By William H. Hoyt, Mark C. Berger, and Paul A. Coomes(*)
From Foresight, Vol. 8, No.4
published 2002
The Kentucky Long-Term Policy Research Center (KLTPRC) and the University of Kentucky Martin School of Public Policy and Administration recently published a report on taxation in Kentucky, ostensibly to improve the discussions being held about state revenue growth and various tax modification proposals.(1) The report included chapters on public finance terminology and theory, trends in state tax policy, sales tax issues, tax equity, local government financing, and business taxes. The chapters were written by some of the participants in the KLTPRC and Martin School conferences held in 2001. Chapter five, written by Professor Lawrence Lynch, focused on business taxes in Kentucky and includes some critical comments on research we performed.(2) In this short essay we respond to Professor Lynch’s comments and use the opportunity to amplify some points we have made elsewhere.
We contest many of Professor Lynch’s statements about our work. We do not feel that he fully and accurately represented our analysis and conclusions. Moreover, when he performed his own limited analysis of business taxes, he found many of the same patterns we found. Professor Lynch repeats several myths about taxation in general, and fiscal arrangements in Kentucky, and we take this opportunity to communicate some important facts about state business taxes.
Professor Lynch did not accurately represent our conclusions. He states in his first paragraph that our study “found that Kentucky has above-average business tax burdens.”(3) We never came to that conclusion. Rather we concluded that Kentucky is “more reliant than its neighboring states on taxes collected from businesses.”(4) The distinction between “tax burden” and “tax collections” is a very important one, and we spend about one half of our report carefully explaining this. Economists agree that all taxes, in the end, are paid by persons, not businesses. Businesses collect many taxes for governments, and then pass them forward in the form of higher prices or backward in the form of lower payments to workers and owners of capital and land. In fact, this point is made in the first chapter of the KLTPRC and Martin School report by Professor David Wildasin when he says, “Businesses as such can never truly bear any tax burdens, they can only distribute tax burdens to natural persons.”(5) Lynch himself makes the same distinction, stating, “Business entities themselves do not bear the burden of taxes,”(6) yet continues to use the term “business tax burden.”
Professor Lynch states that our results are skewed because of our “peculiar choice of tax base,”(7) private earnings. We certainly do not believe that our measure is perfect, but we believe it is as good as any other alternative and certainly better than some. It is clearly better than personal income, a base Professor Lynch seems to prefer. As Professor Lynch notes, “Private earnings consist of wages and salaries, employee benefits, and proprietors’ income.”(8) He evidently prefers using total state personal income as the tax base, because “total personal income includes property income (dividends, interest, and rent) and transfer payments (such as Social Security benefits) as well (Lynch’s italics).”(9) Another important distinction between private earnings and total personal income that Lynch does not mention is that total personal income includes the earnings of all government workers in the state while private earnings do not.
While it would be desirable to include measures of profits and rents realized by Kentucky businesses in our measure, the property income included in total personal income in Kentucky measures the profits and rents received by Kentucky residents, not profits and rents earned by Kentucky businesses. Given that individual household portfolios, pensions, and retirement plans are likely to be invested in firms located throughout the world, there is little reason to believe that the property income of residents of Kentucky is a good measure of the property income derived from Kentucky businesses. By contrast, our earnings base is on a place of work basis, and therefore reflects economic activity in the state.
Why would one want to include the pay of government workers, Social Security payments to retirees, and Medicaid payments to poor persons as part of a business tax base? Transfer payments to Kentuckians in 1999 amounted to $15.5 billion, 30 percent of the amount of earnings paid to workers and proprietors in the private sector. Given that Kentucky has the second highest ratio of transfer payments to private earnings among the eight states in our analysis (see Figure 1), using personal income as the tax base in a calculation of business taxes definitely biases the resulting effective tax rate down. Again, Professor Lynch notes “…Kentucky’s ratio of private earnings to total personal income is lower than in five of the seven contiguous states. Thus, using private earnings … inflates (our italics) Kentucky’s tax-burden ranking.”(10) In other words, Professor Lynch says that using private earnings makes it appear as though Kentucky’s business tax collections are higher than they really are. Professor Lynch is correct in pointing out the greater reliance on transfer payments in Kentucky. However, he errs in his interpretation—use of personal income deflates Kentucky’s tax collection ranking: the use of personal income makes Kentucky’s business tax collections appear lower than they really are. Now which tax base is “peculiar”? (See Figure 1.)
Professor Lynch states that “Private Gross State Product (GSP) would have been a better tax base, both conceptually and in fairness to Kentucky.”(11) While there are some advantages to using GSP, there are numerous problems with using it as a measure of business taxpaying capacity that Professor Lynch failed to mention. Professor Lynch defines Private GSP as “the sum of all value added by businesses (total sales minus purchases from other businesses).”(12) The Bureau of Economic Analysis (BEA) defines GSP as “the value added in production by the labor(13) and property located in a state.” There are, we think, some important distinctions between GSP (and even private GSP) and an ideal measure of a business tax base that need to be made clear.
The BEA prepares GSP for 63 industries. For each industry, GSP is composed of three components: 1) compensation of employees; 2) indirect business tax; and 3) property-type income. The compensation of employees found in GSP is also included in our measure of private earnings and, in fact, comes from the same source as the BEA uses in calculating GSP. Indirect business taxes are not included in our measure but are included in GSP. The problem with inclusion of these taxes, specifically federal business taxes, is that these are not resources available to the state government to tax—what the federal government taxes cannot be taxed by state governments. Indeed, a Department of Treasury determination of total taxable resources specifically notes this as a shortcoming in using GSP as a measure of taxable resources.(14)
The third component of GSP is property-type income. One component of this, proprietors’ income, is also included in our private earnings measure. A second component is capital charges or, more familiarly, corporate profits. This, we admit, would be a useful measure to include; however, caution should be taken in interpreting how corporate profits are measured. In contrast to private earnings, corporate profits are not obtained by information reported by businesses, but instead are calculated based, in part, on wages and salaries, our measure of tax base.(15) Perhaps what is most problematic with using GSP as a measure of the tax base for businesses is the inclusion of imputed rent from owner-occupied housing in GSP. Nationally, real estate makes up a huge share of GSP, 13 percent in 1998.(16) Much of this is attributable to the imputed rent from owner-occupied housing. While the rental stream of housing services is certainly a component of the value added in production by the labor and property, it is not a measure of the tax base of businesses in the state.
The base for state business taxes, as we note in our article, cannot be summarized by a single simple measure. The base depends, in part, on how states choose to structure their taxes, that is, what they choose to tax. What we offered, private earnings, is a proxy for the taxable resources from businesses in a state. State Private GSP, Personal Income, and total taxable resources are also proxies. Every proxy, including private earnings, is incomplete and imperfect, but we see no reason why the difficulties that arise from using private earnings are any worse than those that arise from using private GSP. There are certainly fewer difficulties that arise from using private earnings than from using personal income.
Professor Lynch says that our omission of local taxes in our study of state taxes is “deceptive.”(17) That’s pretty strong language and is also unfair. Any reader can see that we computed business tax collections both ways—state taxes only, and state and local tax collections combined. We explain Kentucky’s very high reliance on tax collection at the state level, how it affects the calculations, and why business tax collections are above average even taking local taxes into account.
Professor Lynch then presents his own estimates, using four different tax bases, but including local taxes in the numerator.(18) But there is an important problem that arises when doing a simple comparison across states using average local tax rates that should be mentioned: local tax structures and rates vary tremendously around Kentucky.(19) Local occupational taxes, city and county property taxes, public school taxes on property and wages, insurance premiums taxes and the like are relatively high in the major urban areas. Much of the rest of the state is characterized by very low property tax rates for local government, no local occupational taxes, and no public school occupational taxes.(20)
Figures 2 and 3 illustrate the differences in local taxes across counties in Kentucky for 1997. These figures show how counties differ in total local taxes as a fraction of private earnings and total personal income. Figure 2 shows a wide variation in local tax rates from 2.5 percent to 13.57 percent based on private earnings (by place of work). Note that if we use personal income (by place of residence) as a base rather than earnings (Figure 3), the effective tax rate is much higher in large metropolitan counties that are employment centers such as Fayette and Jefferson. Given that most of the mobile businesses in Kentucky are located in the urbanized areas, any study of business taxes should be weighted towards those higher tax localities. Averaging in the rural, low local tax areas with the urban, high local tax areas biases the result in a state like Kentucky. From this point of view, when studying state tax collections from businesses, which was the purpose of our study, one should examine state and not local taxes. State tax rules and rates are the same for every Kentucky business, whereas local taxes vary widely from place to place. Instead of being “deceptive,” our statewide analysis provides useful information about taxes in Kentucky, regardless of where the business locates.
Figure 2: Total Local Tax Rates as a Percentage of Private Earnings, 1997
Figure 3: Total Local Tax Rates as a Percentage of Total Personal Income, 1997
Finally, Professor Lynch says that our “study gives a misleading picture of Kentucky’s relative business tax burden.”(21) But we did not study the business tax burden! What we did do is present a fairly accurate comparison of state business tax collections in Kentucky and surrounding states, and found Kentucky’s collections to be somewhat higher than the average. Evidently, if one points out something other than the conventional wisdom, then, according to Professor Lynch, one is misleading policymakers.
* Professors Berger and Hoyt are economists at the University of Kentucky and Professor Coomes is an economist at the University of Louisville. Here, they respond to an article by Professor Larry Lynch, an economist at Transylvania University, in Financing State and Local Government, in which he challenged their conclusions about business taxes in Kentucky. Dr. Lynch stands by his work. Interested parties are urged to read both the original UK-UL study and Dr. Lynch’s article and draw their own conclusions. Return to text.
1 David E. Wildasin, Michael T. Childress, Merl Hackbart, Lawrence K. Lynch, and Charles W. Martie, Financing State and Local Government: Future Challenges and Opportunities (Frankfort: Kentucky Long-Term Policy Research Center and University of Kentucky Martin School of Public Policy and Administration, 2001). Return to text.
2 William H. Hoyt, Mark C. Berger, and Paul A. Coomes, Statutory and Economic Incidence of Taxes in Kentucky and Surrounding States (Lexington/Louisville: University of Kentucky Center for Business and Economic Research and University of Louisville College of Business and Public Administration, 2001). Return to text.
3 Lawrence K. Lynch, “Business Taxes in Kentucky,” in Wildasin et al., 57. Return to text.
4 Hoyt, Berger, and Coomes (2001) iii. Return to text.
5 David E. Wildasin, “Tax Reform in Kentucky: Principles and Practice,” in Wildasin et al 7. Return to text.
6 Lynch 58. Return to text.
7 Lynch 64. Return to text.
8 Lynch 64. Return to text.
9 Lynch 64. Return to text.
10 Lynch 64. Return to text.
11 Lynch 65. Return to text.
12 Lynch 65. Return to text.
13 Bureau of Economic Analysis (BEA), online help file http://www.bea.doc.gov/bea/regional/gsp/. Return to text.
14 For a detailed discussion of the problems associated with using GSP as a measure of taxable resources, see Michael Compson and John Navratil, “An Improved Method of Estimating the Total Taxable Resources of the States,” Department of Treasury Research Paper No. 9702 (1997). Return to text.
15 We thank George Downey of the BEA for helpful discussions on this point. Return to text.
16 Authors’ calculations based on BEA data. Return to text.
17 Lynch 65. Return to text.
18 In five out of the eight cases he considers, Kentucky’s business tax collections are above the average of Kentucky and the contiguous states (Lynch, Tables 10 and 11). The mean percentage across the eight examples is 102 percent. In other words, Lynch finds that Kentucky’s business tax collections are above the average of Kentucky and the contiguous states, just as we do. Return to text.
19 For a detailed analysis of this issue, see William H. Hoyt, “Differences in Tax Bases and Tax Effort Across Kentucky Counties,” 2001 Kentucky Annual Economic Report (Lexington: Center for Business and Economic Research, University of Kentucky, 2001). Return to text.
20 David E. Wildasin, “Tax Reform in Kentucky: Principles and Practice,” in Wildasin et al, 73-101. Return to text.
21 Lynch 65. Return to text.