Why do firms pay dividends? A new approach to old puzzles

Starting out with the current puzzles in the international research literature, our project reconsiders the determinants of cash payout to stockholders in an environment which is particularly well suited for this purpose. First, we address both public and private firms, which is very unusual in the literature. This allows us to analyze two firm types which differ widely in terms of major dividend determinants, i.e., information asymmetry, agency costs, stock liquidity, and financial constraints. Including private firms in the sample also allows us to better understand a firm type which has been ignored in the finance literature, despite our finding that it represents four times more employment, four times more sales, and twice as much assets as public fir ms. Second, we observe the full ownership, board, management. and family structures of every firm, which allows for an unusually close study of how dividend policy interacts with corporate governance mechanisms. Third, we will choose sample firms in which the tax system has a neutral effect, both regarding dividends vs. capital gains and dividends vs. wages. This means we can ignore complicating interactions between taxes and other dividend determinants, which has been a problem in the literature, particu larly in the US. Fourth, our data set is considerably wider, deeper, and more accurate than what has been available internationally in the past. We study the population of Norwegian public and private firms with limited liability over the period 1994-2007 . This represents almost 80,000 active firms with consistent, publicly audited data per year, which also allows us to account for all kinds of interactions between dividends and tunneling. Finally, we account for potential endogeneity between dividends, i nvestment, and financing, which is particularly important in privte firms.