by Gene Steuerle

In his latest economic perspective column, Tax Notes economic consultant Gene Steuerle continues his look at the application of zero tax rates to returns on new investment and learning; this week looking at "taxing everything, yet taxing nothing."

Date: Sep. 23, 3-96

====== FULL TEXT ======

ZERO TAX RATES ON RETURNS
TO NEW INVESTMENT & LEARNING

PART 4: TAXING EVERYTHING, YET TAXING NOTHING

[1] If zero tax theories are followed to their logical conclusion, they imply that it is possible for government to gather large amounts of revenue without really taxing much of anything. With some significant, yet hardly revolutionary, changes from current law, returns from new physical investment, from education and learning, and from research and development could all be made to forego taxation. Yet the full implications of these theories have not been fully thought out. Where exactly would tax collections come from? And can government operate essentially by enacting a series of surprises, even confiscations, without affecting the behavior of individuals?

[2] One of the first applications of this theory was to argue that integrating corporate and individual taxes would be wasteful, as the second layer of tax had already been paid by owners of old corporate capital. This second layer of tax no longer had any effect on a variety of financial and saving decisions, so the theory said, and should not even be counted in measuring the cost of (new) capital.

[3] A second application was to show the advantage of a tax system with investment tax credits and the higher tax rates required to support them. The tax credits would allow tax to be forgiven on new physical capital, while simultaneously reducing the value of old capital and allowing the government to collect taxes on its output. A third application, often used when discussing tax reform, has been to propose that conversion to a consumption tax without transition rules would impose a windfall loss on owners of old capital; when they consumed out of this capital, they would not have benefited from earlier, immediate write-offs of their investments. In effect, a new double tax could be imposed on old capital. By way of analogy, I have carried this consumption tax treatment over to a fourth application: the taxation of human capital. The cost of almost all education and training, whether provided by government or employers, already is deducted immediately under current law. Practically all research expense is also written off immediately (and much is subsidized). If current law were extended slightly to allow deductions for out-of- pocket educational expenses (only a small proportion of the total), essentially all investment in human capital would receive consumption tax treatment.

[4] If all returns from new physical and human capital investment were given this consumption tax treatment, then exactly where would government be collecting its revenues? Economic theory has not, in my view, done a good job of examining this question -- in part because these questions, much less the answers, are not easily discernible in the models and equations that underlie zero tax theories.

[5] Let us assume, therefore, that all returns from new physical and human capital investment are essentially untaxed in a consumption tax world. Then what is left? Essentially there is old capital, there is pure physical labor, and there are rents and extraordinary returns to capital. Even more than old capital, none of these is well defined or measured.

[6] The existing physical capital stock is sometimes estimated at about $25 trillion. Suppose government designs its consumption tax so that 40 percent of this old capital (or the consumption the capital eventually makes possible) can be viewed as being owned by government. A rate of 40 percent is roughly equal to the effective corporate plus individual tax rate applying to physical capital income today. If individuals consumed the $25 trillion immediately, then government would collect $10 trillion. But consumption tax treatment allows these collections to be deferred into the future. For example, owners of old capital might delay this tax by foregoing consumption temporarily -- that is, by reinvesting this capital and the returns on it. Now government should be able to obtain the same rate of return as other owners of assets. Let us suppose that this real rate of return equals 5 percent. Then government could collect $500 billion per year forever rather than take its $10 trillion up front. If government wanted its revenues to grow at the rate of growth of the economy, then it might keep and spend only half of this annual take, "reinvesting" the rest. In sum, it could collect about $250 billion per year out of old capital, and that amount could grow each year so that government's share of national output remained constant forever.

[7] Next turn to "old" human capital. The amount of this capital is some multiple of physical capital. Let's assume that it's about four times greater than physical capital. This implies that in a world with 40 percent tax rates, government could collect another $1 trillion per year out of this old human capital and maintain its share of national output over time. Add together the $250 billion from taxing old physical capital and the $1 trillion from taxing "old" human capital, and the total tax collections are well in excess of what is raised by the federal corporate and individual income taxes today.

[8] Not all returns to labor, however, represent returns to human capital. There are also some returns to pure manual or physical labor. Unfortunately, this type of labor almost always requires some complement of human knowledge or capital, so it's not easy to separate components. The returns to pure physical labor, however, cannot represent a very large share of total returns to labor, since so-called "unskilled" jobs typically have fairly low wage rates. These returns would add moderately to the total tax collections possible in a "zero tax" world.

[9] Finally, economists sometimes talk about "rents" and "extraordinary" returns to capital as being subject to tax in a consumption tax world. This issue has not been well engaged. Certainly rents from land (ignoring new development expenses) represent returns from old capital that would be subject to tax. By rents, however, economists are also referring to returns that derive from monopolies and oligopolies and similar special circumstances. Large or extraordinary returns often derive from economies of scale -- in the modern economy, more from advertising and name recognition and setting standards than from large production processes. However, there are also many investments that don't pay off: all those firms that failed to become the Microsoft or Wal-Mart of the day. To complicate matters even further, many extraordinary returns may show up as higher-than-average wages as much as higher-than-average returns to capital. Those higher wages, in turn, may represent the ability of some to capture the returns to specialized knowledge even while other workers may receive lower-than-average returns to their own investments in human capital.

[10] In effect, once we add together winners and losers, it is not clear whether "extraordinary" returns are even positive in the economy as a whole. If there are returns to risk-taking in both physical and human capital, nonetheless, government should gain some share of those returns. Consider government's "assets" as its tax share of future consumption. A higher rate of growth in the economy means a higher rate of return for the government on its deferred tax collections.

[11] From still another perspective, one might argue that national income at a point in time represents almost entirely returns to old physical and human capital and from pure physical labor. Competition and market pricing would dissipate extraordinary returns if this economy exhibited no growth. Only growth (per capita) over and above existing income or output represents returns to additional investment (per capita) in physical or human capital and, if it exists, net positive extraordinary returns. Accordingly, zero tax theory may imply that consumption tax treatment of all physical and human capital investment would result in tax collections only on old physical and human capital, pure physical labor, and some share of growth.

[12] Admittedly, we are skating on thin ice, exponentially raising a number of unanswered questions as we proceed. We have not, for instance, compared how much old capital or growth is taxed by an income tax relative to a consumption tax. We are on empirically weak ground when trying to divide up returns to labor between returns to human capital and returns to pure physical labor. Many of the returns to human capital, moreover, represent not returns to knowledge, but to the acquisition of knowledge that is temporarily scarce. Nonetheless, if claims are going to be made for the efficiency and equity of alternative tax systems, then we ought to make clearer exactly what we are going to tax and not tax. This type of Garden of Eden analysis is required.

[13] There is one last issue that I wish to address. Many consumption tax advocates proclaim the efficiency gains that derive from zapping old capital. This, however, is dangerous ground for at least three reasons.

[14] First, zapping old capital is not necessarily a conservative argument. Raising tax rates in almost any tax system zaps old capital to some extent. The corollary is that lowering tax rates in almost any tax system creates windfall gains for old capital that formerly has been zapped. If you need evidence, just go back to the literature arguing that a double tax on corporate income adds nothing to the cost of capital and that higher income tax rates should be used to support existing investment credits. Therefore, while zapping old capital is thought by some to be a conservative argument for eliminating taxes on capital, it is simultaneously an argument for raising or maintaining higher tax rates.

[15] Second, if one believes in zapping old capital, it is not clear why the debate should be confined to income taxation. The belief clearly provides some support for rent controls on "old" property and for property taxes on land (the idea of concentrating property taxes on land, associated with Henry George, representing a much earlier attempt to apply zero tax theory to the zapping of old capital). This type of theory, therefore, has a long history and many applications. One could extend it further in all sorts of directions, for instance, to argue that license fees should be charged only on professionals who formerly obtained their credentials or, for that matter, to any onetime, outright confiscation of property.

[16] Finally, the belief that windfall taxes and confiscation can be efficient relies heavily on the notion of surprise and false expectations -- that individuals will not change their behavior in response to such changes. Remember that the models underlying zero tax theory almost always assume that there is only one new zapping of old capital. This new tax system remains in place for eternity, and nobody ever expects another zapping. Once one builds in changes in expectations, the efficiency gains from zapping old capital are reduced, eliminated, or even made negative.

Tax Analysts Information

Code Section: Tax Policy
Jurisdiction: United States
Subject Area: Legislative and Policy Issues
Index Terms: investment incentives
Author: Steuerle, Gene
Institutional Author: Tax Analysts
Tax Analysts Document Number: Doc 96-25861 (2 pages)
Tax Analysts Electronic Citation: 96 TNT 186-79
Cross Reference: