President Bush’s tax proposal has three elements that would promote economic growth. First, he would eliminate the double taxation of corporate income by not taxing dividends and by reducing taxable capital gains to reflect already-taxed corporate earnings that were retained for reinvestment. Second, he would move forward the effective date of the marginal tax rate reductions enacted in 2001 that have yet to be phased in. Third, he would increase the amount of investment in equipment that businesses can count as an immediate cost (expensing) instead of having to depreciate it over time. The President has also recommended extending IRA treatment to cover more saving. Much has been written about the elimination of the tax on dividends. We should also recall why marginal tax rate relief is important for growth. What are marginal tax rates? Marginal tax rates are the tax rates that apply to the next dollar of income one might earn. The statutory marginal tax rates prior to the 2001 Act were 15%, 28%, 31%, 36%, and 39.6%. The first Bush tax cut carved out a 10% rate from a portion of the 15% bracket at the bottom, and scheduled phased reductions in the four top rates. These have become 27%, 30%, 35%, and 38.6% for 2002 and 2003, with further cuts scheduled for 2004 and 2006. The full cuts will reduce the rate structure to 10%, 15%, 25%, 28%, 33%, and 35%. These statutory tax rates do not tell the whole story. Where income is double taxed, as with corporations and shareholders, the combined marginal tax rate must be considered. Ending the double taxation of corporate income would reduce marginal tax rates on saving and investing. The same is true for wages that are subject to the income and payroll taxes. State income taxes must count as well. Also, various provisions of the federal income tax alter taxable income or tax liability by phasing out certain exemptions or credits as one’s income exceeds certain thresholds. Examples of exemptions or deductions that are eliminated for upper-bracket taxpayers as their incomes rise include personal exemptions, itemized deductions, and eligibility for deductible IRAs, among others. Credits that are lost as income rises include the earned income tax credit (EITC) for low-income wage earners, and the child credit, Hope scholarship credit, lifetime learning credit, and many others for upper-bracket taxpayers. These phase-outs cause taxable income or tax liability to rise faster than the jump in actual earnings would otherwise bring about, and they act as a hidden increase in the tax rate on the added income. Why do marginal tax rates matter? People make decisions based on after-tax rewards, and they make them on incremental changes—whether to save or consume the last available dollar, whether to work a few more hours a week or take the time off, whether to add an additional machine to the shop or not. These are incremental decisions, not all or nothing choices. It is the additional tax—the marginal tax rate—on incremental income that determines how much one would get to keep, after taxes, if one were to work a bit longer or save a bit more. It is the rise or fall in these after-tax rewards for additional effort that affect behavior. Taxes on non-marginal units of income or output do not affect work, saving or production decisions. Keynesian economists thought that tax changes worked by giving people money to spend, or by taking it away, affecting their “disposable” and “demand” for goods and services. They concerned themselves with tax revenues and average tax rates. But over thirty years ago, Prof. Milton Friedman pointed out that when the government cuts taxes without cutting spending, it just borrows the tax cut back. The public has no more money to spend, and the demand-side or “pump-priming” effect of tax cuts is nil. The same is true of additional government spending, because the government must either raise taxes or borrow more from the public to pay for it. Tax and spending changes can “pump up” demand only if the Federal Reserve steps in with new money to buy the added federal debt, which is really a change in monetary policy, not a change due to the tax cut or spending per se. Tax changes really affect the economy by altering incentives to work, save or invest, that is, to supply labor, capital, and output. That is why neoclassical or “supply-side” economists believe that tax changes affect the economy by altering marginal tax rates, not tax revenues. Lowering marginal tax rates boosts the supply of labor and capital, and makes it possible to increase employment and production. As people are paid for their added production, they can then use their increased income to buy their increased production. Supply rises first, and then demand follows. An excise tax for illustration It is the marginal tax rates that affect the point at which the supply and demand for a product, for labor, or for capital come together, and the quantities employed or produced. (Chart 1.) When an excise tax is imposed (or increased) on the sale of a product, the cost of the product to the consumer rises (so they want to buy less—see the demand curve), and the price received by the producer falls (so they want to produce less—see the supply curve). The spread between the consumer’s price and the producer’s receipt equals the tax per unit. If a tax is removed (or reduced) on a product, the price paid by the consumer falls and the price received by the producer increases. Output and consumption of the product will increase. In Chart 1, the initial, low tax rate per unit (t1-t1) causes a drop in production. The area of the corresponding shaded rectangle is the tax revenue, which is the tax per unit times the new reduced quantity of output. Note that the tax revenue is less than would have been the case if the tax had not reduced the quantity produced. The triangular area to the right of the revenue rectangle is called the “tax wedge”. It represents the loss of value of economic output — the “dead weight social loss” — of the tax. This value of the lost production exceeds the value of the resources that are released for other uses by the decline in output. When the tax is imposed at a low rate, it brings in revenue while reducing output only slightly. As the tax rate is increased (t2-t2), output falls further, but the revenue (rectangular area) will rise if the percent increase in the tax rate exceeds the percent drop in output. At higher tax rates (t3-t3), the percent drop in the tax base may exceed the percent rise in the tax rate, and revenue will fall. The Laffer Curve (Chart 2), named after Prof. Arthur Laffer, demonstrates this relationship between the tax rate, tax base, and tax revenue (Chart 2). The top of the curve is the revenue maximizing point. The economically optimal tax rate is less than the revenue-maximizing rate, because the total cost of the tax to the society includes the dead-weight loss of real output caused by the tax as well as the tax revenue sent to the government. Government should only spend until the last thing it buys is worth more than its dollar cost, in fact, is worth the sum of the dollar cost and the marginal dead-weight loss of output caused by the tax collection. Marginal income and payroll tax rates (Charts 3, 4.) Taxes on labor and capital income act just like an excise tax, only with more widespread consequences. The tax in question could be the personal income tax or the payroll tax for Social Security and Medicare. These taxes raise the cost of labor and capital to businesses, and cut rewards to workers and savers. Employment, capital formation, and output all fall. It is these “excise effects” of the income and payroll taxes that affect economic behavior, not the disposable income effect that the Keynesians used to talk about. The quantity of capital is more sensitive to taxes than is the quantity of labor. The value of adding capital to the economy is very high, and its usefulness declines very slowly as more is added (demand falls slowly). Capital is easily reproduced (elastic supply). People are very willing to finance capital formation if the returns look good, but they are equally willing to consume instead of save or to invest abroad instead of in the United States if the rate of return on saving and investment is driven down by rising taxes. If a tax is imposed on capital income, the capital stock will fall until its scarcity drives up its productivity and its earnings to pay for the tax and still leave savers the normal after-tax return. It takes a large reduction in the quantity of capital to raise its rate of return sufficiently to compensate investors for the added burden of the tax. Workers have fewer options than savers, and their supply is less “elastic” (note the steep supply curve). Most people must work to have a satisfactory income, and to some degree they must conform their hours of work to the requirements of their employers. They have some choices—such as to take or reject overtime, to contribute a second earner from the family to the labor force, how long a vacation to take, and when to retire. Employers, to some extent, can substitute capital for labor. Hence, a tax on labor is largely born by the workers, less by the employers — that is, the after-tax wage is driven down a good deal by the tax. Workers suffer most when capital is taxed In these labor and capital charts, the “demand” curves for labor or capital are labeled “marginal product”. Marginal product is what an additional unit of labor or capital adds to the value of output, so marginal product is what workers and savers are paid for their services. Marginal product falls as more of an input is added, if other inputs do not rise in proportion. For example, for any given stock of capital, adding more labor causes output to rise, but at a diminishing rate (law of diminishing returns) because the expanded work force has less capital per worker. Hence the marginal product of labor line slopes down as we move along it. The whole marginal product of labor line would shift up, however, if the capital stock were to increase, because any given amount of labor would have more capital to work with and would be more productive. An increase in the capital stock would increase wages and employment. If we increase the tax rate on capital, the stock of capital would fall, a lot, because it is sensitive to tax. Investors would decide to consume instead of invest, an acceptable second best use of income. As the capital stock declined, the remaining assets would earn more, and the investors would end up with much the same rate of return as before the tax hike, albeit on a smaller amount of capital. Labor, however, would be hurt. It would become less productive. The entire marginal product line for labor, and the demand for labor would fall, driving down the wage. Because the amount of capital affects the productivity of labor, wages, and employment, labor bears most of the economic burden of any taxes that are levied on capital income. Labor’s after-tax income is highest under a tax system that does not tax the earnings of capital. What to do with the tax code Congress should not only lower statutory marginal tax rates, it should eliminate phase-outs that discriminate against upper income taxpayers and Social Security recipients, and correct the tax biases against saving and investment that are inherent in the income tax. Accomplishing all that will require fundamental tax reform. The Bush plan is a step in the right direction. More needs to be done.