or no leverage, vastly increasing managers' opportunities to adopt policies that benefit
themselves at stockholders' expense."
Moreover, the paper makes the important distinction (made before by Jensen & Meckling 1976 and Easterbrook 1984) that earned equity is in someways different than contributed equity (external financing). Notably, contributed equity comes with investor imposed monitoring and the so-called market discipline that is provided when firms must raise new money. Therefore, firms with higher levels of earned equity should pay out larger dividends since these firms have (ceteris paribus) a greater likelihood of a free cash flow problem.
Sure enough, the authors find that "For the 25 longstanding dividend payers discussed above, the median ratio of earned to total equity is 97%, suggesting that this measure does in fact identify historically profitable firms with potentially large agency problems. Our evidence is uniformly and strongly consistent with the prediction that the probability of paying dividends increases with the amount of earned equity in the capital structure."
Which really should not surprise anyone.
This importance of earned equity is important even after controlling for growth, cash on hands, and other factors thus "indicating that the impact of earned equity on the decision to pay dividends that we document here is an empirically distinct phenomenon from other factors that have previously been shown to affect the dividend decision."
BTW Jensen's 1986 free cash flow problem paper is one of my favorite papers of all time. So much so that I did my dissertation on firms with high cash--finding that investors believe that firms that build up cash reserves do in fact tend to waste them as measured by lower Q values. Thus, this paper by DeAngelo, DeAngelo, and Stulz fits perfectly into my semantic network of managers, excess cash, and dividends. Here is a bad version of a paper based on my dissertation in case anyone is interested. Yeah right!