Dividend and Corporate Taxation in an Agency Model of the Firm 

In our paper Dividend and Corporate Taxation in an Agency Model of the Firm, which is forthcoming in the American Economic Journal: Economic Policy, we propose a simple model based on the agency theory of the firm (Jensen and Meckling 1976) that provides an alternative to the two leading theories of corporate taxation – the “old view” (Harberger 1962, 1966, Feldstein 1970, Poterba and Summers 1985) and the “new view” (Auerbach 1979, Bradford 1981, King 1977). Our model is motivated by empirical studies of the 2003 dividend tax reform in the U.S. (e.g. Chetty and Saez 2005), which found that: (1) the 2003 dividend tax cut caused large, immediate increases in dividend payouts, and (2) the increases were driven by firms with high levels of share ownership among top executives or the board of directors. These empirical findings are difficult to reconcile with the two leading theories of corporate taxation.

The critical new feature of our model is a divergence between the preferences of managers and shareholders. We model this divergence as arising from perks and pet projects, although the underlying source of the conflict between managers and shareholders does not matter for our analysis. Shareholders can provide incentives to managers to invest and pay out dividends through costly monitoring and pay-for-performance. Only the large shareholders of the firm choose to monitor the firm in equilibrium (Shleifer and Vishny 1986, 1997). In this model, a dividend tax cut leads to an immediate increase in dividend payments because it increases the manager’s preference for dividends relative to the pet project and increases the amount of monitoring by large shareholders. Firms where managers place more weight on profit maximization – either because the manager owns a large number of shares or because there are more large shareholders – are more likely to increase dividends in response to a tax cut.

After showing that the predictions of the agency model fit the recent evidence on dividend taxation, we characterize its implications for the efficiency costs of dividend and corporate taxation by deriving empirically implementable formulas for excess burden. We obtain two results that challenge intuitions from existing old and new view models. First, dividend taxes create an efficiency cost even if the marginal source of investment is retained earnings by distorting the tradeoff between pet project investment and dividend payouts. Second, if the contract between shareholders and the manager is inefficient – as is the case in a model with diffuse shareholders – dividend taxation creates a large (first-order) efficiency cost. In contrast, the corporate tax may generate only small (second-order) efficiency costs because it does not amplify the manager’s incentive to hoard cash for pet projects. This suggests that corporate taxes may be a more efficient way to generate revenue than dividend taxes. Indeed, our analysis suggest that a Pigouvian dividend subsidy would be desirable to correct the negative externality created by agency problems in firms.

Our analysis suggests that the main source of inefficiency from increasing the dividend tax rate is the misallocation of capital by managers because of reduced monitoring, and not the distortion to the overall level of investment emphasized in the “old view” model. From a policy perspective, if agency problems are prevalent, dividend taxation should be used relatively little if the government has other tools – e.g., progressive income taxation integrated with corporate taxation – that have similar distributional effects but do not create large (first-order) distortions.

The full paper is available for download here.