Full Text: DRI/McGraw-Hill's Remarks at W&M Hearing on 1998 Budget Proposal

IMPACT OF CAPITAL GAINS TAX REDUCTION

TESTIMONY OF

Before the House Committee on Ways and Means

Mr. Chairman, members of the Committee:

[1] Thank you for inviting me to discuss capital gains taxation and its impact on the economy. DRI/McGraw-Hill has long been skeptical about any claims that tax cuts will raise more revenue than they cost. The capital gains tax is the one tax, however, where this may be true. At least for moderate changes in the capital gains tax rate, our analysis indicates that the impetus the lower rate gives to asset valuation and to the economy roughly offsets the lower static tax revenues. Within a reasonable range of tax rates, the capital gains tax rates appears roughly revenue neutral.

[2] This is one area where dynamic scoring really seems required. Dynamic scoring is, generally, the correct way to look at any form of taxation, but the problem with it historically has been the ability to distort the dynamic scores by letting wishful thinking dominate economic analysis. We feel by consistent use of economic techniques it is possible to do dynamic scoring. It is questionable whether it is possible to do it within a political framework.

[3] The capital gains tax is revenue neutral not primarily because of its impact on the economy, but because of its impact on asset values. In general, we believe that the impact on the economy is relatively small. This would be expected. After all, the capital gains tax cut is only worth about $15 billion on a static basis, and even with substantial leverage it is hard for $15 billion to have an enormous effect on a $7 trillion economy. Our analysis suggests that, even after ten years, the impact on overall GDP is 0.4%, clearly not an enormous number, but hardly a trivial one. In fact, over that ten year period, real GDP would be raised by an aggregate of over $200 billion.

[4] This may be a small share of GDP, but it is certainly a very large share of the $15 billion annual static cost of the capital gains tax. Moreover, some additional tax revenue will come from the higher value of existing assets. When the capital gains tax rate is cut it increases the value of owning shares of stock. As the price of those shares rises, owners pay taxed on that increased amount. Those additional taxes offset the static revenue loss.

[5] Additional revenue also comes from higher turnover. Capital gains taxes lock owners into their stock positions. People are unwilling to sell stock because if they sell the stock, they will have to pay capital gains at a 28% rate. Sub-optimal stock positions are thus locked.

[6] The estimates of higher turnover are probably the most questionable part of any analysis of the impact of capital gains. In the past, capital gains cuts have usually come as part of larger tax packages. It is very difficult to disentangle turnover from the overall impact of the rest of the tax code. Moreover, capital gains realizations depend heavily on what has been happening in the stock market which is also moved sharply by changes in the capital tax. The unlocking effect seems clear from anecdotal reports and simple logic but its magnitude is somewhat questionable. We feel our estimates are conservative -- additional turnover of only 5% of unrealized capital gains over the next ten years.

[7] There will be some negative effect as taxpayers shift income away from ordinary income and into capital gains. If the capital gains rate is reduced, people have a greater incentive to take income in the form of tax-advantaged capital gains rather than ordinary income. The restriction of passive loss deductions, however, makes this a more difficult game to play than in the 1980s. We estimate the impact at about $2 billion a year. Prior to the 1986 tax law we would have estimated a substantially higher figure, $5 to $10 billion annually.

[8] Overall, we believe that a moderate capital gains reduction, on the order of that proposed in the Hatch-Lieberman bill, will result in a small gain in revenue. Too large a cut in rates, however, is no longer revenue neutral. Much of the clawback from revenue comes from applying the capital gains rate to the higher asset values. If the rate is reduced too far, the tax on the rise in assets is not sufficient to offset the static revenue loss. Our analysis of HR 14, for example, indicates that the bill would be a significant revenue loser, costing well over $100 billion on net over a ten-year period. The Dreier bill would have a greater impact on the economy and on the stock market, but because the stock market gains are taxed at a lower rate, it would not raise enough revenue to offset its cost.

[9] This is not an argument against indexation. Indeed, our study suggests that indexation is more effective, dollar-for-dollar, than rate cuts. Indexation cuts rates only on future gains, and does not reward the past. We do caution against trying to cut the rate and introduce indexation, however.

[10] One argument against a capital gains rate cut is that it accrues only to the rich. Although it is certainly true that the rich have more capital than the poor, the poor don't lose in a capital gains cut. Pension portfolios will gain in value, and the middle class have a rising portfolio of 401K plans. The great bulk of the assets of the middle class, and particularly the younger middle class, is tied up in their homes. More direct relief on capital gains in housing could spread the benefits more equally. But even if households have no capital assets themselves, they will benefit from the higher investment and resulting higher productivity created by the capital gains rate cut.

[11] The capital gains cut helps most people and hurts no one. The overall impact on the economy, though small, is clearly positive. Reducing capital gains will not remake the economy, but it helps modestly at no cost to the Treasury. We find it difficult to understand why the bill should not be passed.

                               TABLE 3

            CUMULATIVE IMPACT OF TAX REDUCTIONS IN S. 66

Real GDP (billions of 1992 $)                        $185

THE CAPITAL GAINS TAX, ITS INVESTMENT STIMULUS, AND

[12] The impact of a capital gains tax cut on the economy and on tax collections has been investigated for decades. The evidence suggests to almost all economists that a capital gains cut is good for the economy and roughly neutral for tax collections. Although static analysis of the tax suggests the federal government loses revenue, a dynamic analysis suggests that the government can gain revenue. Such dynamic gains depend on the time interval examined and on the feedback effects from a stronger economy. By encouraging investment, a capital gains tax reduction becomes a true supply-side tax cut, and perhaps the only cut that really might fully pay for itself.

[13] Cutting the capital gains tax will boost investment by lowering the cost of capital to businesses. The only preferences now offered for individuals are a maximum rate of 28% for gains versus 39.6% for ordinary income, and the deferral of taxation on accrued gains until the underlying asset is sold. A new preference that included just 50% of the gain in ordinary income would thus cut the effective top marginal rate paid by capital gains recipients to 19.8% from the current 28%. Together with a reduction in the corporate rate on gains from 35% to 25%, this shift would, in a static analysis, lose about $15 billion of federal revenue per year. (From the perspective of an investor facing both state and federal capital gains taxes, the current marginal tax rate on capital gains is 32%, assuming a 6% state and local rate deductible against the federal income tax. The new effective rate would be 23.5%.)

[14] The lower tax rate on capital gains unambiguously raises the value of assets subject to the tax because the same, related stream of pre-tax earnings is now worth more after-tax to the investor. This affects the stock market most. DRI analyzes share prices as the after-tax, discounted present value of dividends and capital gains (driven by retained earnings).

[15] If all holdings were subject to the tax, the proposed lower tax rate on gains would need to raise share prices 8% to equalize the risk-adjusted rate of return on shares with that on bonds. This calculation assumes that bond yields would not be affected by the lower capital gains rate. It assumes that all of the correction required to avoid an unjustifiable yield differential is produced by a rise in the price-earnings ratio of stocks. This assumption of unchanged bond yields is reasonable when considering a solitary change in capital gains taxation. That is because bond yields represent the discounted value of future short-term interest rates, adjusted for risk, and there is no immediate reason why short- term rates should be affected by the solitary tax change.

[16] However, we recognize not all asset holdings are subject to capital gains taxes. Tax-sheltered investments (such as pension funds and 401K and IRA holdings) and foreign investments now account for about half of all equity holdings. Assets passed through an estate without a sale before death escape capital gains taxes, although heavy estate taxes apply to any accrued appreciation plus the original principal invested.

[17] As share prices rise in response to a tax rate cut in the above-noted scenario, the expected pre-tax rate of return of equity would drop, inducing non-taxable investors to shift into bonds. If we assume that the degree of risk aversion, and thus the trade-off between risk and return, is the same for these investments as for non-tax-exempt holdings, the impact on share prices will be reduced by their ratio to total holdings. This implies that in an environment with half the investors exempt from taxation, equity price-earnings ratios will rise by only 4%, half the 8% gain calculated above. Note that this would be a permanent increase in the price-earnings ratio corresponding to any given bond yield. Experience suggests that the move will not occur until the bill is signed by the President, but then it will come quickly.

[18] The capital-gains tax cut would lower the net cost of capital by about 3%. This 3% estimated reduction is a blend of an 4% reduction in the cost of equity and an unchanged cost of debt finance. The core cost of equity equals the dividend yield (reduced 4%, as indicated above) plus expected dividend growth (assumed to be unchanged at about 6%). Other factors equal, this shift would raise the level of business spending by about 1.5%, or $18 billion in 2007. Over a 10-year period, the capital stock would rise 1.2% above its baseline level, increasing productivity by roughly one-third this percentage, or 0.4%.

BUDGET IMPACT

[19] The effect on capital gains tax collections is complicated and potentially controversial. However, a decomposition of the impact into discrete components can replace some of the emotion in the debate with rationally discussible magnitudes:

1. In a STATIC ANALYSIS, the lower rates will reduce tax revenues

2. However, lower rates unlock assets, creating HIGHER TURNOVER and

3. The higher value of assets, and thus ASSET PRICES, raises total

4. INCOME RECLASSIFICATION from ordinary income to capital gains will

5. The stronger resulting economy and HIGHER GDP raises total tax

[20] The static loss is easy to calculate since it is simply the change in the tax rate times the level of capital gains receipts. There are only two complications: first, there may be an impact on state tax revenues that are tied to federal income definitions and, second, capital gains taxes are not broken out separately in the statistics. Using estimates based on 1991 returns, the lower rate would lose about $11 billion per year. Based on gains since 1991, we estimate the current impact at $15 billion.

[21] The unlocking/higher-turnover effect will increase revenues in the early years of the program. When the capital gains tax rate is reduced, assets become more liquid in the sense that the tax loss involved in selling them is a lower percentage of the asset's value. This switch makes individuals ready to sell more often, and by increasing the turnover raises revenue in the early years of the program. Estimates based on past CBO papers suggest that this unlocking could add $10 billion to revenue in the first year. The additional revenue would diminish rapidly, however, since the change only moves forward the realization of capital gains. Under some models, the impact turns negative in the third and fourth year. In the long run, there is still some positive effect in our analysis because a higher turnover implies that fewer capital gains expire on the death of the owner.

[22] The higher asset value will be the primary positive contributor to tax revenues. A 4.0% share price increase will raise total stock market valuation by $280 billion. If 15% is sold within the first year (about the current turnover rate in the stock market) and taxed at 19.8%, this effect would raise capital gains revenue by $8 billion. Taxable gains on other assets (primarily privately held businesses) might rise by about one-half this amount.

[23] The increased gap between the ordinary income and capital gains rates will induce individuals to reclassify income in order to lower their tax liability. This income reclassification was, in fact, one of the primary reasons cited for going to an equal treatment in the 1986 tax bill. There is evidence to suggest that the change was successful in widening the income-tax base. Reversing this 1986 reform would clearly increase the incentive to convert ordinary income into capital gains, for example, by deferring income in closely held enterprises or shifting from wage income to stock options. The size of the impact is uncertain, but it could easily cost $5 billion per year in reduced tax revenue. At least, this was the magnitude cited in 1986 when the change was made in the opposite direction. We have assumed only half this effect, since we believe the changes in passive loss rules make income-shifting less easy.

[24] The stronger growth of the economy produced by a solitary change in capital gains taxation will add to both capital gains and ordinary income tax collections. Our model suggests that after 10 years real GDP could be 0.4% higher than in the baseline, because of the capital gains tax rate change and its repercussions on investment. In 1992 dollars, this extra growth would add $34 billion to national income. In 2007 dollars, the impact would be $116 billion, adding $30 billion to federal revenue.

[25] The usual static analysis does not consider either the increase in asset values or the impact of a stronger real economy. Only the rate reduction and the higher turnover and income reclassification are usually included in the analyses done by the Congressional Budget Office or other government agencies. The difference between the estimates illustrates the importance of using dynamic estimates of the impact of tax changes. Much of the literature from the extreme right, which indicates that the capital gains tax is a huge money-maker, illustrates the dangers as well.

[26] This analysis so far has examined only the macroeconomic implications. The distribution of the benefits is also an important political and economic issue. A narrow view of the capital gains tax cut is that it will favor the wealthy, who have more capital than the poor. Although the rich get most of the gains, no one loses. Often overlooked benefits flow to all workers and middle income citizens, and the overall economy wins. The middle class will benefit from greater appreciation in their pensions, now increasingly structured as defined contribution plans in which all investment returns are captured by the employee. Small businessmen will gain from more generous tax treatment of the gains on their enterprise. And all employees will see wage gains tied to investment-driven higher productivity.

                               TABLE 1

                         1998   1999    2000    2001    2002    2003

                          [Table continued]

                         2004   2005    2006    2007   Total

INDEXING CAPITAL GAINS

[27] The proposal to index capital gains has a similar impact to the exclusion from tax. It also cuts the cost of capital to the firm, and raises the value of assets, including especially common stocks. The impact will depend on the amount of inflation expected by the market, and on the details of the law.

[28] Over the past, inflation has accounted for about half of capital gains. That average, however, includes the high-inflation 1974-1985 period. We believe the market is currently discounting a 3% inflation rate, which would only account for about one-third of anticipated capital gains.

[29] Our analysis assumes the indexation comes on top of a reduction in the capital gains rate to 14%, as in HR 14. Thus, the impact, or the cost of capital, is less than half as great as if it were done on its own. It also, of course, loses only half the revenue it would otherwise. Unfortunately, because the higher stock values are clawed back at a much lower tax rate, the dynamic scoring of the bill does not offset its static loss. A net loss of about one-third the static loss results.

[30] This is not an indictment of indexation. In fact, our model indicates that indexation gives a greater impact on investment per dollar of tax revenue lost. The analysis does, however, suggest that although capital gains rates are revenue-neutral for moderate changes, if they are reduced too far they will lose revenue.

[31] The additional static tax loss is very small in the near term, since the bill does not index for past (pre-1997) inflation. The impact builds over time, however, as more of the gains can be indexed. By 2007, the static loss from indexing is estimated at $8 billion.

                               TABLE 2

                         1998   1999    2000    2001    2002    2003

Total                      $4    ($4)    ($9)   ($15)   ($16)   ($15)

                          [Table continued]

                         2004   2005    2006    2007   Total

Total                    ($14)  ($15)   ($14)   ($14)  ($112)

[32] Asset prices will rise proportional to the effective reduction in the capital gains rate in future years. This means that although indexation has little immediate impact on the effective capital gains tax rate, it does have full impact on asset prices and capital costs. The increase in turnover is difficult to estimate in this bill. In general, indexation would have only a small near-term impact on turnover, since pre-1997 inflation is not forgiven.

[33] Income reclassification is less likely with indexation than with exclusion. There is no advantage to reclassifying income as capital gains unless the underlying asset is held long enough and its base price is high enough to make the inflation adjustment attractive.

[34] The impact on the economy is proportional to the impact on asset values. The economic impact results from the lower cost of capital to business, which increases investment. The marginal impact of adding indexation to the 50% exclusion is about one-third the impact of the exclusion. (Note that these magnitudes would have changed if done in the opposite order.)

SUMMARY

[35] Our analysis shows that the Hatch Lieberman bill is essentially revenue-neutral. The static losses are largely offset by the increase in the value of assets and the stronger domestic economy over the ten-year period. Our analysis shows a marginally positive result, but well within the margin of error.

[36] In essence, the bill does nobody any harm and does business and does a capital owner some good. The proposal lowers capital costs and thus raises business investment improving productivity and income growth in the long run. The higher incomes raise tax revenues offsetting about one-third of the static revenue loss. Most of the rest of the revenue loss is offset by the higher turnover of capital assets and by the increase in the value of the assets caused by the lower tax rate.

[37] Combining a sharper reduction in the rate with indexation, as in HR 14, is not revenue neutral. There is a proportionate impact on the economy, and the stronger domestic economy offsets nearly one- third of the rise of the static revenue loss, as in our analysis of Hatch-Lieberman. Asset values also rise, but these are clawed back at a lower tax rate and, therefore, do not offset as much of the revenue. Similarly, the higher turnover is taxed at only 14% instead of 19.8%, yielding a smaller impact on revenues. We, thus, find that the bill loses $112 billion over a ten-year period, with only about two-thirds of the static loss made up by income gains elsewhere.