Copyright (c) 1992 Tax Analysts

Tax Notes

 

MAY 18, 1992

 

LENGTH: 13622 words 

 

DEPARTMENT: Special Report (SPR) 

 

CITE: 55 Tax Notes 957 

 

HEADLINE: 55 Tax Notes 957 - THE CONSUMPTION OF CAPITAL GAINS. 

 

AUTHOR: Johnson, Calvin

 Tax Analysts 

 

SUMMARY:

 

   Calvin H. Johnson is professor of law, University of Texas. 

 

   Professor Johnson argues that a reduction in capital gains rates will cause the consumption of productive capital that could otherwise be preserved. The case for reduction in the maximum capital gains rates, Johnson argues, depends fundamentally upon a presumption that the proceeds of sales benefiting from the reduction will be reinvested in some other productive investment. Under common law property concepts, proceeds of the sale of the feudal land had to be reinvested as part of the corpus interest. Economists continue to assume that the proceeds will be reinvested even when there is nothing in law or custom that now requires it. Economists argue, for instance, that lower capital gains rates will unlock capital; unlocking capital serves productivity, Johnson argues, only if the capital moves to more productive investment. A number of economists in the current capital gains debate have presented models assuming that capital gains will be reinvested and ignoring the consumption of the capital gains. 

 

   The presumption that proceeds of sale will be reinvested, Johnson argues, is unsound. In a competitive economy, it is not usually rational to sell one property to reinvest in another if there are any material transaction costs in the change. Sales that occur, where tax is material, are sales made at least in part to consume the proceeds. Johnson argues that much, probably most, of the tax cut will benefit sales proceeds that are consumed by investors with all deliberate speed. 

 

   A cut in the maximum tax rate, Johnson argues, focuses the tax cut on the cases where consumption by the wealthy is most likely and reinvestment is least likely. A cut in the maximum tax rate commonly leaves the investor with too high transaction costs for a sale to reinvest to be rational, even after the tax reduction. Forgiving the last of small amounts of tax due would at least better focus the tax cuts on sales where reinvestment is possible. Professor Johnson also argues that lower capital gains rates should, at minimum, be conditioned on a requirement that proceeds of sales benefiting from the preferential rate be rolled over into new productive investments. Absent a rollover requirement, productivity would be better served by high taxes that lock capital into existing investments and preserve the capital than it would be served by tax cuts that induce consumption of the capital.

 

   The author wishes to thank Doug Laycock, Eugene Steuerle, Joseph Dodge, and Tom Evans for helpful comments, and to thank Tobin Olson 

 

TEXT: 11 MAY 92 

 

   The Bush administration proposal to cut the maximum tax rate on capital gains to 15.4 percent /1/ invites a review of the structural issues in the taxation of capital gain. Structural issues are independent of the level or distribution of tax. Within any level or distribution of tax, the structure should not give unintended or indefensible incentives. Whether revenue is to be raised or lowered, the structure of tax should not be made worse. This article is the second of a trilogy dealing with the structural issues involved in the taxation of capital gains. The first article, entitled 'The Undertaxation of Holding Gains,' /2/ argued that holding gains are undertaxed. The third article, entitled 'The Private Advantage of Money-Losing Investments Under Cut-Rate Capital Gain,' /3/ argues that the capital gains cut will distort the allocation of capital. All three articles are critical of the impact of the proposed tax cuts on the structure of the tax system. 

 

   This article argues that preferential tax rates for capital gains need to be conditioned on a requirement that the proceeds of a sale qualifying for capital gain be reinvested in some new productive investment. The most influential arguments in favor of a capital gains tax cut rest on an assumption that the proceeds of a  capital gain sale will be reinvested. Without a statutory requirement that the proceeds be reinvested, however, studies show that a large proportion, probably most, of the capital represented by the sale will be diverted to consumption or consumption assets and not reinvested. In a competitive market in which returns from alternative investments are in equilibrium with each other a sale solely to reinvest the sales proceeds is not usually rational. For the sale to occur then, the investor must have planned to consume at least some of the proceeds. Reductions in the maximum tax rate, moreover, focus the cuts on sales where reinvestment is least likely and consumption by the wealthy is most likely. A reduction in tax which leaves the investor with substantial transaction costs will largely benefit sales where sale to reinvest is irrational. In absence of a requirement that capital gains be reinvested in some new productive investment, in any event, high taxes on sales that lock in capital better preserve the capital and serve productivity than do low taxes that encourage investors to consume their capital. 

 

   Section A shows how the reinvestment presumption underlies the most important arguments in favor of lower capital gains. Section B reviews the existing evidence about consumption of capital gains sales. Section C argues that consumption is a necessary requirement of a sale, within an efficient market. Section C also argues that tax cuts that do not bring the transaction costs within the scope that allows rational sales will be wasted on sales to consume.

 

 A. The Reinvestment Presumption 

 

   The most influential arguments in favor of reducing the tax rate on capital gains presume that the proceeds from a sale qualifying for lower rates will be reinvested. Supporters of a preferential capital gains rate argue, for example, that taxes imposed on sale of property 'lock in' capital, preventing the capital from shifting from old to better investments. /4/ Capital gains taxes are a toll charge on the turnpike from old to new investments. An investor will sell an old investment to travel to the new one, it is aruged, only if the new investment gives a return that exceeds the return from the old by enough to pay the toll charge. /5/ If the toll charge were lowered, the argument goes, capital would shift from old to new more readily. Even opponents of preferential rates have called the lock-in argument the most persuasive of the traditional arguments for preferential rates on capital gains. /6/ 

 

   The lock-in argument presumes that capital will be reinvested. Capital gains taxes are described as taxes on reinvested capital, even in the face of statistical evidence showing that the gains are increasingly consumed. /7/ Staff economists for the Joint Committee on  Taxation, /8/ the Library of Congress Congressional Research Service, /9/ and the Congressional Budget Office /10/ all have published pieces assuming or arguing that capital from capital gains sales will be reinvested. The staff advising Congress is thus not likely to reconsider the assumption or to test what proportion of the unlocked capital will be consumed. 

 

   Reinvestment of existing capital plays an important role in justifying the proposals in part because of the weakness of the case for the cuts in capital gains taxes as incentives for savings. The tax cuts would apply to capital that has already been created and to investments that have already been made. Because of the nature of time, it is impossible for a subsidy, no matter how generous, to increase the creation of capital in years already past. /11/ For many years after the proposal is adopted, even new savings that might be influenced by lowered tax rates will be a miniscule percentage of the stock of capital that would benefit from the cut. The retroactive operation of the cuts forecloses many arguments that might otherwise be made in favor of tax subsidies or incentives. 

 

   Even if the cuts were better targeted to new investments that might be influenced by lower rates, the cuts are unlikely to improve savings. The consensus in the economic literature is that personal savings show little if any positive response to increases in the after-tax returns on investment. /12/ Aggregate savings decline when after-tax returns rise. Most savings, apparently, are designed to meet specific targets. When investment returns go up, savings go down, because savers can meet their future goals with lesser immediate savings. /13/ As argued in the third of this trilogy of articles, cuts in capital gain tax would worsen rather than improve the allocation of new capital. /14/ Thus, the tax cut has to be justified by its influence on shifts of capital already committed to investment, rather than by its influence on the creation or channeling of new capital. 

 

   The presumption that capital gains are reinvested also gives force to the argument that tax cuts will be good for everyone if the tax cuts cause government revenue to increase. The issue of whether tax cuts will raise or lower revenue is hotly contested. /15/ The prospect that a tax cut might raise revenue is the most politically attractive aspect of the capital gains cut. The reinvestment presumption comes into the debate with the assumption that a revenue rise will be an unalloyed good. If a cut in rates increases Treasury revenue, proponents argue, the revenue gain will be a free lunch for everyone. The fact that capital gains are distributed disproportionately to wealthy taxpayers would not matter, proponents argue, because everyone would be better off, and no one will be worse off. /16/ Even if government revenue decreases some, proponents argue, the improvement from shifts to better investments would be well worth the cost. /17/

 

   To the extent that wealthy sellers consume the proceeds of sales, these arguments lose their force. Under the lock-in argument, the purpose of a capital gains cut is to cause the reinvestment of capital in some better use. To the extent that sales proceeds are consumed, however, a sale of an investment means that capital will disappear, rather than shift to a more profitable investment. Lock-in is plausibly better than unlocking capital if the unlocked capital ceases to be productive investment. Consumption reduces aggregate investment. /18/  If the existing investment makes any contribution to productivity, it contributes more than will the capital that has disappeared. To the extent that the proceeds of sale are consumed, a tax cut that increases sales will increase the reduction in capital. 

 

   Similarly, if gains are consumed, an increase in government revenue is not a free benefit for all. Looking only to its budgeted accounts, the government may well gain revenue from increased sales, but that would be an unalloyed good only because the accounting fails to take account of the reduction in aggregate capital caused by the easier consumption of capital. A tax hike that lost revenue, but prevented the consumption of the seed corn, could be the better choice. If capital gains are consumed, government revenue might increase, but only because the government too will share in the increasing liquidation of our capital heritage. 

 

   Consumed capital gains could also quite reasonably be taxed at rates above the average. Consumed amounts are the least controversial heart of the tax base /19/ and capital gains are highly concentrated in the hands of the richest taxpayers. /20/ Consumed capital gains are consumed by our richest taxpayers. It is, therefore, the least desperate needs /21/ that are satisfied by the reduction in tax. Taxes on disinvested capital gain are a 'most rational luxury levy.' /22/ Reducing the tax on consumed capital gains just lowers the sluice gates to let out more of backed up demand for greater, largely luxury consumption. If capital gains are consumed, reducing the tax may just be part of the political demand for immediate gratification, that is, part of the problem of our country's 'Maypo syndrome' /23/ rather than part of the solution. 

 

   The assumption that capital gains are reinvested is based on traditional property law doctrine. Under feudal property systems, the proceeds from the sale of appreciated property (usually land) were 'corpus' or 'capital' that had to be reinvested for the benefit of remaindermen. The gains were not part of the income interest (usually the harvest from the land) that could be distributed and consumed by the income beneficiaries. /24/ Under modern law, the question of whether gains are part of the corpus or part of the income interest is under the full control of the instrument setting up the interests. /25/ Modern property is only rarely subject to entailments or other restraints resembling feudal tenures. Property-law doctrines, moreover, have little  meaning in the design of a reasonable tax system. /26/ Still, the traditional requirement that gains be reinvested has had continuing intellectual influence. The adoption of preferential rates for capital gains, as a historical matter, is best understood as a compromise with the view that capital gains, because they had to be reinvested, were not within the purview of a tax on income at all. /27/

 

 B. Enforcing the Reinvestment Presumption 

 

   Capital gains arise when the investor liquidates a prior investment pulling the sales proceeds from the investment. Nothing in law, the proposals, or nature requires that the proceeds be rolled over into a new investment. In the absence of an express reinvestment requirement, the next sections show, loss of capital to consumption of the sale proceeds will be a serious problem, especially where the most tax is at stake. Sales qualifying for the proposed rate cut will then represent not just a sound reinvestment of old capital, but rather the liquidation and consumption of the old estate. Since reinvestment is such an important tenet in the justification of preferential capital gains rates, reinvestment should be added as a condition for eligibility for the lower rates. 

 

   Under a number of different 'rollover' provisions under current law, a taxpayer's gain from cash received is not taxed, under the specified circumstances, provided the taxpayer reinvests the cash ('rolls the cash over') into a new replacement property that is like the property the taxpayer has left. /28/ The rollover idea would work as effectively as a condition for preferential rates as it does as a condition for full nonrecognition of tax. /29/ 

 

   Replacement property qualifying a seller for the preferential rate would have to be defined to exclude consumption assets like yachts, cars, and houses, as well as collectibles. Such consumer durables are technically investments because they give continuing value beyond the end of the year, but they should be treated as consumption for the purpose of a rollover requirement. The true consumption of durables is in the imputed rent from use of the durable. But it is impossible to tax the imputed rent from yachts, cars, or houses directly, by reason of a combination of inadministrability and politics. The initial purchase price of the durable thus needs to be treated as the consumption as a proxy to reach the consumption in the use of the durable. 

 

   Consumer durables are also not the kind of investments that can be expected to increase worker productivity or international competitiveness or to generate technological change. They are unlikely to generate rich externalities, benefiting someone other than the owner. The current administration proposals exclude 'collectibles' from the benefit of the proposed preferential rates, on the ground that there is no public reason to subsidize collectibles. /30/ Neither collectibles nor consumer durables merit the subsidy. Accordingly, if the sales proceeds are reinvested in houses, cars, or other consumer durables or in collectibles, the proceeds should not qualify for preferential treatment reserved for rolled-over capital.

 

 C. The Evidence of Consumption 

 

   Although the evidence of what proportion of the proceeds of capital gains sales are consumed is inconclusive, the evidence points toward a conclusion that a large proportion of capital gain proceeds, probably most, will be diverted from productive investment and into consumption or consumption durables. One of the most interesting studies is that of James Poterba of MIT, who examined the data from the leveraged buyouts of  the late 1980s. The 1986 Tax Reform Act shifted the balance for established companies away from stock capitalization and in favor of debt capitalization. /31/ Corporations responded to the Act by incurring debt on the order of $100 billion to buy back their stock from their shareholders with borrowed cash. Poterba has estimated that shareholders spent between 40 and 59 cents in consumption of every dollar that they received. Almost all of that consumption was in the form of the purchase of consumer durables. /32/

 

   The Poterba study is supported by increasing evidence in the literature that there is a strong 'mail box' effect, /33/ under which checks in the mail increase consumption, even though they represent just a conversion of prior wealth from noncash to cash form. Under the life cycle hypothesis, consumers maximize the utility of their income if they spread out the consumption of their total projected income evenly over their entire lifetime. /34/ Under this hypothesis, consumers should vary consumption according to changes in permanent income or in wealth (i.e., the discounted value of all projected income), but not according to cash received. Under the model, a mere increase in cash should not increase consumption unless the cash is associated with an increase in wealth or permanent income. 

 

   An increasing amount of empirical evidence, however, shows that consumer behavior is inconsistent with the life-cycle or permanent income thesis. /35/ Consumption is more closely related to receipts, averaged over a time-horizon of just a few years, than it is to permanent wealth. Consumption is more volatile than allowed by the life-long time-horizon of the permanent income or life-cycle hypothesis. A large proportion of income is received by individuals who consume according to current income rather than according to wealth or permanent income. /36/ Most consumers seem to smooth out atypically large cash receipts over a period of only a few years. /37/ They keep only enough savings to buffer uncertain incomes in the next few years. /38/ 

 

   Skinner and Feenberg, looking at capital gains realized for a panel of taxpayers in 1986, concluded that half of the proceeds from the capital gains sales in 1986 were reinvested in temporary interest-bearing accounts. /39/ But proceeds of sales parked in temporary accounts for a while, pending increased consumption  that can liquidate the accounts, are part of the problem. A liquidation of the old estate that allows a dramatic increase in rate of consumption would be a concern, even if the proceeds must be parked temporarily in some way-station account and drawn down for consumption only over some period of time after the year of the original sale. The consumption of the proceeds does not occur all in one period, but the amount of the consumption, meaningfully measured in time value terms, has increased dramatically because of the sale. /40/ 

 

   Consumption of a large proportion of realized capital gain can be expected in part because households consume so much of their income. In 1988, for instance, households in the United States consumed 96 cents of every dollar of after-tax personal income and saved only four cents. /41/ The current low rate of savings in the United States is part of a long-term drop in savings rates. Individuals are consuming an ever larger portion of what they get their hands on. /42/ U.S. savings rates are said to be a tenth of the level of Japanese savings rates. /43/ To pile on the bad news, savings do not seem to respond to increases in real interest rates, even at record high levels. /44/ Even the rich, who are the people who receive capital gains, consume what they get. /45/ Sales proceeds are cash, just like most other income, and they are not likely to be restricted as to use. Cash can be expected to increase household consumption quite dramatically. 

 

   The Tax Reform Act of 1986 also plausibly increased the proportion of capital gains proceeds that are consumed by making it more expensive to borrow to have the liquidity to consume. The act phased out the deduction of interest incurred for consumption. /46/ Prior to 1986, some consumption could be met by borrowing on appreciated assets; after 1986, without the interest deduction, the borrowing becomes too expensive and the owner must sell appreciated assets to satisfy his consumption demand. 

 

   The loss of the interest deduction also commonly means that it is rational to sell assets to pay off past consumption borrowing. The combination of high borrowing interest rates and nondeductibility of consumption interest means that it is commonly cheaper to liquidate even preferentially taxed investments, rather than to continue to carry expensive consumption borrowing. Large sales to pay off prior consumption borrowing should be classified as consumption and not reinvestment in determining whether a preferential rate applies.

 

     Given the state of the data, assumptions about what portion of gains are consumed will have to depend in part on theory. As argued in the next section, sales to reinvest are not likely where the tax at stake is significant. Consumption is usually a necessary motive for a sale.

 

 D. Consumption of Capital Gains in an 'Efficient Market' 

 

   1. The irrationality of sales to reinvest.Consumption forms a necessary motivation for many sales qualifying for capital gain treatment because a sale solely to reinvest the proceeds is often not rational. If returns from a new investment come into equilibrium with returns from an old investment, an investor cannot profit by moving from the old to the new investment. Perfect equilibrium among returns and prices would drive out all sales to reinvest. While returns in the world do not reach perfect equilibrium, returns yearn toward equilibrium with enough vigor to dampen the rationality of sales to reinvest, at least where there is significant tax or other transaction costs at issue. 

 

   Proponents of the lock-in argument assume that new investments with better returns are available to a seller. /47/ They also sometimes assume that the market adjusts only slowly to news, so that an investor can get a bargain by changing investments before the market reacts. A recent model of stock trading, for instance, constructed by Donald Kiefer of the Congressional Research Service to study lock-in, excludes consumption of the sales proceeds. Kiefer then explains sales that in fact occur by assuming a very slow 7.5 percent per year movement of returns on some stocks toward the general fair market rate of return. /48/ A rapid adjustment of sale prices to news would drive out the bargains Keifer assumes that traders have available to them, thereby extinguishing all sales under his model. /49/ 

 

   If, however, there is anything to the 'efficient market thesis,' the assumptions that higher returns are available by switching investments and that price adjustments will be slow are dubious. In bargaining between buyers and sellers, prices of assets tend to adjust so that the returns from all investments, adjusted for risk, reach equilibrium with each other. In the stock market or other efficient markets, the adjustment of prices to news is extraordinarily fast. Once relative prices of investments have adjusted, an investor who must bear any material tax or other transaction costs in the trade cannot come out ahead by selling one investment solely to reinvest the proceeds of sale in another investment. If sales solely to reinvest are irrational, then consumption becomes a necessary element. Price adjustments do not plausibly make all sales to reinvest irrational, but they do put stringent limits on the scope of rational trades.

________________________________________________________________________________

 

 

          A Numerical Example. Assume, for instance, that Investor

     Able owns an acre of irrigated farmland 'A' that will give a

     harvest of $ 10x per year in perpetuity, net of all expenses. At

     a 10-percent discount rate, the acre is worth $ 100x. Investor

     Beta owns four acres of farmland 'B,' which is also worth $ 100.

     Farmland B is drier than A so that it takes four acres of B to

     generate $ 10x net per year in perpetuity. As far as anyone

     knows, farmland A and B are of equal risk. Assume there are

     costs of $ 10x to sell one land and purchase the other,

     including, for instance, taxes on the realized gain, costs for

     information about farmlands, and brokers' fees.

          If Able sells farmland A to get farmland B, A will get

     $ 100x from the sale, but will lose the $ 10x costs of the

     transaction. Able will then end up with $ 90x invested in

     farmland B (for 9/10ths of four acres) and will get a harvest of

     only $ 9x per year. It is far better for Able to stay put and get

     $ 10x per year from A.

          The same is true for Beta. Although farmland A is more

     productive per acre than the land that Beta holds, by selling

     acres of B to buy A, Beta will drop from $ 100x to $ 90x invested

     and from $ 10x to $ 9x in net annual harvest.

          Now assume that the price of irrigation water goes up by $ 1

     an acre. Farmland A' then generates only $ 9 harvest per year,

     after deducting the new cost of water. Assume also that new

     hybrids and dry farming techniques allow farmland B' to yield

     $ 11x harvest for the four acres. Should investor Able' go from

     A' to B', now that B' is better?

          If buyers bidding for A' know about the new water prices,

     they will pay only $ 90x at 10-percent discount rates for a $ 9x

     perpetual stream of income. If costs of sale are still $ 10x,

     then investor Able' will have only $ 80x left to invest in B'. If

     Beta', the current owner of B', knows about the new dry hybrids

     and dry farming techniques, she will sell the $ 11x income stream

     for nothing less than $ 110x. Beta' will sell only 8/11 of B' for

     $ 80x, leaving investor Able' with only $ 8x in annual harvest. It

     is better for Able' to keep farmland A' with its $ 9x per year.

     Able' has suffered a loss by

     the change in events, but she does better to absorb the loss,

     than to change investments incurring further costs. /50/

________________________________________________________________________________

 

   Because the 'efficient market thesis' has a point, an investor should assume that the price of stock sold on an established market accurately reflects the returns that can reasonably be expected to be made on the stock. There is a substantial volume of empirical research suggesting that the stock market is an informationally 'efficient market.' /51/ In an 'efficient market,' the price of a stock quickly adjusts to information about the value of stock as it becomes available. /52/ Once prices reflect available information, further changes are random, absent new information. A portfolio selected by a blindfolded chimpanzee throwing darts at a list of stocks will perform as well as a list chosen by the experts. /53/ The current price can certainly prove to be wrong -- it reflects the weighted average of many estimates about the probabilities of uncertain future events. /54/ But, no investor in an efficient market can be confident beforehand in what direction the price will be wrong. /55/ 'There is no other proposition in economics,' it is said, 'which has more empirical research supporting it than the efficient markets hypothesis.' /56/

 

   Kiefer's explanation of sales that take place by appealing to a slow, multiyear process by which returns approach the general rate of return is inconsistent with the rapid adjustments in price that in fact take place. If anything, the stock markets are getting faster in their speed of adjustment. Foot and Perold, /57/ for instance, looked at changes in price over 15-minute intervals in the price of marketed stock over the period from 1983 to 1989. In the early years of their study, they found that stock prices exhibited trend lines. The change in price over a prior 15 minutes was a good predictor for the change in price that would occur in the next 15 minutes. By the late 1980s, however, the trend lines were weak or nonexistent and the correlation of one change to the last was nearly random. The existence of index funds and other financial innovations has meant that news is disseminated to and digested by the market far more rapidly than ever. Stock market prices seem to adjust to news and reach randomness in some period of less than 15 minutes. /58/ 

 

   The mechanism by which the market adjusts to value places limits on how closely the price can approach true value. Information about value is itself a commodity. If the costs to find out about news and to effect trades are greater than the available bargain, no buyer can profit from a trade and the disparity between  price and value will remain. Similarly, no seller would incur the costs of research to determine that the stock is overpriced if the costs of the research are larger than the premium over real value that he might achieve. The cost of research thus creates a limit on how close the market comes in its approach to true value. /59/ 

 

   Tax also affects whether it is rational to sell to reinvest the sales proceeds. Many sales for reinvestment are made by taxpayers who bear little or no tax. Much of the stock market is now made up of tax-exempt institutions, such as pensions and charities, and of low effective tax entities, such as insurance companies and banks. /60/ In addition, individual investors with capital losses from unrelated transactions may sell appreciated property without tax to absorb their otherwise unrecognizable capital losses. /61/ For short-term traders who have small gains on any transaction, there is tax, but the tax on the small gains is so trivial in proportion to the sales proceeds that it will not prevent trades. Because big institutional traders can amortize their research costs over many shares and get the lowest price per share with mega-block trading, they will undoubtedly often be the marginal buyer who sets how closely price can approximate knowable value. The activity of low-cost, low-tax traders seems sufficient to allow the market to adjust to news and stay efficient. 

 

   Once taxes and other transaction costs become a significant fraction of the sales proceeds, however, sales for reinvestment become less likely. A seller might rely on some optimism by the buyers, but it is unlikely they could convince buyers to be euphoric enough to pay for substantial transaction costs. Similarly, a good analyst may have small advantages in markets that only partially incorporate knowable information, but substantial transaction costs will wash away the advantage. If the capital gain tax and other transaction costs constitute 20 percent of the sales proceeds, an investor has to be far more confident of her judgment about mispricing than is justified by a competitive market if she is to come out ahead even after forfeiting 20 percent of her capital. /62/ In sum, as the tax at issue grows more substantial, reinvestment is less likely to be a sufficient motive to justify the sale. 

 

   2. A proposal for cuts from the bottom of tax. Within the confines of the efficient market thesis, it might be possible to design a tax cut that focuses on sales in which reinvestment is somewhat more likely than usual. A tax cut focused on sales in which reinvestment is more likely would be tax cut 'at the bottom,' that is, the cut would forgive the tax only when taxes, computed under normal means, are a small fraction of the proceeds of sale. One might, for instance, provide that any taxpayer whose tax is under two to three percent of the sales proceeds would have tax on sale forgiven in full. After a tax cut 'at the bottom,' an investor might fit within the range where marginally better information or other factors allow a sale to reinvest to be rational. 

 

   A tax cut 'off the top' of a large amount of tax due, by contrast, will usually leave the investor in a position in which even after the tax cut, it is still irrational to sell property to reinvest. A reduction in tax from 34 percent to 15.4 percent of the proceeds of sale, for instance, would probably leave tax and other transaction costs too high to justify a shift. If the cut in tax does not bring the transaction within the range where a sale to reinvestment is rational, the tax cut will be wasted because it will benefit consumption. /63/ For the same amount committed to the cut, a tax cut at the bottom would more efficiently serve sales to reinvest than would a tax cut off the top.

 

     3. The impact of rational sales to reinvest. 

 

   (a) Inefficient markets. Sales to reinvest are arguably more likely if the asset is sold on a market that is not efficient. Not all assets are sold on markets that digest information as quickly and accurately as do the stock markets. For stock and commodity markets, there is a large cadre of professional traders who rapidly digest fairly subtle information within a sophisticated framework and who can intervene with enough capital to change the market price. /64/ Stocks are fungible and evaluated by standardized measures. Stock or bond traders can shift massive amounts of capital at a push of a computer button in reaction to changes in risks or interest rates that amount to a fraction of a percentage point. The empirical research shows efficiency on the stock market, not on other assets. 

 

   Other markets might well react more slowly, although undoubtedly more rapidly than the 7.5-percent change in price per year that Kiefer assumes. Real assets, like buildings and business assets, have too many idiosyncratic features to be evaluated by standardized measures or sold on a large scale. Where the market is not that large, it will not support the large cadre of information brokers needed to collect and analyze news. Real assets, such as business equipment or buildings, cannot be created or replaced so rapidly, nor even purchased or sold so rapidly to reflect subtle changes in the available information. /65/ 

 

   The competitive forces that put prices and returns into equilibrium in efficient markets, however, also work in less efficient markets. The price of any asset moves toward efficiency because perceptible bargains and premiums shrink or disappear in bargaining between self-interested sellers and buyers. Sellers eschew perceptible bargains and buyers eschew perceptible premiums over value no matter what the asset is. To get a true bargain, a buyer must not only outsmart the current owner of the property, but must also outsmart all other potential buyers as well. To assume that a sale to reinvest would be rational, an investor must be confident that she can either overcharge her buyer or outsmart the whole world as to the replacement. As in efficient markets, one cannot assume, in absence of a rollover requirement, that sales benefiting from a cut in capital gains tax are sales made for reinvestment. Buyers and sellers may be groping in the dark, but the fog over information does not necessarily mean that sales to reinvest are rational, given the competition on both sides for a good price. 

 

   (b) Irrational trades. Some sales to reinvest may be irrational and yet occur even in the face of substantial tax on the sale. Unsophisticated investors seem to undertake a good deal of irrational churning, that is, sales to reinvest that leave the investor with wealth reduced by the broker's fees, taxes, and other costs of the transaction. Many investors ignore the basic rules of investing. They do not diversify; they are often 'noise traders,' buying on the basis of tips that are either untrue or already reflected in price. They often pay high fees to mutual funds or advisers who do not beat the market. /66/ Much of the activity may be harmless amusement. The efficient market thesis implies that stock purchased will ordinarily be as valuable as the stock sold, so that the investor will lose only the identifiable transaction costs in the trade. But it is also plausible that noise traders increase risk in the market and impede the adjustment of prices to real economic events. /67/ Even if irrational activity is harmless sport, however, it can reasonably bear normal taxes. Proponents of state-run lotteries, for instance, argue that a state-run lottery with odds stacked very much against the 'investor' is a 'painless tax' because the purchase of a lottery ticket is a voluntary act, even if it is irrational. /68/ Irrational turnovers of assets may just be a kind of high-class lottery. 

 

   (c) Rational reinvestment. Sales to reinvest are commonly rational, notwithstanding transaction costs and rapid price adjustments in a competitive market. The rationality of sale to reinvest under the circumstances does not preclude consumption of the sale proceeds, but it does mean that reinvestment of the sale proceeds might occur. One can list a number of phenomena that might justify a rational trade: 

 

   (i) Superior information or analysis. Some investors, especially insiders, might in fact have better information or be better able to analyze available information than the marginal investors whose judgments set prices. It is a tautology that the market as a whole cannot do better than the market average. For every investor who does better than average, someone must do worse. But investors with better information or analysis might be able to improve return by selling to reinvest. If a trader in corporate stock is relying on insider information, we might want to maintain a high rate of tax to protect public shareholders hurt by disloyal  insiders. /69/ Not all trades based on superior information or analysis, however, are a breach of loyalty. 

 

   (ii) Bubbles. Market prices may be fundamentally inaccurate because they accurately reflect the way market participants evaluate risks. Humans tend to misevaluate risks by overweighing or overgeneralizing from the most recent news. Emotional overvaluation may result in bubbles of uncertain duration, in which prices rise rapidly and later collapse. Unemotional, rational investors might not be able to correct bubbles because even fairly dramatic differences in price may make such a small difference in rate of return. Their investments against the trend may not be profitable enough to justify their risk, especially because they have no way of knowing how long a bubble will last. /70/ 

 

   (iii) Changing utilities. The utility of real assets to the owner changes because of changes in the owner's needs. Almost all real assets have different values to different owners. Differing utilities do not necessarily justify resales because only the highest utility owners will buy the asset in the first place. But, tastes or needs or the organization into which the asset fits can change so that the investor optimally sells the asset to use the proceeds elsewhere. The owner of a widget factory may be the highest and best user of a new, improved widget machine, for instance, but if the owner goes broke or retires or decides to change her line of business, then the machine is best sold for whatever the market will bear. 

 

   (iv) Diversification. A rational investor diversifies her portfolio with investments that tend to move in opposite directions from each other. If one does not know whether there will be war or peace, for instance, one should buy both butter stocks and gun stocks so that losses on the one will be offset by gains on the other. /71/ What investments are needed for diversification depend on the investor's existing portfolio. Different investors will thus have different values for an investment presenting the same nominal risk to all. When a single investment has different value or gives different risk adjusted returns to different investors, no single price for the investment can take away all bargain values. Investments can become too concentrated or too uniform in their risks and diversification can  change by chance and inadvertence so that an investor commonly needs to sell assets to rediversify her portfolio, even though the prices of assets reflect all available information. 

 

   (v) Individualized risk and discount rates. Choices about when money is needed and how much risk is tolerable are choices individual to the investor, and cannot be met by a uniform market discount rate. So long as an investor is charged a premium for borrowing over what can be made by investing, or borrowing is rationed, an investor's choice of what discount rate to use will depend upon personal decisions about when and how desperately money is needed. /72/ A single investment cannot satisfy both the 'widows and orphans in need of sure and immediate succor and...the cowboy capitalist willing to wait for the big score.' /73/ Either a change in the risk or discount rate of the investment or a change in the investor's needs would make a sale for reinvestment rational even when prices reflect all available information. 

 

   Some of the rational demand for sales to reinvest can be fit into nontaxable transactions even under current law. Property that is leased, for instance, may be rerented to another tenant when needs change, without a taxable sale. When leasing is available, the user of property does not need to be the tax owner. The economic ownership of urban buildings, for instance, is usually fragmented into a number of tiers, bearing different risks and rewards, and only the ultimate tenant, with what is usually the shortest-term interest in the property, uses the space. /74/ Tax law allows the user of the property who is nominally only a tenant to bear almost all of the risks and benefits of real ownership. /75/ The change of a lease or replacement of a tenant is not a taxable event to the 'owner,' so that the use and risks of the asset can be shifted to a new higher user without tax if the utility of the property to the user changes. /76/ Section 1031, moreover, providing for nonrecognition of tax in like-kind exchanges, allows a seller of real estate who is willing to keep sales proceeds invested in real estate to move from one property to another without tax. /77/ Capital losses, finally, are usually available to allow an investor whose portfolio becomes too concentrated by reason of market fluctuations to sell gain assets without tax. High appreciation that generates concentrated portfolios is associated with high volatility that generates capital losses. The losses should ordinarily be sufficient to allow the turnovers needed to maintain diversity. /78/ Finally, the capital losses generated by fluctuations can only be used to exempt capital gains from tax, with a modest exception. /79/ For those transactions that can now, with some planning, be channeled into tax-exempt form, a merely preferential capital gains tax rate on sales to reinvest will not be that attractive. The opportunities for tax-free shifts in capital or change in use of assets, however, do not fully cover all of the sales that might rationally be made to reinvest in another asset. 

 

   The central point is not that reinvestment of sale proceeds bearing significant capital gain tax will never occur, but only that reinvestment cannot account for all of the proceeds of all sales and that consumption is thus sometimes a necessary motive for sales. The possibility  of a rational sale for reinvestment, moreover, does not mean that the sale in fact was fully for reinvestment or that no consumption of capital occurred. Consumption sometimes occurs even when it is not logically necessary for the sale. 

 

   A rollover requirement, providing that preferential rates on capital gain sales will be available only to the extent that the proceeds of the sale are reinvested in productive investments, would mandate reinvestment. Proof that sales proceeds were reinvested should be required even where sales to reinvest are rational. The rationality of a sale to reinvestment does not mean that they in fact were for reinvestment. Sales proceeds can and are consumed even when they might have been reinvested. Since the arguments for preferential tax rates for capital gains depends upon a premise that the sales proceeds will be reinvested and not consumed, a reinvestment requirement is a necessary part of the case for lower tax rates. A rollover requirement would exclude only proceeds which ought to be excluded from a soundly designed preferential tax.

 

 E. Exemption for Rollover Investments 

 

   There have been proposals in the past to exempt all tax on sales if the gains are reinvested, while taxing at ordinary rates all proceeds withdrawn from investment for consumption. /80/ Complete tax exemption for rolled-over sales proceeds is not on the current political agenda in part because there is no possibility of increasing government revenue from a tax rate of zero, no matter how many more sales are induced. 

 

   Nor would the exemption for rollover sales be a fair idea. In an income tax system, most investments are made with 'hard,' after-tax money. An exemption for previously untaxed gain rolled over into new investments would give investors with existing appreciated capital a source of 'soft' untaxed money to put into any investment. /81/ The soft money privilege is a benefit that can be expected to be as valuable as exempting income from tax. /82/ A tax system that gave the soft money privilege to investors with existing wealth, while requiring new investors to invest with hard money, would not pass even rudimentary notions of fairness. /83/ As explained in the first article of this trilogy, 'The Undertaxation of Holding Gains,' gain built into unsold property is undertaxed, relative to other investments. Extending the exemption of unsold gain to realized gain would make the undertaxation worse.

 

 F. Conclusion 

 

   The most influential arguments in favor of preferential rates for capital gains rest on the assumption that the proceeds of a capital gains sale will be reinvested in a higher yielding, more productive investment. Capital gains taxes are considered to be toll charges on the movement of capital to more efficient uses, which lock in capital in old, less efficient investments. But nothing in law or nature in fact requires capital gain to be reinvested. In absence of an explicit requirement, a significant portion, probably most of capital benefiting from a reduction in the tax, will be diverted to consumption or consumer durables and will not be returned to productive investment. Taxpayers consume a large portion of cash receipts. In an efficient market, sales to reinvest are not rational where transaction costs are significant, so that consumption is often a necessary motive for a sale. Taxes on disinvested capital are a most rational luxury tax. 

 

   The administration proposals, accordingly, should condition their preferential tax rate on a requirement that the proceeds of a capital gain sale be rolled over into some productive investment. Tax cuts targeted to the lock-in problem, moreover, need to be reductions of the last significant amount of tax rather than reductions in rate which leave significant tax to be paid.

 

 

________________________________________________________________________________

 

                                   FOOTNOTES

________________________________________________________________________________

 

   /1/ President Bush, 1992 State of the Union Address, 92 Tax Notes Today Electronic Edition 92 TNT 20-5 (Jan. 29, 1992); Department of the Treasury, General Explanation of the President's Budget Proposals for Fiscal Year 1993 Affecting Receipts at 7 (Jan. 1992). The 15.4-percent maximum rate is reached by excluding 45 percent of the gain and imposing a maximum rate of 28 percent to the remaining 55 percent. 

 

   /2/ Johnson, 'Undertaxation of Holding Gains,' Tax Notes, May 11, 1992, p. 807 (hereinafter 'Undertaxation of Holding Gains'). 

 

   /3/ Johnson, 'The Private Advantage of Money-Losing Investments Under Cut-Rate Capital Gain,' Tax Notes, May 25, 1992 (hereinafter, 'Private Advantage'). 

 

   /4/ See, e.g., Cong. Budget Off., Capital Gains Taxes in the Short Run (1991); Department of the Treasury, General Explanation of the President's Budget Proposals for Fiscal Year 1993 Affecting Receipts at 4 (Feb. 1, 1992); Schmalbeck, 'The Uneasy Case for a Lower Capital Gains Tax: Why Not the Second Best,' 48 Tax Notes 195, 200 (1990); Blum, 'Rollover, An Alternative Treatment of Capital Gains,' 41 Tax L. Rev. 383, 387-388 (1986); R. Goode, The Individual Income Tax 197-208 (rev. ed. 1976); Wetzler, 'Capital Gains and Losses,' in Comprehensive Income Taxation 115, 135-138 (Pechman, J. ed. 1977); Holt & Shelton, 'The Lock-In Effect of the Capital Gains Tax,' 15 Nat. Tax J. 337 (1962); Brown, 'The Lock-In Problem,' Papers on Federal Tax Policy for Economic Growth and Stability,' Subcommittee on Tax Policy, Joint Comm. on the Economic Report, 83d Cong., 1st Sess. at 381 (1955). 

 

   /5/ The following table shows how much better the return on the new investment must be than the return on the existing investment if the investor is going to sell the old asset to buy a new one. The required new return is stated as a multiple of the return on the old investment. The tax on a sale depends upon tax rate and also upon what fraction of the sales proceeds is taxable gain, so that the required multiple varies with both tax rate and the fraction of the sales proceeds which taxable gain represents.

________________________________________________________________________________

 

           TABLE. ANNUAL RATE OF RETURN ON NEW INVESTMENT

               REQUIRED TO MAKE A SWITCH ADVANTAGEOUS,

              SELECTED TAX RATES AND GAINS ON OLD ASSET

                    AS FRACTION OF MARKET VALUE.

           (Required new rate of return stated as multiple

                    of return on old investment.)

Tax Rate

on Sale      1/10      3/10      5/10      7/10      9/10

10%          1.01      1.031     1.053     1.075     1.099

25%          1.026     1.081     1.143     1.212     1.290

35%          1.036     1.117     1.212     1.324     1.460

50%          1.053     1.176     1.333     1.538     1.818

________________________________________________________________________________

 

   R. Goode, supra note 4, at 200. 

 

   The multiple of the old rate of return that the new rate must meet is calculated as the ratio,

 

 

________________________________________________________________________________

 

                                M / [M-t(M-C)],

________________________________________________________________________________

 

 

 

 where M is the current value of the old asset, t is the capital gains tax rate, C is the cost basis of the old asset, so that M-C is the taxable gain on sale. Id. at 200, n.30. 

 

   Cong. Budget Office, supra note 4 and Holt & Shelton, supra note 4, analyze the required rate of return on the new investment on the assumption that either the old or the new investment will be held until death. 

 

   /6/ Dodge, 'Restoring Preferential Capital Gain Treatment Under a Flat Rate Income Tax: Panacea or Placebo?' 44 Tax Notes 1133, 1137 (1989). See also, Blum, 'A Handy Summary of the Capital Gains Arguments,' 35 Taxes 247, 257 (1957) (characterizing the lock-in argument as a 'formidable indictment'). 

 

   /7/ Wallich (Professor of Economics, Yale), 'Statement' in 2 General Tax Reform: Panel Discussions Before Comm. on Ways & Means: Capital Gains and Losses at 276, 93d Cong., 1st Sess. (Feb. 6, 1973) (asserting that 'the capital gains tax reduces the supply of savings because the gains upon which it falls tend to be predominantly saved,' but conceding that '[s]tatistical tests...increasingly find evidence of spending out of capital gain'). 

 

   /8/ Auerbach & Hassett, 'Corporate Savings and Shareholder Consumption,' in National Savings and Economic Performance 75 (B. Bernhein & J. Shoven ed. 1991) (arguing that consumption will not increase just because shareholders have involuntary cash realization of their wealth). Auerbach, the first listed coauthor, is the economist for the Staff of the Joint Committee on Taxation. 

 

   /9/ Kiefer, 'Lock-In Effect Within a Simple Model of Corporate Stock Trading,' 43 Nat. Tax J. 75, 76 (1990) (excluding sales for consumption from the model). 

 

   /10/ Cong. Budget Office, supra note 4. 

 

   /11/ See, e.g., Cong. Budget Office, supra note 4, at ix (arguing that limiting the cut to newly purchased assets is better than retrospective tax cut applied to old investments); Kotlikoff, 'The Crisis in U.S. Saving and Proposals to Address the Crisis,' 43 Nat. Tax J. 233, 239 (1990) (arguing that capital gain cut is poorly designed because of windfall benefit to those who accrue gains before act). 

 

   /12/ See Skinner & Feenberg, 'The Impact of the 1986 Tax Reform Act on Personal Savings,' Nat. Bur. of Econ. Research Working Paper No. 3257 at 10-17 (1990) (Consensus in the economic literature is that any positive response of savings to interest rate increase is 'fragile and fleeting.'); Hall, 'Intertemporal Substitution in Consumption,' 96 J. of Pol. Econ. 339 (1988) (finding no savings response to increased interest returns and explaining away apparent findings that savings respond to increased interest); Howrey & Hymans, 'The Measurement and Determination of Loanable-Funds Savings,' in What Should Be Taxed: Income or Expenditure? (J. Pechman, ed. 1980) (finding no response to increased interest). 

 

   /13/ See Bovenberg, 'Tax Policy and National Saving in the United States: A Survey,' 42 Nat. Tax J. 123, 128 (1989) (reviewing literature finding that high interest rates decrease savings); Bosworth, Burtless & Sabelhaus, 'The Decline in Saving: Evidence from Household Surveys,' 1991-1 Brookings Papers on Econ. Activity 183. 

 

   /14/ See 'Undertaxation of Holding Gains,' supra note 2, at 807; 'The Private Advantage of Money-Losing Investments,' supra note 3. 

 

   /15/ See, e.g., 'Undertaxation of Holding Gains,' supra note 2, at 21. 

 

   /16/ Auerbach, 'Capital Gains Taxation and Tax Reform,' 42 Nat. Tax J. 391, 391 (1989). 

 

   /17/ Schmalbeck, supra note 4, at 201. 

 

   /18/ Capital that a seller withdraws from an investment and consumes is replaced by capital that the buyer puts into the investment: a change in ownership of any asset or investment does not affect the total supply of capital. But the buyer's capital is a given, which cannot be affected by the tax treatment of the seller. If the buyer does not buy this investment, she will put her capital elsewhere. When buyer's capital is fixed, the seller's consumption of capital reduces aggregate capital. 

 

   /19/ Indeed, reducing private consumption is said to be the primary purpose of taxes. A. Lerner, The Economics of Control 307-308 (1944); Andrews, 'A Consumption-Type or Cash Flow Personal Income Tax,' 87 Harv. L. Rev. 1113, 1121 (1974). 

 

   /20/ Capital gains are, not surprisingly, gains achieved disproportionately by the wealthiest taxpayers, who by definition have the most property that might appreciate. Almost two-thirds of the benefits of a reduction in capital gains tax go to the top 9/10ths of one percent of the population; 80 percent of the tax benefits go to the top 3.3 percent of the population. The bottom 80 percent of the population gets only eight percent of the benefits and the bottom half of the population gets only five percent of the benefits. Letter of Robert Reischauer, Director, Congressional Budget Office, to Rep. William Gradison, R-Ohio, cited in Democratic Study Group, 'The Capital Gains Tax Cut,' Special Report No. 101-18 at 4 (Sept. 15, 1989).

 

   /21/ See, e.g., Sen, 'On Ignorance and Equal Distribution,' 63 Am. Econ. Rev. 1022 (1973) (showing that any decline in the marginal utility of extra income as wealth increases entails that equality of distribution of income will maximize aggregate utility). 

 

   /22/ Groves, Post War Taxation and Economic Progress 294 (1946). 

 

   /23/ Darman (Director of the Office of Management and Budget), Address to the National Press Club, July 20, 1990, 89 Tax Notes Today Electronic Edition 150-12 (decrying the 'Maypo Syndrome' in American life, i.e., the demand for immediate gratification). The Maypo Syndrome is named after a 1960s TV commercial that featured a baby crying 'I want my Maypo and I want it now.' 

 

   /24/ See Annotation, 'Rights of Life Tenant and Remainderman Inter Se Respecting Increase, Gains and Enhanced Value of the Estate,' 76 ALR 2d 162, 167 (1961); 51 Am. Jur., '2d Life Tenants and Remainderman,' section 162-171 (stating capital gains are presumed not to be distributable to income beneficiary); Kornhauser, 'The Origins of Capital Gains Taxation: What's Law Got To Do With It?' 39 S.W.L.J. 869 (1985); L. Seltzer, The Nature and Tax Treatment of Capital Gains and Losses 26-35 (1951). Cf. Isaacs, 'Principal -- Quantum or Res?' 46 Harv. L. Rev. 776 (1933) (discussing the ambiguity of whether capital refers to property or to cost invested in trust accounting). In Merchant's Loan & Trust Co. v. Smietanka, 255 U.S. 509 (1921), the Supreme Court held that capital, constitutionally immune from an 'income' tax, was the prior investment, rather than the starting fair market value of the sold property. Nevertheless, a number of instances of starting 'fair market value' basis remained after the decision. See Johnson, 'The Legitimacy of Basis from a Corporation's Own Stock,' 31 Am. J. of Tax Policy 1, +11-28 (forthcoming 1992). 

 

   /25/ Annotation, supra note 24, at 168 ('There can be no doubt' that instrument controls). 

 

   /26/ Within an income tax that attempts to be comprehensive, 'capital' is now largely synonymous with the tax term, 'basis,' and 'basis' is a monetary account that generally means amounts invested that have not yet been deducted and excludes appreciation of the property. See, e.g., IRC section 1022 (basis is in general cost) and section 1016 (basis must be adjusted downward for depreciation). Gains, even capital gains, are not 'capital' because the gains are the amount that is in excess of the taxpayer's basis -- even when it is appreciated corpus land that is being sold. IRC section 1001 (defining taxable gain as the excess of amount realized over basis). In 1921, the Supreme Court held that capital gains fell within the purview of the Sixteenth Amendment, authorizing a tax on income, even assuming that the income tax could not be applied to tax 'capital.' Merchants' Loan & Trust Co. v. Smietanka, 255 U.S. 509 (1921). Within the norms of an income tax, moreover, amounts that are invested have ordinarily been subjected to tax once before they ever became investable capital. See, e.g., Johnson, 'Soft Money Investing Under the Income Tax,' 1989 Ill. L. Rev. 1019 (1990). Labelling something as 'capital' is thus now a better reason to subject the item to tax or to ordinary rates than it is to exempt the item from tax or to reduce the rate of tax on the item. 

 

   /27/ Preferential rates for capital gain were first enacted in the Revenue Act of 1921, 42 Stat. 227, 232, the same year the Supreme Court held that capital gains were not capital immune from income tax. See Kornhauser, supra note 24, at 893- 896, 901; L. Seltzer, supra note 24, at 26-35.

 

   /28/ IRC section 1033 (insurance or condemnation proceeds from 'involuntary conversion' are not taxed if proceeds are rolled over into replacement property); section 1034 (proceeds of sale of principal residence are not recognized if reinvested in replacement); sections 402(a)(5)-(7) (lump sum or plan termination distributions from qualified retirement plan are tax- free if rolled over into new qualified plan). 

 

   /29/ Blum, supra note 4, discusses the technical issues that would need to be resolved to implement a rollover requirement. 

 

   It would not be necessary to require that capital rolled over into a new investment stay there for any length of time. If the capital is quickly withdrawn and reinvested, still the capital remains in productive use. If the capital is withdrawn for consumption or a consumption durable, the ordinary rates will be imposed on the sale and the initial sale and reinvestment will have delayed the ordinary tax for only a short time. 

 

   /30/ See Department of the Treasury, supra note 1, at 7. 

 

   /31/ See, W. Andrews, ALI Fed. Income Tax Project (June 1, 1989); Warren, 'Recent Corp. Restructuring and the Corporate Tax System,' Tax Notes, Feb. 6, 1989, p. 715; Joint Committee on Taxation, Federal Income Tax Aspects of Corporate Financial Structures (Comm. Print Jan. 18, 1989). 

 

   /32/ Poterba, 'Dividends, Capital Gain and the Corporate Veil: Evidence from Britain, Canada and the United States,' in National Savings and Economic Performance 49, 63 (B. Bernhein & J. Shoven ed. 1991). 

 

   The corporate buyouts were structured as mergers to be involuntary for any one shareholder once the requisite number of shareholders approved of the transaction. Auerbach & Hassett, 'Corporate Savings and Shareholder Consumption,' in National Savings and Economic Performance 75 (B. Bernhein & J. Shoven ed. 1991) argue that there is no increase in consumption from the buyout, apart from its effect on shareholder wealth, in part because the buyouts are involuntary. Under a life cycle model of savings and consumption, consumers average out consumption of the aggregate income over their lives. Auerbach and Hassett argue under the model that no rational consumer would increase consumption merely because her wealth is reduced to cash, in part because a consumer could always sell wealth to consume. Under the efficient market thesis, discussed in text in section D, no one would sell except to consume. The Auerbach and Hassett findings also appear to be reconcilable with Poterba's because Auerbach and Hassett exclude consumer durables from their definition of consumption, whereas Poterba does not. 

 

   /33/ Bernhein, 'Introduction' in National Savings and Economic Performance 1, 5 (B. Bernhein and J. Shoven ed. 1991). 

 

   /34/ Modigliani, 'The Life-Cycle Hypothesis of Saving, the Demand for Wealth and the Supply of Capital,' 33 Social Research 160 (1966). See also Auerbach & Hassett, supra note 32, arguing that under the life cycle hypothesis, cash is no reason to increase consumption. 

 

   /35/ Carroll and Summers, 'Consumption Growth Parallels Income Growth: Some New Evidence,' Nat. Bur. Econ. Research Working Paper No. 3090 (1989) (permanent income or life-cycle hypothesis is inconsistent with the grossest features of the data on actual consumption patterns); Flavin, 'The Adjustment of Consumption to Changing Expectations About Future Income,' 89 J. of Pol. Econ. 974, 1006 (1981) (data rejecting the permanent income or life-cycle hypothesis is significant); Skinner, 'Risky Income, Life Cycle Consumption and Precautionary Savings,' J. of Monetary Econ. 237 (1988). 

 

   /36/ Campbell and Mankiw, 'Permanent Income, Current Income and Consumption,' Nat. Bur. of Econ. Research Working Paper No. 2436 (1987) estimate that 40 to 50 percent of income is received by individuals who consume according to current income rather than according to wealth or permanent income. 

 

   /37/ See sources cited, supra, note 35 and Bosworth, Burtless & Sabelhaus, 'The Decline in Saving: Evidence from Household Surveys,' 1991-1 Brookings Papers on Econ. Activity 226 (summarizing the research). 

 

   /38/ Carroll and Summers, supra note 35, at 27-28. 

 

   /39/ Skinner & Feenberg, supra note 12, at 25, 28 determined that for every dollar of capital gain reported in 1986, there was 38 cents more interest income in the following year but no added dividend income. Since 3.8 percent was about half of the prevailing interest rate, they conclude that half of their gains represent a shift from non-income reporting investments such as real estate and tax shelters into interest bearing assets, such as savings accounts. The increases in interest did not, however, survive into future years. (The year 1986 was a record shattering high year for sales of assets. Cong. Budget Office, supra note 4, at 36. Rational high- bracket investors, ahead of the trends, should have been selling off their tax shelters in 1986 in anticipation of the Tax Reform Act. Since the 1986 act raised capital gains rates, sellers with pent up consumption needs should have been liquidating their estates to beat the tax rise.) The Skinner and Feenberg data did not capture all investments nor all consumption. Shifts from a tax shelter partnership into a yacht or other consumer durable, for instance, would not have been captured on either side of the shift. Their data does not support a hypothesis that capital gain sales reduce taxable investment income. The data is consistent with many hypotheses, however, including the hypothesis that atypical spurts are parked for a while and that gains are invested in consumer durables. 

 

   /40/ Consumption that does not fall into the same period that the sale takes place presents heuristic problems because it can be difficult to match the sale with the consumption that motivated it. Rates of savings over too short a period will understate consumption increases where a seller temporarily reinvests the sales proceeds. The sales may generate much greater consumption, meaningfully measured in time value terms, but only within a time horizon that is not easy to identify. 

 

   /41/ Calculated from Bureau of Economic Analysis, 'Table 694, The National Income and Product Accounts, Relation of GNP, Net National Product, National Income, Personal Income, Disposable Income and Personal Savings: 1970-1988,' in Dept. of Commerce, Statistical Abstract of the United States, 1990 at 428 ($144.7/3477.8 billion). For the five years, 1984-1988, individuals saved 3.9 percent of disposable personal income. Id. (689.7 savings/17,632.4 billion.) 

 

   The NIPA figures treat purchase of consumer durables as consumption. Consumer durables are investments because they give continuing value (although as noted in text accompanying supra note 30, they are not the kind of investments that can be expected to increase productivity or international competitiveness or to generated externalities worth subsidizing). Kotlikoff, supra note 11, at 236, finds savings rates that are 150 percent higher when durables are considered savings. The NIPA figures also exclude capital gains from disposable personal income. Skinner, 'Precautionary Saving, Wealth Accumulation, and the Savings Downturn of the 1980s,' 43 Nat. Tax J. 247, 250, Table 1 (1990), adds realized capital gain to disposable personal income and finds savings to be four percent of disposable income in 1985-89. The adjustments are not material in determining what to expect regarding consumption of realized capital gain. No matter what the adjustments, a high 94 to 96 cents of every dollar of available income is consumed. 

 

   /42/ Current U.S. savings rates are a quarter of what they were in the 1950s through 1980s. Kotlikoff, supra note 11, 43 Nat. Tax J. at 233. 

 

   /43/ Summers & Carroll, 'Why Is U.S. National Saving So Low,' 1987-2 Brookings Papers on Econ. Activity 607. 

 

   /44/ See authorities cited supra note 12, especially Hall, 96 J. of Pol. Econ. 339 (emphasizing the historically high real interest rates). 

 

   /45/ The marginal propensity to consume is substantial in magnitude as income rises, even though the marginal propensity to consume is less than average when income is high. See Darby, 'Consumption Function,' in 1 The New Palgrave: A Dictionary of Economics (1987). 

 

   A tax cut aiding investment, moreover, should have a closer relationship to reinvestment than merely that it enriches rich people. It does not seem a sufficient justification for government that it gives more income to rich people, without restrictions, simply because they are rich. 

 

   /46/ IRC section 163(h)(1) & (2), as enacted by Pub. L. No. 99- 514 section 511(b)). The deduction of personal interest was phased out over the years 1987 through 1990. IRC section 163(h)(5) by reference to section 163(d)(6)(B). IRC section 163(h)(3), however, allows deduction of interest on up to $1.1 million of home mortgage debt. 

 

   /47/ See, e.g., Goode, supra note 4, at 200 (showing how much higher the return on the new investment must be to allow an advantageous switch). 

 

   /48/ Kiefer, supra note 9, at 76 (excluding consumption), 79 (assuming stock prices move toward equilibrium rates at only 7.5 percent per year). 

 

   /49/ But see id. at 93, n.16 (shareholders might trade upon irrational expectations). 

 

   /50/ A change in investments is not rational if price reflects value, even if values change with time. Assume, for instance, that the dearer water and better dry farming come in only over time and keep on going, such that the harvest on farmland A' declines by five percent per year perpetually (e.g., $10 to $9.50 to $9.05 to $8.57 to $8.14, etc.), and that new dry tolerant farming means that the harvests on farmland B' increase by two percent per year perpetually. Does it not make sense to sell the declining investment to buy the climbing one? Not if both buyers and sellers know of the trend. A harvest of $10 that declines perpetually by five percent per year is worth $66.67 at a 10-percent discount rate. The formula for a perpetual annuity that changes perpetually by a fixed percent is A/d- g, where A is the starting annuity amount -- $10 here -- d is the discount rate -- 10 percent here -- and g is the rate of growth or shrinkage. See J. Van Horne, Financial Management and Policy at 32 (5th ed. 1980). Here, $10/[10%-(-5%)] = $10/15% = $66.67. A harvest that rises perpetually at two percent per year from $10 is worth $125 at a 10-percent discount rate. The formula, A/d-g, where g is +2% becomes $10 / [10%-2%] = $10/8% = $125. If investor Able' bears $10 cost in the trade, she can buy only $56.67 worth of B' which is only 56.67/125 or 45 percent of farmland B'. If she had stayed in investment A', then she would have $10 in the first year and hold an investment worth $66.67. Again Able' has taken a loss on A', but incurring costs to change investments just adds extra costs and increases her loss. 

 

   /51/ Malkiel, 'Is the Stock Market Efficient?' Science 1313, 1317 (March 10, 1989). See also Fama, 'Efficient Capital Markets: A Review of Theory and Empirical Work,' 25 J. Finance 383 (1970); Note, 'The Efficient Capital Market Hypothesis, Economic Theory and the Regulation of the Securities Industry,' 29 Stan. L. Rev. 1031 (1977). The evidence that the stock markets are efficient includes evidence that future prices of stocks are independent of past prices, that prices adjust rapidly to newly disclosed information and sometimes anticipate (or incorporate leaked) information, that changes in accounting methods do not affect stock prices, and that market professionals cannot consistently perform better, after research expenses, than randomly selected diversified portfolios do. 

 

   /52/ In theory, the value of stock is the discounted present value of future distributions on the stock. The price that future buyers will pay for the stock is in turn just a proxy for the distributions that buyers can expect after they buy the stock. See, e.g., Debondt & Thaler, 'Anomalies: A Mean-Reverting Walk Down Wall Street,' J. Econ. Perspectives 189 (Winter 1989); Pogue & Lall, 'Corporate Finance: An Overview in S. Myers,' Modern Developments in Financial Management at 27-28 (1976). Measurements of the efficiency of the market also simultaneously test the pricing of stock according to discounted value of cash (see, e.g., G. Foster, Financial Statement Analysis 363 (1978)), that is, stock prices, under the efficient market thesis, reflect what is known publicly about future distributions and appropriate discount rates. 

 

   /53/ Malkiel, supra note 51, at 1313. 

 

   /54/ See, e.g., L. Thurow, Dangerous Currents: The State of Economics 152 (1983) (price reflects a wide distribution of possibilities with a long tail such that, for instance, investors who owned IBM stock in the 1930s would be very wealthy by now). 

 

   /55/ Verrecchia, 'Consensus Beliefs, Information Acquisition and Market Information Efficiency,' 70 Am. Econ. Rev. 874 (1980) argues that the weighted average reflected in the market will be better than any of the participants' estimates about the future. 

 

   /56/ Jensen, 'Some Anomalous Evidence Regarding Market Efficiency,' 6 J. Fin. Econ. 95, 95 (1978). See also id. at 96: 'In the literature of finance, accounting, and the economics of uncertainty, the Efficient Market Hypothesis is accepted as a fact of life, and a scholar who purports to model behavior in a manner which violates it faces a difficult task of justification.'

 

   /57/ Foot & Perold, 'New Trading Practices and Short-Run Market Efficiency,' NBER Working Paper No. 3498 (Oct. 1990). 

 

   /58/ The absence of a negative correlation between 15-minute price changes indicates the market was also not over-reacting and then correcting itself. Id. at 1. 

 

   /59/ See Easterbrook & Fischel, 'The Proper Role of a Target's Management in Responding to a Tender Offer,' 94 Harv. L. Rev. 1161, 1166 (1981); Grossman & Stiglitz, 'On the Impossibility of Informationally Efficient Markets,' 70 Am. Econ. Rev. 393 (1980) (arguing that the markets reach not perfect knowledge, but an 'equilibrium degree of disequilibrium'). See also Gilson & Kraakman, 'The Mechanisms of Market Efficiency,' 70 Va. L. Rev. 549, 610 (1984) (arguing that as costs of various kinds of information decline, the market becomes more efficient). 

 

   /60/ See Cary & Eisenberg, Corporations: Cases and Materials 196 (6th ed. 1988) (estimating that 54 percent of corporate stock was held by institutions in June 1986, but that 20.7 percent of stock was held by mutual funds or banks, as trustees, which pass through their capital gains tax); Columbia University Law Center for Law and Economics Studies Institutional Investors Project, Growth of Institutional Investors in U.S. Capital Markets 9-14 (1988) (estimating institutional ownership of stock from 31 percent in 1970 to 43 percent in 1986). 

 

   /61/ IRC section 1222(11) allows capital losses to offset capital gains, but only $3,000 of capital losses in excess of gains may be deducted. IRC section 1211. Unused capital losses are carried over to future years. IRC section 1212. 

 

   /62/ Costs that are large in amount might be small in impact because they represent merely a short anticipation of the cost. Thus, a seller who sells property this year for consumption next year will rationally put the sales proceeds into the bank or some other liquid investment for a year. On the fact of it, the seller might have incurred a large cost on the sale and has in fact reinvested the sales proceeds. But, the real cost of selling to reinvest was only the time value of money on the transaction costs for a year. The rest of the transaction costs, after a year's interest on the cost, was a cost of selling to consume. 

 

   /63/ See also discussion accompanying infra notes 80-83, considering and rejecting proposals to exempt sales from tax entirely if the proceeds are reinvested. 

 

   /64/ See Gilson & Kraakman, supra note 59, at 569 (emphasizing the role of professional brokers in making the stock market efficient). 

 

   /65/ See, e.g., L. Thurow, supra note 54, at 144-150 (emphasizing the speed of price adjustments for financial instruments contrasted with the sluggishness of prices and persistent return variations on tangible business assets). 

 

   /66/ Delong, Shleifer, Summers & Waldmann, 'Noise Trader Risk in Financial Markets,' 98 J. of Pol. Econ. 703, 704 (1990). 

 

   /67/ Id.

 

   /68/ Dennis Farney, 'More States Bet on Lotteries to Increase Revenue as Popularity of this 'Painless Taxation' Grows,' Wall St. J., Feb. 6, 1986 at 42. 

 

   /69/ Corporate insiders with access to good news about their corporation that has not yet been disseminated to the market can dramatically improve their returns by selling some investment to purchase stock of their corporation before the news is known and taken into account in the price. (For studies finding that insiders can reap superior profits, see, e.g., Seyhun, 'Insider Profits, Costs of Trading and Market Efficiency,' 16 J. Fin. Econ. 189 (1986); Baesel & Stein, 'The Value of Information: Inferences from the Profitability of Insider Trading,' 14 J. Fin. & Quantitative Analysis 553 (1981); Finnerty, 'Insiders and Market Efficiency,' 21 J. of Finance 1141 (1976); Jaffe, 'Special Information and Insider Trading,' 47 J. of Business 410 (1974)). Insiders can also avoid significant losses by bailing out of their stock before bad news hits the street. Securities law, however, commonly forces insiders to disgorge their profits made by trading on inside information. See Clark, Corporate Law, sections 8.6, 8.8, 8.10, at 293-300, 306-309, 316-340 (1986). Insider profits come at the expense of the profits that public shareholders expect from their investments. In theory, it is the public shareholders and not their agents, the managing officers of the corporation, who are the owners entitled to the rewards of stock ownership. Allowing insiders to strip extraordinary gains from public shareholders by early purchases (or slough off extraordinary losses by early sales) would deter public investors from providing needed capital to corporations. Seynum, supra, at 211; Scott, 'Insider Trading: Rule 10b-5, Disclosure and Corporate Privacy,' 9 J. Leg. Stud. 801, 808 (1980); Clark, supra, section 8.2.4, at 273-275. 

 

   H. Manne, Insider Trading and the Stock Market (1966) has argued that abnormal insider profits are appropriately part of the compensation to insiders whose productivity produced the profits and that insider trades signal to the market the true worth of the corporation. The profits come at the expense of stock investors, however, whose capital investment was also responsible for corporate profits. Insiders can also make money (or avoid loss) trading on knowledge of their own incompetence, as well as on their own competence. It is their fiduciary position and not their productivity that gives them their extraordinary profit. See also Gilson & Kraakman, supra note 59, at 629-634, arguing that insider profits too inefficiently transmit information or appropriate price changes to the investing public. 

 

   Since corporate security law is commonly willing to impose confiscation -- a 100-percent tax -- on gains on sales made rational by inside information, it would seem awkward to give those same sales a preferential tax rate. 

 

   /70/ See generally, Repetti, 'The Use of Tax Law to Stabilize the Stock Market: The Efficacy of Holding Period Requirements,' 8 Va. Tax. Rev. 591, 622-626 (1989) (reviewing the literature); Shiller, 'Fashion, Fads, and Bubbles in Financial Markets,' in Knights, Raiders and Targets 56, 59 (J. Coffee, L. Lowenstein & S. Rose- Ackerman, eds. 1988) (arguing that profit opportunity from bubbles of uncertain duration may not be great enough to induce response from rational trader); Arrow, 'Risk Perception in Psychology and Economics,' 20 Econ. Inquiry 1 (1982) (arguing that market participants overgeneralize from current news in evaluating risk). See also Corcoran & Wallich, 'The Analytic Economist: Bursting Bubbles,' Scientific American 125 (May 1991) (arguing that home prices exhibit bubbles).

 

   /71/ See, e.g., Van Horne, supra note 50, 50-54, 163-167 (demonstrating that if risks are to offset each other, the correlation between future value of the two investments must be negative). An investor will not bear risks of an investment if she can offset the risks with other risks that move in the opposite direction. If the market gives high returns to cover risks, the investor can achieve the high return without bearing the risks. Risks that competing buyers can readily avoid by diversifying investments should in theory not be reflected in the market price of the asset, because the buyer with the highest utility for the asset can avoid the risk and sets the price. See, e.g., Hu, 'Risk, Time and Fiduciary Principles in Corporate Investment,' 38 U.C.L.A. L. Rev. 277, 290-295 (1990). If the market filters out diversifiable risks, then an investor trapped in a portfolio that does not offset risks will bear the risks of asset volatility without getting a return rate that pays for them. 

 

   /72/ See, e.g., Auerbach, 'Taxation, Corporate Finance and the Cost of Capital,' 21 J. of Econ. Lit. 905, 908-909 (1983); A. Sen, On Ethics and Economics 29-57 (1987); J. Hirshleifer, Investment, Interest and Capital 195-205 (1970). 

 

   /73/ Hu, supra note 71, at 287 (1990). 

 

   /74/ See Kronovet, 'Characterization of Real Estate Leases: An Analysis and Proposal,' 32 Tax Lawyer 757 (1979). 

 

   /75/ See Wolfman, 'The Supreme Court in the Lyon's Den: A Failure of Judicial Process,' 66 Cornell L. Rev. 1075 (1985) (discussing Frank Lyon Co. v. United States, 435 U.S. 561 (1978)). 

 

   /76/ The fact that leasing is not the dominant form of use of property can sometimes be explained by tax factors. Single family personal residences, for instance, generate considerable net deductions in investments that are not in fact losing money because the owner can exclude the major return from the investment -- the chance to live in the house -- from tax, but deduct the major costs of the house -- interest on the mortgage and sales tax. High bracket individuals could not afford to forego the deductions that shelter outside income and add so much value to the investment, even if leasing were more rational in absence of tax. 

 

   /77/ On its face, section 1031 gives nonrecognition only to 'barters' in which the investor gives up her real estate in exchange with another owner who has real estate she wants, but the rules have been liberalized in practice so that they apply to brokered transactions that are exchanges by the investor only in a formal sense. See, e.g., Biggs v. Commissioner, 632 F.2d 1171 (5th Cir. 1980); Rev. Rul. 90-34, 1990-1 C.B. 154; Proposed Treas. Reg. section 1.1031(a)-3(f)(4) (1990) (authorizing taxpayer's use of an agent who for a fee facilitates deferred exchanges). But see Blum, supra note 4, at 389 (arguing that allowing roll over of cash proceeds, without requiring the form of an exchange, would simplify what are not baroque commercial transactions). 

 

   /78/ Turnover rates on the Vanguard Index Trust, which just matches the Standard & Poor 500 Index of stocks, were eight percent, 10 percent, and 15 percent of assets in 1989, 1988, and 1987, respectively. Prospectus, Vanguard Index Trust at 4 (April 24, 1990). Portfolios, however, do not need to maintain an exact match of an index to remain diversified; with less exacting rules for the portfolio, the turnover needed to maintain diversity should be considerably lower. 

 

   /79/ IRC section 1211 (allowing individuals to deduct $3,000 of capital losses per year against salary or other ordinary income, but otherwise requiring the losses to be used against capital gains or carried forward to future years). 

 

   /80/ D. Smith, Federal Tax Policy 151-155 (1961); Clark, 'The Paradox of Capital Gains: Taxable Income That Ought Not Be Currently Taxed,' in 2 Tax Revision Compendium: Compendium of Papers on Broadening the Tax Base, submitted to Ways & Means Comm. 1243. (Comm. Print 1959). Blum, supra note 4, at 383 (1986) (discussing the technical issues that would need to be resolved to implement a rollover requirement). 

 

   /81/ See, e.g., Johnson, supra note 28. 

 

   /82/ Brown, 'Business Income and Investment Incentives,' in Income, Employment and Public Policy: Essays in Honor of Alvin H. Hansen 300 (1948); Andrews, supra note 19, at 1126-1128. 

 

   /83/ The fairness objection, of course, applies to preferential rates as well as to exemptions. In a world in which start-up investors must make their investments out of fully taxed hard money, a capital gain preference that gave investor with existing unrealized appreciation a chance to reinvest that money upon the payment of low capital gains rates would give them an undue advantage over their upstart competitors. 

 

 

________________________________________________________________________________