The Economic Impact
of an Oregon Sales Tax
By Richard K. Vedder, Ph.D.
The proposed new five percent Oregon sales tax may have far-reaching adverse economic consequences, primarily by reducing the substantial cross-border purchases of goods in Oregon. Oregon has an unusual ratio of retail sales to income almost certainly reflecting a high migration of non-Oregonians into the state to buy non-taxable goods.
Oregonians are considering the imposition of a five percent sales tax. This study analyzes some possible economic effects that the tax imposition may have on the citizens of the state. Special emphasis is placed on the impact cross-border retail sales activity has on Oregon, and what a sales tax might do to that activity.
The sales tax proposal has a number of other provisions that would have significant economic impacts, including some reductions in property taxation, an increase in corporate income taxation, and reductions in income taxes via a new earned income credit. Similarly, the measure provides some limits on state government expenditures. This paper does not analyze the impact of these other changes, but it is acknowledged that they may well be material.
Criteria for Evaluating Taxes
While experts on public finance might disagree on some things, virtually all would agree that the following three factors are important in evaluating the impact of any given tax: administrative and enforcement costs, equity or fairness, and economic effects. Some would add still other factors, but the three mentioned above are factors universally cited.
Most experts would give the sales tax in general (without reference to Oregon) fairly high marks with respect to administrative costs and perhaps economic effects, but low marks with respect to equity considerations. In the case of Oregon, however, there is reason to believe the administrative cost/economic effects problems may play a much more significant role.
Speaking first to equity considerations, what is a "fair" tax is a matter of opinion. A tax that some think is quite fair, others might view as inequitable. Nonetheless, a large body of opinion accepts the "ability to pay" principle of taxation, which says people with greater ability to pay (higher incomes), should pay a larger share of the tax burden. Taxes should be progressive, meaning they take a larger part of the income of higher income individuals than poorer persons.
The evidence is clear that sales taxes are regressive, burdening the poor more than the rich. The removal of food and drugs from the sales tax base alleviates the problem somewhat, but on balance sales taxes remain regressive.(1) Moreover, the food exemption raises administrative costs and interesting policy and compliance problems. For example, should "snacks" (e.g., potato chips, tortilla chips or candy bars) be taxed? Should carry-out restaurant food be taxed, while carry-out grocery food items are not taxed?
"The evidence is clear that sales taxes are regressive, burdening the poor more than the rich."
Less known is the extent to which sales taxes violate the principle of what economists call horizontal equity. Consider two persons of identical economic circumstance, one who spends much of her money on clothes, restaurant meals and appliances, while the second spends much of her money on out-of-state trips. The first person will pay substantially more tax than the second.
The previous example points to another problem with sales taxation, namely the fact that they have a distortive impact on the use of resources. To the extent purchases of clothes in Oregon stores is taxed, but out-of-state travel is untaxed, the State of Oregon is raising the price of clothes purchases in Oregon stores relative to out-of-state trips. This impact is unintended, as there seems to be no compelling reason to discourage persons from shopping in Oregon or to encourage them to take trips outside the state (indeed, it would be expected that Oregonians would prefer to use the tax system to encourage exactly the reverse).
Relatively small states with a large proportion of their population living near state borders face still another situation, namely the existence of cross-border economic activity to avoid taxation. Economic theory would suggest that cross-border activity would be greater when:
People are unlikely to incur the costs of crossing the border from one state to another if they have to drive 100 miles. At the same time, if the lower taxed state is just across a bridge or street, the individual is likely to cross borders to shop. If State A has a sales tax of 5 percent and State B has a sales tax of 5.5 percent, the individual is far less likely to cross borders than if the tax differential is 0 versus 5 or 6 percent. The individual will not cross borders if he or she is unaware of prices or goods available in the alternative designation. They likely would not cross borders to buy a low valued item (say, light bulbs), but would be more likely to do so to buy a carload of goods, or, better yet, the car itself, where the tax advantage likely would reach hundreds of dollars.
At present, Oregon has been favorably blessed by several of the aforementioned economic factors, thereby supporting an immense increase in sales activity in some Oregon establishments from what otherwise would be the case, as will be demonstrated below. Surrounding states, most notably Washington, but also California, Idaho and Nevada, all have sales taxes at a five percent rate or higher, while Oregon has none.
A number of studies indicate that consumers living near borders are quite "price-sensitive", i.e., they respond to tax-induced price differentials in making their purchase decisions. In general, these studies suggest that states with potentially high cross-border activity will assume a significant loss in retail sales when they create or increase sales or excise taxes.(2) States with a tax advantage should hesitate before giving it up by raising rates. As one study put it recently, "Relatively small states can benefit substantially from having low excise and sales tax rates relative to neighboring states."(3)
Oregon's Border Economy
Oregon's lack of a sales tax means that citizens of bordering states can and do cross the border to take advantage of lower prices. This would not have too much meaning if the population of Oregon border areas was sparse, representing only a small proportion of the state's population; the in-migration of outsiders spending money would impact only a small part of the state's economy. However, that is not the case.
If we define the population area likely to be impacted by cross-border activity as being those counties that border on another state, it turns out that 17 counties are involved. These counties have 39 percent of the state's population and nearly 41 percent of its retail sales (1991).(4) More importantly, the state's largest city, Portland, is adjacent to the state of Washington. Such important shopping centers as the Jantzen Beach Mall (Multnomah County) are easily accessible to Washington residents, and the same is true of several smaller Oregon communities outside the Portland area.
If there were absolutely no cross-border activity between states, we would presume that retail sales in each state would be roughly the same proportion of total disposable income. This presumes each state's residents have roughly the same propensity (desire) to consume out of income. However, as indicated, all states surrounding Oregon levy significant (five percent or greater) sales taxes, the incidence of which falls on consumers. Theory would predict that residents of these other states would come to Oregon to buy goods. Thus retail sales should be relatively high in Oregon in relation to disposable income in the state, because of the purchases of goods by outsiders; this ratio should also be higher in Oregon than in "normal" states that do not have such an interstate tax advantage, and also higher than in states bordering on Oregon who lose sales to Oregon by cross-border migration.
Figure 1 shows what is termed the "sales/income ratio" for Oregon, neighboring states, and the U.S. It records retail sales as a proportion of disposable personal income. Note that the ratio for Oregon is significantly higher than that of any surrounding state or the U.S. as a whole. Indeed, within the 48 contiguous U.S. states, only Maine, which borders on high-sales tax Canada, has a higher ratio of sales to income.
Other than tax-related cross-border activity, there are two other possible explanations for the extraordinarily high sales/income ratio for Oregon. First, it is possible that Oregon attracts unusual amounts of tourist business, a special form of cross-border activity unrelated to taxes. The evidence, however, shows this is not the case. Indeed, the data on hotel and motel receipts suggests that Oregon is actually less involved in tourist activity than the U.S. average and less than two of its neighbors, California and Nevada.(5) Its tourist activity in relation to population is only marginally greater than that in Washington state.(6) Oregon's sales/income ratio is far higher than neighboring states with larger tourist orientations. Moreover, estimates of revenues by Oregon's Legislative Revenue Office assume only a small proportion (well under five percent) of revenues would be derived from non-Oregonian tourists, suggesting the tourist presence is not extraordinary or of a magnitude to greatly impact on the sales/income ratio figure. (7)
"The 1991 estimated retail sales per capita for Oregon was $8,184,more than 10 percent above the U.S. total of $7,323, despite the fact that Oregon has a lower personal income per capita than the national average."
Another possible explanation is Oregonians have a peculiarly high propensity to spend their income on goods and services sold within the state, meaning that they do not save much, travel elsewhere, and so forth. While a difficult hypothesis to test, no reliable evidence exists to support the conclusion that Oregonians have a fundamentally different consumption behavior pattern than other Americans. Since personal savings nationwide is only about five percent of disposable income, variations in savings habits almost certainly cannot explain more than a fraction of the huge variation in the ratio of retail sales to income noted in Oregon relative to more typical states.
Therefore, the hypothesis that Oregon retail sales are significantly and positively impacted by tax-related cross-border activity seems to be not only plausible but highly likely. To provide another test, it would be expected that the sales/income ratio would be higher in Oregon border counties (recipients of much cross-border activity) than in interior counties, which are longer distances from state borders and thus less likely to attract such activity. Figure 2 shows the ratio for the two areas. Retail sales in the border counties equaled 62.9 percent of spendable income in 1991, much higher than observed in the interior counties (55.7 percent), and clearly higher than in any state in the Union. This is highly consistent with the view that border counties derive substantial retail sales from out-of-state residents seeking a tax advantage.
"Oregon could see a drop in annual retail sales of $3.2 billion, and a loss of 32,000 jobs"
Another way of looking at the data is to examine retail sales per capita. The 1991 estimated figure for Oregon was $8,184, more than 10 percent above the U.S. total of $7,323, despite the fact that Oregon has a lower personal income per capita than the national average.(8) Oregon's per capita estimate is above all four of its bordering states, including relatively high income and tourist-rich Nevada.
Finally, one category of sales that might be particularly affected by a sales tax differential is automobiles. To examine this issue, data were gathered on the percent of disposable income spent on automobiles in Oregon and Washington. The results are startling. Even though Washington has a law that requires its citizens to pay Washington state sales tax on their Oregon purchases, 8.01 percent of retail sales there were in the form of automobiles, whereas in Oregon it was 11.24 percent. Oregonians seemingly spent 40 percent more of their retail sales dollars on autos than Washingtonians! In reality, we suspect spending was similar in both states, but that many of the Oregon sales came from Washington residents crossing the border. Per capita automotive sales (measured in dollars) were 23 percent higher in Oregon than in Washington, despite the fact that per capita income was nearly 10 percent higher in Washington.
Effects of Sales Taxation:
What are the possible economic effects of a new five percent sales tax in Oregon? Looking first at the impact on economic activity of a new sales tax, let us develop an estimate based on the following assumptions:
Under such a scenario, the sales/income ratio differential between Oregon and the U.S. would sharply decline, with the Oregon ratio reaching 50.27 percent (from 58.40 percent). Oregon could see a drop in annual retail sales of $3.2 billion, and a loss of 32,000 jobs.
These estimates should probably be considered upper bound values. Even if people are sensitive to price changes (and this implies a high price elasticity of demand), it does take time for adjustments to occur. Perhaps in some areas of, say, Washington state, there are not adequate retail sales establishments, so customers would come to Oregon despite the disappearance of a price advantage (although in time that deficiency would likely be overcome by new stores built in Washington to take advantage of rising demand there). Regarding employment, some employers may not cut work staffs proportionately with sales reduction (although failure to do so would adversely affect their profits and income). It is possible that some of the observed sales-income ratio is unrelated to cross-border activity. Perhaps in some cases as sales volume did fall, workers would take pay cuts or work fewer hours, but still remain employed.
"A particularly worrisome dimension of the proposed sales taxis the three percent rate applied to sales of machinery and equipmentnecessary for production. This is not a tax on consumption, but on production."
An extreme critic of this approach might argue that none of the adverse behavioral effects observed above relate to differential sales taxes, that retail sales would not be impacted in Oregon, and the unemployment effects in retail trade would be negligible. Thus a minimum estimate of the job loss might be zero, while the 32,387 figure represents a maximum value. The fact remains, however, that even a middle range estimate between these two extremes, would represent the loss of over 16,000 jobs, enough to raise the state's unemployment rate by a full percentage point.
If the long-term impact of a sales tax is to reduce retail sales by an excess of $3 billion a year (on which sales tax collections might amount to $150 million), it is possible that Oregon's Legislative Revenue Office estimates of receipts from the new tax are materially overstated. This author does not have adequate information on the assumed cross-border effects used in making the official estimates, but his past experience with other states suggests that the possibility of over-estimation should not be excluded.
The Sales Tax and Business Enterprises
Oregon's Legislative Revenue Office estimates calculate the sales tax (and other changes) on households and on businesses.(9) If in fact businesses absorb their estimated share of the tax burden, this lowers profits, the rate of return on business investment and, ultimately, the attractiveness of Oregon as a location for capital investment. If, however, businesses are able to shift the costs of the tax forward to consumers, the actual household tax burdens are greater than what the official estimates indicate.
A particularly worrisome dimension of the proposed sales tax is the three percent rate applied to sales of machinery and equipment necessary for production. This is not a tax on consumption, but on production. As such it raises the price of productive capital investments, discourages the in-migration of capital from other states, and encourages Oregon businesses to locate machinery and production facilities elsewhere. Public finance authorities generally believe a general sales tax should apply only to consumption goods and services, and should exclude resources used to produce those goods and services.
Economic Effects of Sales Taxation: The Expenditure Side
One might argue that the adverse impacts noted above are offset by positive effects of spending generated from the tax increase. For example, more school teachers may be hired, raising the disposable income of those involved, leading in turn to greater spending on retail goods. Offsetting that, however, the disposable income of Oregonians in general will fall by the fact that any given income will buy fewer goods than before. It is not clear whether the negative effects on disposable income are greater or lesser than the positive ones.
Proponents of the new tax might legitimately argue that "human capital"--knowledge and skill--are important resources, and that increases in human capital tend to lead to increased economic growth. Even accepting that argument, however, it must be acknowledged that any gains in productivity from greater skills will be a long time coming, since the recipients of the spending would be children 5, 10 or 15 years of age, still several years away from entering the labor force. This, of course, assumes that the new tax monies will in fact be used for increased education spending, a proposition this author finds dubious, even with earmarking, since spending of non-earmarked funds for education may decline (nationwide, there has been a decline in the proportion of state and local spending that has gone for education over the past generation).
A more fundamental objection to the notion that increased educational funds will improve the learning of young Oregonians is a large amount of research exists that indicates virtually no relationship between educational expenditures and learning.(10) Moreover, this includes massive studies performed by some of the nation's top social scientists. John Chubb and Terry Moe in their seminal study of effective American high schools concluded, "When other relevant factors are taken into account, economic resources are unrelated to student achievement."(11) Similar conclusions are reached by James S. Coleman and associates in an analysis of public and private schools.(12) This author's own research with colleagues using Ohio schools reinforces these findings, noting in addition that a modest, positive relationship between spending and learning is dissipated by the law of diminishing returns and by the fact that much spending goes for unproductive, non-instructional purposes.(13)
In addition to this, there is evidence that the increased spending occasioned by a new sales tax would lower economic growth if the increased funds are largely dispersed to public employees in the form of higher compensation, something that clearly occurred nationally and in Oregon in the 1980s.(14) In fact, roughly one-third of all state and local public employees in Oregon work for the public school system, and their compensation constitutes the majority of public school spending in the state.(15) The redistribution of income from relatively productive private sector employees to less productive workers in the public sector can have debilitating effects on economic performance.(16) If Oregon uses some or all of the incremental funds for public pay compensation, it would be anticipated that this would lower the growth of income and wealth in the state.
Oregonians should be aware that the proposed sales tax almost certainly will impose significant costs on them, particularly in border areas. Thousands of jobs--16,000 being a conservative estimate--will likely be lost in retailing as cross-border shopping declines in importance. That is the equivalent of raising the Oregon unemployment rate by one full percentage point. The tax, applying as it does to machinery, discourages capital investment. Moreover, the national evidence suggests that spending more on schools is not likely to lead to greater student performance. These considerations need to be weighed by Oregonians as they evaluate the latest of nine sales tax proposals since the 1930s.
1. The reader is referred
to any of a number of standard texts or references in public finance at this
point. See, for example, Edgar K. Browning and Jacquelene M. Browning, Public
Finance and the Price System, Third Edition (New York: Macmillan, 1987),
Joseph E. Stiglitz, Economics of the Public Sector (New York: W.W. Norton,
1986), or Joseph A. Pechman, Federal Tax Policy, Fourth Edition (Washington,
D.C.: Brookings Institution, 1983).
2. Two recent examples are Richard Vedder, David Klingaman and Lowell Gallaway, The Economic Impact of Excise and Sales Taxation in Ohio (Columbus, OH, 1992) and Tax and Fiscal Policy Task Force, American Legislative Exchange Council, "Voting With Their Feet II: The Economic Consequences of Cross-Border Activity in the Southeastern U.S.," The State Factor, August, 1993.
3. Price-Waterhouse, "Voting with Their Feet: A Study of Tax Incentives and Economic Consequences of Cross-Border Activity in New England," American Legislative Exchange Council, The State Factor, August, 1992, p. 59.
4. The statistics in this report are mostly derived from estimates prepared for the U.S. Department of Commerce. The author used the compilations reported in Sales & Marketing Management Magazine's annual Survey of Buying Power Demographics USA. That magazine uses a concept called "effective buying income" which is a measure of spendable household income derived from Commerce Department data that is the equivalent of the U.S. government's "disposable personal income." Retail sales estimates are updated by the magazine from data provided in the U.S. Bureau of the Census, 1987 Census of Retail Trade.
5. U.S. Census data suggest that in 1987 Oregon had $154.90 in hotel, motel, and other lodging receipts per capita, compared with a U.S. average of $213.07. The figures for California were $218.00, and exceeded $5000 for Nevada! Similarly, Oregon had 10.42 hotel and motel employees for each 1,000 workers in 1987, compared with a national average of 12.55 and 12.14 in California; the Nevada figure was an extraordinary 219.66. The statistics were derived from U.S. Department of Commerce, Bureau of the Census, 1987 Census of Service Industries, Hotels, Motels and other Lodging Places (Washington, D.C.: Government Printing Office, 1991); employment and population statistics were obtained in the Bureau of the Census, Statistical Abstract of the United States, 1989 and 1990 editions.
6. For example, hotel and motel receipts per capita in 1987 were $144.24 in Washington, about seven percent lower than Oregon's $154.90. See ibid.
7. See Oregon Legislative Revenue Office, "Legislative Tax Reform Proposal, HJR 10/HB2500/HB2443" (August, 1993) for details of official estimates.
8. The per capita sales estimate for Oregon is derived by dividing the 1991 retail sales estimate of $23,262 billion by the 1990 population of 2,842,321. Use of estimated 1991 population statistics would not change the observed relative relationship between Oregon and the U.S.
9. "Legislative Tax Reform Proposal ", p. 4.
10. See, for example, Eric A. Hanushek, "The Economics of Schooling: Production and Efficiency in Public Schools," Journal of Economic Literature, September, 1966, pp. 1141-1177, or his "The Impact of Differential Expenditures on School Performance," Educational Research, Vol. 18 (1989), especially p. 47.
11. John E. Chubb and Terry M. Moe, Politics, Markets & America's Schools (Washington, D.C.: The Brookings Institution, 1990).
12. James S. Coleman, Thomas Hoffer, and Sally Kilgore, High School Achievement: Public and Private Schools Compared (New York: Basic Books, 1982).
13. Luther M. Boggs, Richard Vedder, and Alfred E. Eckes, Testing and Education Achievement: Ohio and the Nation (Contemporary History Institute, Ohio University, 1992).
14. Wendell Cox and Samuel A. Brunelli, "America's Protected Class II: The Widening Public-Private Pay Gap," The State Factor, American Legislative Exchange Council (Washington, D.C.: January, 1993). See also: Richard Vedder, "Economic Impact of Government Spending: A 50-State Analysis," NCPA Policy Report #178 (Dallas: National Center for Policy Analysis, 1993). It is interesting to note that upon review, estimates of per capita income in Oregon would have been $1,516 higher in 1990 if public employees in the state had received compensation increases in the 1980s that only equaled those of private sector persons (rather than greatly exceeded them).
15. The percentage of public employees who work for schools is derived by dividing the number of public school employees listed in The Oregonian article, "Schools and the Sales Tax," September 26, 1993 (as reported by the Oregon Department of Education), by the U.S. Census Bureau annual report of the "Number of Public Employees by Level of Government and State," October, 1990 (Table 6). According to another Oregonian article ("School Funding," April 4, 1993), a majority of public school spending is in the form of public employee compensation--"About 80 percent of most district general funds go for salaries and benefits."
16. There is an immense amount of literature arguing that where public sector activity is relatively high or growing (as measured by taxes and/or expenditures), economic growth is relatively low, presumably reflecting differential productivity between the two sectors. A couple of representative studies include Alaeddin Mofidi and Joe A. Stone, "Do State and Local Taxes Affect Economic Growth?" Review of Economics and Statistics, November, 1990, and Bruce L. Benson and Ronald N. Johnson, "The Lagged Impact of State and Local Taxes on Economic Activity and Political Behavior," Economic Inquiry, July, 1986.
Professor Richard K. Vedder has written over 100 articles and six books and monographs. His latest book is Out of Work: Unemployment and Government in Twentieth-Century America co-written with Lowell E. Gallaway. He has written extensively on the economic impact of state and local taxes and spending.
He is currently distinguished professor of economics and faculty associate,
Contemporary History Institute, Ohio University. His other books include The
American Economy in Historical Perspective and Poverty, Income Distribution,
the Family and Public Policy.
Cascade Policy Institute is a tax-exempt educational organization as defined under IRS code 501(c)(3). Cascade neither solicits nor accepts government funding, and is supported by individual, foundation, and corporate contributions. Nothing appearing in this document is to be construed as necessarily representing the views of Cascade, or as an attempt to aid or hinder the passage of any bill before any legislative body.
Copyright 1993, Cascade
813 SW Alder Street
Portland, Oregon 97205
Phone (503) 242-0900
Fax (503) 242-3822