New View Integration
64 Pages Posted: 10 Aug 2017
Date Written: August 3, 2017
This paper examines the design of corporate integration systems, comparing integration limited to equity issued after enactment (New Equity Integration or NEI) to integration that applies to all equity (complete integration). It shows that NEI achieves all of the efficiency benefits of complete integration at a fraction of the cost. NEI, unlike complete integration, is, moreover, supported by both the traditional view and the new view of the effects of dividend taxation. From an efficiency perspective, NEI is strictly better than complete integration.
The problem NEI systems face that complete integration systems do not is distinguishing new equity from old and preventing churning, transactions designed to allow old equity to get the integration benefits given to new equity. Churning, I argue, is like any other type of avoidance. Few systems completely stop avoidance, and we generally try to limit avoidance rather than allow it freely, which is what complete integration effectively allows. To understand the viability of new equity from old, the paper considers three NEI systems: an explicit transition tax, a tracing method described in the American Law Institute’s 1982 and 1989 corporate tax reports, and a set of methods based on the economics of consumption taxes. Although all three present trade-offs and administrative problems, all three are feasible. As a result, studies of integration should focus on NEI rather than complete integration.
The double-level tax on corporate earnings is thought to discourage the use of the corporate form, to encourage corporations to retain earnings, and to encourage the use of debt financing in place of equity financing. Eliminating or reducing these distortions by eliminating one of the two levels of taxes, a policy known as corporate integration, is one of the central components of many tax reform proposals.
Examinations of corporate integration tend to take a dichotomous view of the available policies: they take the view that we should either have what I will call complete integration or none. Complete integration is an integration system that applies to equity existing at the time of enactment (existing equity or old equity) and to equity issued after the time of enactment (new equity). As will be explained below, complete integration is thought to be premised on empirical support for a model of the effects of the corporate tax known as the Traditional View instead of an alternative model, known as the New View. If the Traditional View holds, the efficiency gains from complete integration are thought to be substantial, while if the New View holds, they are modest. Because of the considerable revenue costs, complete integration is only desirable with the efficiency gains that come with the Traditional View. If instead, the New View holds, integration may not be desirable because the revenue costs of eliminating the double-level tax might exceed the efficiency gains.
My goal here is to reexamine the arguments for integration, focusing on the implications of the Traditional and New Views. My core conclusion is that what I will call New Equity Integration or NEI, which is integration applied only to equity issued after the date of enactment, is preferable to either complete integration or no integration. NEI achieves all of the efficiency gains of complete integration, is supported by both the New View and the Traditional View, and would only cost a fraction of the cost of complete integration.
The central problem with NEI is that it requires distinguishing new equity from old. New and old equity in the same class of shares cannot be distinguished by observable features. And even if there were a method of distinguishing new and old equity – suppose that the law required new equity to be of a different and identifiable class – corporations could redeem their old equity and issue new equity with the same economic rights. A corporation that does this would not have changed its capital structure, its investments, or anything else that might matter – it would have simply moved pieces of paper around. It would, however, argue that it has new equity eligible for the benefits of NEI.
That taxpayers will engage in transactions to allow old equity to get new equity treatment, which I will call churning, reduces the benefits of NEI. The prospect of churning may mean that NEI systems need elaborate rules to prevent it. Churning that remains notwithstanding these rules will increase the revenue cost of NEI and distort financial structures. Although it is lost in the fog of time, it seems that the reason that NEI has not received significant attention is that the problems of churning are thought to be so difficult that they cannot be overcome. As a result, perhaps complete integration is a better option than NEI.
Churning, I will argue, however, should be analyzed and treated like other tax avoidance problems. Any time the tax law has to draw a line, taxpayers will seek to structure transactions to be on the favorable side of the line, generating economic distortions and revenue losses. Our general approach to this sort of tax avoidance is to try to limit it. Only rarely do we say that a particular tax avoidance problem is so severe that it is better to just allow it rather than incur the costs of trying to prevent it. Yet complete integration does exactly that. It automatically grants the benefits of churning to old equity. It is, effectively, NEI plus free churning. While allowing free churning might be the right choice, it is not the standard approach to line drawing, and, at a minimum, we need to at least analyze the costs of churning and anti-churning rules before deciding.
To understand the costs of churning, we need to consider methods of limiting integration to new equity. I will examine three: (1) a method based on a suggestion by Alan Auerbach in a 1990 paper to combine complete integration with an explicit tax on existing equity that equals, in present value, the tax that would have been imposed had the dividend tax been retained on that equity; (2) a tracing mechanism proposed by William Andrews in a 1982 and a 1989 American Law Institute study, a version of which was used in Sweden for more than 30 years and Finland for almost 20; and (3) a set of methods based on the economics of cash-flow taxes and the transition to a consumption tax. I will conclude that none of these methods is without flaws but that any of them could be made to work. I conclude that NEI is likely superior to complete integration even once we consider churning and implementation costs. Studies of corporate integration should focus on NEI rather than complete integration.
This paper has five parts. To understand the comparison between NEI and complete integration, we need a clear understanding of the economic distortions caused by the current corporate tax system and the complete integration proposals to fix them. Parts 1 and 2 provide this background, for the most part reviewing existing literature. Part 1 examines the distortions caused by current law. Part 2 considers complete integration methods, focusing on the most prominent methods that have been proposed or have been used in other countries.
Part 3 takes up new equity integration. It makes two claims. First, Part 3 shows that NEI can achieve all of the efficiency benefits of complete integration at a lower cost. While this claim is not novel, it seems to be widely forgotten, ignored, and sometimes disputed, including by some of the major studies on integration. The second claim is that churning should be treated like other forms of avoidance: merely because most NEI methods will allow some churning does not means we should have complete integration which amounts to free churning. Part 4 considers the three NEI methods mentioned above. Part 5 concludes.
Suggested Citation: Suggested Citation