SERIES: Brookings Policy Brief Series | # 31 of 186 « Previous | Next »

Don't Buy the Sales Tax

The tax reform panacea in vogue several years ago was the flat tax. This go around, it's the retail sales tax. Advocates praise the simplicity, the potential for economic growth, the enforceability, and the fairness of a tax system based on consumption. But when examined closely, the simplicity breaks down, payments would be close to impossible to collect, and the tax's fairness would be, at best, questionable. Nonetheless, more plausible consumption and income tax reform plans should not be ignored.

POLICY BRIEF #31

The politics of tax reform are cyclical, and once again, we're hearing the call for sweeping change. The flat tax and the idea of fundamental tax reform dominated policy discussions in 1995 and the early part of 1996. These proposals aimed to replace the income tax, drastically simplify taxes, and spur economic growth by flattening tax rates, eliminating tax preferences, and taxing consumption rather than income. In 1997, however, Congress and the President moved in essentially the opposite direction. The Taxpayer Relief Act of 1997 showered new deductions and credits on various groups of taxpayers, made taxes more complicated, and raised effective marginal tax rates for many people.

Now, as 1998 begins, there are once again renewed calls to tear out the income tax and start over. Although the flat tax has certainly not disappeared, the newest plan to attract significant attention is a national retail sales tax. Representatives Dan Schaefer (R-Colo.) and Billy Tauzin (R-La.) have proposed a 15 percent sales tax which they claim will replace the personal and corporate income taxes and the estate tax. A group called Americans for Fair Taxation has launched a multimillion dollar campaign to replace income, estate, and payroll taxes with what they claim would be a 23 percent sales tax. These plans would tax almost all private consumption and all government expenditures. Ways and Means Chairman Bill Archer (R-Texas) has indicated general support for a sales tax but to date has officially kept his options open.

Sales tax advocates claim their plans would dramatically boost prospects for economic growth, simplify the tax system to the point where states could collect the tax and the IRS could be abolished, reduce costs of compliance by an order of magnitude, and provide Americans with a tax system that is fair, visible and understandable. All worthy goals, but that's a very tall order. Can it be filled?

A National Retail Sales Tax

Retail sales occur when businesses sell goods or services to households. Business-to-business transactions are not retail sales because the purchase is used as an input, not as household consumption. Household-to-household sales of goods or services are not included, either, since tax should have already been paid when the good was originally sold to a household from a business. For example, the sale of a newly constructed home to a family that will occupy it is a retail sale. But the sale of that same newly constructed home to a business that is planning on renting it to others is not a retail sale. Nor is a sale of an already existing home from one occupant to another.

Measuring the sales tax rate would seem to be a simple problem. Suppose a good costs $100 and there is a $30 sales tax placed on the item. Most people would probably consider that to be a 30 percent sales tax, since the tax is 30 percent of the selling price. This is known as the tax-exclusive tax rate. An alternative would divide the $30 tax payment by the total cost to the consumer ($100+$30), making it a 23 percent rate. This is known as the tax-inclusive tax rate. Sales taxes are typically quoted as tax-exclusive rates. Income tax rates, however, are typically quoted as tax-inclusive rates. For example, a household that earns $130 and pays $30 in taxes would normally think of itself as facing about a 23 percent (30/130) income tax rate.

Although there is no one true method of reporting the rate, it is crucial to understand which approach is being used, since the tax-inclusive rate will always be lower than the tax-exclusive rate and the difference grows as the rates rise. At a rate of 1 percent the difference is negligible, but a 50 percent tax-exclusive rate corresponds to a 33 percent tax-inclusive rate.

Both the Schaefer-Tauzin (S-T) and the Americans for Fair Taxation (AFT) plans would attempt to tax all goods or services purchased or used in the United States. Exemptions would be provided for business purchases for resales, purchases to produce taxable property or services, and exports. Expenditure on education would be exempted on the grounds that it is an investment. The sale of newly constructed single-family homes would be taxed, as would improvements to existing single-family houses. Already existing owner-occupied housing would be exempted. All financial intermediation services would be taxed. The tax would exempt expenditures abroad by U.S. residents and half of foreign travel, but tax domestic expenditures by non-residents. Notably, the proposals would tax all federal, state, and local government consumption items, as well as government investment in equipment and structures.

The plans would offer demogrants, or rebates, for each family, equal to the tax rate times the poverty line, the amount of income a family needs officially to stay out of poverty. Both plans would allow businesses who remit taxes to government to retain a small percentage of revenues collected as compensation for collection costs.

A national retail sales tax structured along these lines would represent a sharp break from the current tax system. First, the tax base would shift. Currently, our system taxes something loosely related to income. Under a pure sales tax, the base would be consumption. Second, marginal tax rates currently rise with income, but would be flat in a sales tax. Third, for reasons of social policy, tax administration and politics, the income tax contains a series of deductions, exemptions, and tax credits. Under the proposed sales tax, all of these would be eliminated. Finally, the sales tax would make vast changes in the tax system's administration, enforcement, and point of collection.

Lessons from the States

Sales taxes already exist in 45 states, the District of Columbia, and over 6,000 localities. State tax rates range from 3 percent to 7 percent. While state sales taxes are widely viewed as successful, they are very poor models for federal reform.

States only tax about half of private consumption of goods and services. Many states exempt goods such as food, electricity, telephone service, prescription medicine and so on. Most states do not tax services very well, if at all. In addition, between 20 and 40 percent of state sales tax revenue stems from business purchases, which are not retail sales. This causes cascading of the sales tax, which distorts relative prices in capricious ways and gives firms incentives to merge with other firms in order to avoid the tax. States often do not require their own government to pay sales taxes. And states do not provide demogrants; instead they help the poor by exempting specific items like food. These findings suggest that running a pure, broad-based sales tax as envisioned by S-T and AFT could be quite difficult in practice.

What?s the Rate?

Determining the appropriate tax rate in a national sales tax can be tricky. For example, if (a) there is zero tax evasion, (b) state and local taxes do not change, (c) the tax base is not reduced by political or other factors, (d) nominal government spending is held constant, and (e) transition issues and economic growth are ignored, the AFT proposal would require a 23 percent tax on a tax-inclusive basis, or a 30 percent tax rate in the more familiar tax-exclusive approach. (All of the following rates are tax-exclusive.) But realistic adjustments for each of the stringent conditions listed above raises the estimated tax rate.

The tax evasion rate under the income tax is between 15 and 20 percent. Evasion issues are discussed in the next section; here we simply note that a 15 percent evasion rate in the sales tax, which is probably conservative, would raise the rate to 35 percent.

In the absence of federal income taxes, states would likely have to convert their own income taxes to sales taxes and conform closely to the federal tax base. Add in the revenues from existing state sales taxes, and the combined federal and state sales tax rate would climb to 45 percent.

Suppose the tax base were reduced by one-third from the pure consumption tax proposed. In light of all the preferences in the current income tax, and the exemptions in state sales taxes, this is probably not an unreasonable assumption. It would come about, for example, if just food and health were exempted, or if just government expenditures were exempted. Cutting the base like this would raise the required rate to 67 percent.

Since a sales tax would likely raise prices, nominal government expenditures on goods and services consumed by government and on transfer payments would have to rise to hold constant the inflation-adjusted effects of government policy. This could add another 10 percentage points to the required tax rate.

Any transition relief provided to households would reduce the tax base and raise the required rate further. For example, the S-T plan would make payments to those on fixed annuity incomes to account for any increase in prices. Economic growth could reduce the required tax rate, but as discussed below, probably not by very much. And if saving rises following the implementation of a sales tax, consumption would have to fall, which would reduce the tax base and raise the required tax rate further.

The clear result, then, is that any realistic sales tax plan would have tax rates much higher than the 23 percent rate promised by the AFT. Sales taxes at such high rates raise crucial questions about enforceability.

Enforceability and Simplicity

Tax simplicity and tax enforcement should be analyzed together—any tax can be simple if it is not enforced. If it could be enforced, the sales taxes would be quite simple for the typical household, but problems would arise for businesses.

The likely rate of evasion in a sales tax is a key issue. Sales tax advocates admit that evasion would be a certainty, yet make no account for it in their estimates and hope that sentiments of fairness will induce taxpayers not to cheat. They also point to low marginal tax rates as an inducement not to cheat, but as shown above, the tax rate would not likely be low.

Another claim is that detection of cheating would rise dramatically since only retailers would have to be audited, but this is misleading. Under the sales tax, businesses that make retail sales would be responsible for sending tax payments to the government, unless the buyer used a business exemption certificate, in which case no tax would be due. But the buyer would have the legal responsibility for determining whether the good is used as a business input or a consumption item. This means that auditing and enforcement would have to focus not just on retailers, but also on all businesses that purchase from retailers, to ensure that business exemption certificates were used appropriately. One study found that, in Florida, where sales taxes have never exceeded 6 percent, 5 percent of all purchases made with business exemption certificates were used inappropriately to exempt personal consumption from taxes. At the much higher tax rate needed in a federal sales tax, a much larger percentage of sales to businesses might be expected to fall into this category.

Most importantly, the sales tax would generate tremendous opportunities for evasion. For example, in the income tax, the rate of evasion is around 15 percent. But income where taxes are withheld and reported to government by a third party has evasion rates of around 5 percent. For income where taxes are not withheld and there is no cross-reporting, evasion is around 50 percent. Since the sales tax would feature no withholding and no cross-reporting, the possibility of high evasion rates needs to be taken quite seriously.

Advocates also assert that the sales tax would be more effective than the current system at raising revenue from the underground economy. The classic example is that of a drug dealer who currently does not pay income tax on the money he earns, but would be forced to pay taxes under a sales tax if he took the drug money and bought, for example, a Mercedes. The problem with this argument is laid out best by Rep. Richard Armey (R-Texas): "If there is an income tax in place, he [the drug dealer] won't report his income. If there is a sales tax in place, he won't collect taxes from his customers" and send the taxes to government. In the end, neither system taxes the drug trade.

Many other countries have attempted to implement a retail sales tax, or variants, and almost all have abandoned the tax and moved to a value-added tax. Governments have gone on record as noting that at rates of more than 12 percent, sales taxes are too easy to evade. The most optimistic assessment would be that there is no historical precedent for a country to enact a high-rate, enforceable, national sales tax. That does not mean it is impossible, but extreme caution would be appropriate.

Finally, some sales tax advocates would eliminate the IRS and have the states administer the tax. Even though the states would keep 1 percent of the revenue they collect, they would have poor incentive to collect federal taxes adequately. Even the Wall Street Journal, no fan of big government, notes that "it is fantasy to think of 'getting rid of the IRS.'"

The Tax Base

For reasons of equity, efficiency, and simplicity, establishing a broad tax base is a key element of sales tax proposals. The broader the base, the lower the rate can be. As the income tax shows, allowing specific exemptions creates a political slippery slope. Loopholes accorded to one group breed additional loopholes, by fueling demand from other groups for equally favored treatment and by weakening congressional resolve to stem the tide of special requests.

But taxing a broad base will be hard. Some items are quite difficult to tax, like imputed financial services. Other items may not be taxed for reasons of social policy. Child credits, rental payments, and some or all food may fall into this category. Finally, some sectors might not be taxed because of strong political influences—housing and health, for instance, though exemptions here may also be related to social policy.

These issues create serious trade-offs. For example, taxing health insurance would raise the number of uninsured by an estimated 6 million to 14 million people. Not taxing it would raise the sales tax rate by several points.

Even with extreme political discipline in avoiding subsidies, it will be difficult to tax more than 80 percent of personal consumption. Retaining some of the major preferences in the income tax could reduce the private consumption base to about 60 percent of personal consumption.

The most controversial and interesting feature of the proposed sales tax base is the provision to tax all consumption and investment purchases made by the federal, state, and local governments. These provisions increase the tax base by more than 40 percent. But taxing government only helps keep the sales tax rate down when government is not allowed to raise the additional amount of revenues it collects to pay the tax. If this restriction takes effect, real government spending would have to fall dramatically.

Economic Growth

Tax reform could increase economic growth in several ways. A broader tax base that eliminated loopholes would help direct assets and efforts to their most productive economic use. Lower tax rates would encourage working, saving, and investing and reduce incentives to avoid or evade taxes. Simpler taxes would reduce compliance costs. Changing the base from income to consumption would enhance the after-tax return to saving and investment.

But growth projections also need to take into account transition relief. Consider a 65-year-old who retires the day that a sales tax takes effect. While she worked, her income exceeded her consumption and she paid income taxes. Now in retirement—when her consumption will exceed income—she will be asked to pay consumption taxes. Should she be given tax relief for consumption financed by drawing down her assets? Giving such people tax relief on their accumulated assets would raise the sales tax rate considerably and reduce growth, but not giving tax relief seems unfair. Politically, some sort of relief seems almost a certainty.

Estimates suggest a well-functioning, broad-based consumption tax, with limited personal exemptions and limited transition relief could raise income per person by up to 2 percent over 10 years. But more generous transition relief or erosion of the tax base would drive the growth effects to zero fairly quickly. These results stand in sharp contrast to AFT's claims that after 10 years of a sales tax, the typical American household would be at least 10 percent and "probably 15 percent better off" than they would have been otherwise.

Would a Sales Tax Be Fair?

The issue of fairness in taxation is often contentious and always subjective. The sales tax would tax consumption at a flat rate compared to the current system, which taxes income at progressive rates. Taxing consumption instead of income is often justified on grounds that consumption may be a better indicator of long-run ability to pay taxes, since income varies significantly from year to year. But for people who face constraints on what they can borrow, the long run may not be the most relevant time period. Clearly, there is nothing inherently fair (or unfair) about having just a single rate.

If households are classified by annual income, the sales tax is sharply regressive. Under the AFT proposal, taxes would rise for households in the bottom 90 percent of the income distribution, while households in the top 1 percent would receive an average tax cut of over $75,000. If households are classified by consumption level, a somewhat different pattern emerges. Households in the bottom two-thirds of the distribution would pay less than currently, households in the top third would pay more. Still, households at the very top would pay much less, again receiving a tax cut of about $75,000. There appears to be little sound motivation for heaping huge tax cuts on precisely the groups whose income and wealth have benefitted the most from recent events, and raising burdens significantly on others.

The proposed sales tax also contains potentially severe marriage penalties. Recall that the demogrant would equal the tax rate times the poverty line. The poverty line is $7,890 for one person, plus $2,720 for each additional person. At a sales tax rate of 30 percent, a conservative estimate, two single people would be entitled to rebates of $2,367 each for a total of $4,734. If they married, they would receive $3,183, so their annual marriage penalty would be $1,551. At higher rates, the penalty would rise.

Advocates like to assert that sales taxes are pro-family relative to the income tax. However, children and families benefit disproportionately from numerous features of the current system, including dependent exemptions, child credits, child care credits, earned income credits and education credits. And the preferential treatment of housing, health insurance, and state and local tax payments also plausibly helps families, since they consume relatively more housing, medical services, and government-provided services such as education. All of these preferences would be eliminated under a sales tax.

Moreover, compared to childless couples, families with kids generally have high consumption relative to income, so switching from income tax to a consumption tax would further raise tax burdens during years when family needs were highest. Based on 1996 data, a recent study found that enactment of a broad-based, flat-rate consumption tax like the sales tax or flat tax would hurt families with incomes less than $200,000, because of the loss of tax preferences, but would help families with income above $200,000, due to the dramatic reduction in the top tax rate. Incorporating the 1997 tax changes—especially the child and education credits-—would only exacerbate these results.

Conclusion

As a replacement for the existing federal tax system, a national retail sales tax is a non-starter. After accounting for evasion, the conversion of state and local taxes, adjustments to keep government benefits constant, and plausible reductions in the tax base, the required rate would be sufficiently high to make enforcement too difficult and evasion too tempting. The historical record should suggest great caution in this regard.

Even if the tax were enforceable at these rates, the implied effects on economic growth would be small at best, and certain sectors of the economy, such as employer-provided health insurance, could be affected significantly. The sales tax would raise burdens on low- and middle-income households and sharply cut taxes on the top 1 percent.

Opposition to the sales tax is not a partisan issue. Robert McIntyre of Citizens for Tax Justice decries the sales tax as unworkable. The Wall Street Journal calls the sales tax a "VAT [value-added tax] in drag" and highlights concerns about administration and enforcement. Bruce Bartlett of the National Center for Policy Analysis concludes that the sales tax is a poor choice for replacing the tax system. Joel Slemrod of the University of Michigan pronounces the sales tax as NAUSEA-I, an acronym for "not administrable at usual standards of equity and intrusiveness."

These problems with the sales tax do not imply that tax reform is a bad idea. Improvements to the tax system are sorely needed. Other types of consumption taxes, like value-added taxes or variants of the flat tax, should receive careful consideration. Many of the gains of moving to a consumption tax could also be obtained through judicious modifications of the income tax. Since income tax reform may also generate much smaller transitional problems, it should be considered seriously, and first.

SERIES: Brookings Policy Brief Series | # 31