Plaintiffs are airlines engaged in interstate commerce. They all appealed the assessed value of their respective properties and, in addition, allege one common claim: that their property is assessed at a higher ratio to fair market value than other commercial and industrial property in violation of 49 U.S.C. § 1513(d).1 The cases were consolidated for trial of this common claim.
In response to complaints of discriminatory taxation, Congress exercised its constitutional authority to regulate interstate commerce by prohibiting the states from taxing railroads, motor carriers and airlines in a discriminatory manner. Although 49 U.S.C. § 1513 allows reasonable taxation of air carrier transportation property, it prohibits state and local taxes which "unreasonably burden and discriminate against interstate commerce." The statute defines specific acts which are prohibited, one of which is:
"[A]ssess air carrier transportation property at a value that has a higher ratio to the true market value of the air carrier transportation property than the ratio that the assessed value of other commercial and industrial property of the same type in the same assessment jurisdiction has to the true market value of the other commercial and industrial property." 49 U.S.C. § 1513(d)(1)(A).2
The airlines claim that the Department of Revenue and Multnomah County violated this provision by assessing the airlines' property at 100 percent of its real market *Page 361 value, while assessing other commercial and industrial property at a ratio of only 25 percent.3
Approximately 95 percent of airline property in Oregon is in Multnomah County. Accordingly, the parties stipulated that the assessment ratio of other commercial and industrial personal property in Multnomah County will be used for the determination of the airlines' claim. The parties also stipulated that the court's determination of the ratio for the 1993-94 tax year will be applied to all years before the court. Finally, the parties stipulated that the airlines' air carrier transportation property is assessed at 100 percent of its real market value. This eliminates the issue of the ratio of the airlines' assessed value to the real market value of their property. Thus, the court need only determine the ratio of assessed value to real market value of other commercial and industrial personal property. At trial, the parties submitted evidence within the parameters set by these stipulations.
One issue that must be clarified is: What does the statute mean by the "same" type of property? The statute defines "commercial and industrial property" to mean:
"[P]roperty, other than transportation property and land used primarily for agricultural purposes or timber growing, devoted to a commercial or industrial use and subject to a property tax levy; * * *." 49 U.S.C. § 1513(d)(2)(D).
Inasmuch as the statute expressly excludes transportation property from the comparison class, it follows that Congress did not intend the words "same type" to mean property with the same function. Rather, "type" must refer to a class of property in the sense of whether it is tangible or intangible, real or personal. The parties have correctly assumed that the airlines' tangible personal property is to be compared with other commercial and industrial tangible personal property, which is consistent with similar claims made regarding railroads and motor carriers. In ABF Freight Sys., Inc. v. Tax Div. of Ark.,787 F.2d 292 (8th Cir 1986), the court held that the assessed ratio of motor carrier *Page 362 personal property may not be compared with the assessed ratio of real property. The court stated:
"Additional support for the proposition that only like categories of property may be compared is found in a recent congressional amendment that prohibits discriminatory state property taxation practices against air carriers. The statute provides that air carrier property may not be assessed at a higher ratio than the assessment ratio of 'commercial and industrial property of the same type.' See 49 U.S.C. § 1513(d)(1)(A). The legislative history indicates that the purpose of the amendment was to make current law prohibiting property tax discrimination against motor carriers applicable to air carriers. See S. Rep. No. 494, 97th Cong.2d Sess., reprinted in 1982 U.S. Code Cong. Ad. News 781, 1188. Thus, it is clear that Congress intended that under the MCA 'property of the same type' is to be compared in determining whether a state taxation scheme discriminates against motor carriers." Id. at 298.
The airlines contend that the ratio of assessment for their property must be compared with the ratio of all other taxable personal commercial and industrial property in Multnomah County. The airlines' evidence consisted primarily of a study by Dr. Roy Bahl and Dr. Dick Netzer, both respected economists.4 The study estimates the market value of taxable, commercial and industrial, personal property in Oregon and Multnomah County and compares that value with the assessed value of such property for the 1993-94 tax year.
The Bahl study uses data collected by the U.S. Census Bureau and the U.S. Department of Commerce to estimate the market value of personal property by type of industry and by class of equipment. One main source of the data is the U.S. Department of Commerce report entitled Fixed Reproducible Tangible WealthIn The United States, 1925-89. This data is collected and compiled by the Bureau of Economic Analysis (BEA) based on information obtained from various industries reporting the replacement cost of their personal property. The reported amount of wealth in each class is based on a perpetual inventory method which *Page 363 consists of adding new "stock" each year and deleting stock which has been discarded. The report states:
"[C]urrent-cost estimates for the net stock represent the depreciated value of all items in the stock at the prices of the current period if assets depreciate in a straight-line manner."
One problem faced by Dr. Bahl is that the data available is only on the national level. To obtain estimates of value for Oregon and Multnomah County, Dr. Bahl allocated the national industry values based on the number of employees in each industry. To make the allocation, he first divided each United States industry value by the total number of employees for that industry in 1993, resulting in a value per employee. However, there were no 1993 employment figures available for Multnomah County. To allocate values to Multnomah County, Dr. Bahl used the 1991 figures. Assuming there was growth in employment in Multnomah County from 1991 to 1993, this would have the effect of attributing less value to Multnomah County. He readily acknowledged that the allocation assumes the value of personal property per employee within an industry is the same in Oregon and Multnomah County as it is in the rest of the United States. Dr. Bahl concluded this assumption was reasonable because there is no evidence to the contrary. Also, he reasoned that an employee in a particular industry in New York will utilize the same amount or value of property as an employee in a particular industry in Oregon.
The focus of the Bahl study is on tangible, taxable, personal property. Dr. Bahl excluded all of the equipment of every manufacturing firm because, in Oregon, much of the manufacturing equipment is classified as real property for taxation. No data exists that would make it possible to ascertain what proportion of manufacturing equipment is classified as real property. Dr. Bahl recognized that by excluding all manufacturing industries, he was excluding substantial amounts of personal property that is taxable and should be included in the ratio. He also excluded certain categories of assets used in nonmanufacturing firms because some of the equipment in those categories is also taxed as real property. Dr. Bahl did not use the BEA *Page 364 data for industries whose property is centrally assessed, such as transportation, communications and utilities. Instead, he assumed that all centrally assessed property is personal property and further assumed that it is assessed at 100 percent of value. Finally, because firms with less than $3,000 of personal property are not taxed under Oregon law, Dr. Bahl allowed a $3,000 exemption to every account. With all of these exclusions and allowances, Dr. Bahl calculates that only 75.5 percent of the value of all commercial and industrial personal property in Multnomah County is assessed. Without these adjustments, the ratio would be only 22.1 percent.
The department's evidence consisted of an explanation of how the personal property tax system is administered in Multnomah County and a ratio study performed by the department. Oregon statutes require all owners or users of tangible, taxable, personal property to file a return each year. Multnomah County employs one auditor full time to audit these returns. The county also contacts business addresses once every three years and checks on new business licenses and other sources of information. The department acknowledges that the penalties for failure to report or file a return are minor.
County property tax administration is subject to performance review by the department. With overall responsibility for supervision of the property tax laws, the department checks the accuracy and completeness of personal property reporting. The counties are required to implement administrative changes recommended by the department. Also, taxation of tangible personal property is based on values using the cost approach to valuation. The department checks the market for costs and prepares depreciation schedules for use by the counties.
The department prepared a ratio study of Multnomah County for these cases. The objective of the ratio study was to determine the general level of assessment of personal property within the county.
The study was performed by taking a random sample of all active commercial and industrial accounts for the 1993-94 tax year. Some 23,707 accounts were stratified *Page 365 by value into five strata, with approximately $275 million of assessed value in each strata. The department then took 100 random samples: 16 from the first strata; 20 from the second, third and fourth strata; and 24 from the fifth strata. Due to various factors, it later took an additional 24 samples from the fifth strata, resulting in a total of 124 accounts being audited. One of the accounts in the fifth strata was excluded as an anomaly, so 123 accounts were used to calculate the ratios.
Each of the selected accounts was examined by two separate sampling methods. In the first method (Sample 1), approximately 10-15 assets were selected randomly from each personal property tax return and their cost was verified against the books and records of the taxpayer. In the process, the auditor also verified the capitalization policies to determine how the taxpayer distinguished personal property from real property. With the second method (Sample 2), each auditor made an on-site inspection, during which the auditor selected 10-15 assets at random. The auditor checked to see whether those selected assets were listed on the personal property tax return. The auditors completed a report for each audit activity and these reports were compiled.
The department devised a formula to calculate a ratio for each strata and an overall ratio. Using this formula, it calculated the mean, median and ratio of aggregates or weighted mean. In the department's view, these calculations showed an overall ratio ranging from 97.69 percent to 100.65 percent or very close to 100 percent. The department also calculated confidence levels of 95 percent.
COURT'S ANALYSIS In reviewing the evidence, the court finds that the Bahl study is interesting but neither pertinent nor persuasive. Even if the study was 100 percent correct in all its calculations and assumptions, it says nothing about the ratio of assessed value to real market value. The study simply indicates that the total assessed value of commercial and industrial personal property in Multnomah County is significantly less than Dr. Bahl's estimate of the value of all such property. It assumes that property that is *Page 366 not assessed must be included in determining the ratio between assessed value and real market value. This is consistent with the airlines' claim that any ratio comparison for purposes of § 1513(d) must include the value of unreported property and property that is not assessed.
1. Whether Congress intended § 1513(d) to include unreported property and property that is not assessed in the ratio is not clear. The language of the statute does not require intentional discrimination such as setting different assessment ratios by statute or regulation. It is not concerned with how the prohibited discrimination occurs, it simply prohibits assessments of higher ratios.
2. The statute requires that a comparison be made between the assessment ratio of air carrier transportation property and the assessment ratio of other commercial and industrial property. An assessment ratio is determined by comparing assessed value with real market value. If property is not assessed, it has no ratio because one of the two elements is missing. If property has no ratio, there is nothing to compare with the assessment ratio of other property.
3. In the railroad cases, which involve very similar language under the 4-R Act, Congress has expressly directed the courts to determine whether the statute is violated:
"[E]ither through statistical sampling of the assessed value and the sale value of individual properties, or through the determination of assessed value and true market value of 'all other commercial and industrial property' in the assessment jurisdiction." Burlington No. R.R. Co. v. Okla. Tax. Comm'n, 481 U.S. 454, 462, 107 S. Ct. 1855, 1859-60, 95 L.Ed.2d 404 (1987).
This language suggests that property which is not assessed is excluded in determining the ratios because such property has no assessed value to compare with its market value.
In Department of Revenue of Or. v. ACF Industries,510 U.S. 332, 114 S. Ct. 843, 849, 127 L.Ed.2d 165 (1994), the United States Supreme Court held that exempt property is not part of the comparison class. Because the 4-R Act *Page 367 limits the definition to property "subject to a property tax levy" the Court found that Congress intended the comparison class to mean "property that is taxed." Id., 95 L.Ed.2d at 175. The Court cited Western Air Lines v. Bd. of Equalization,480 U.S. 123, 131, 107 S. Ct. 1038, 1042, 94 L.Ed.2d 112 (1987). In a footnote the Court states that § 1513(d) is "identical for all relevant purposes" to provisions of the 4-R Act. See alsoBurlington Northern R.R. Co. v. Wis. Dept. of Revenue,59 F 3d 55, 56 (7th Cir 1995) (referring to the Supreme Court's analysis in ACF Industries to support its finding that subsections (b)(1)-(3) of the 4-R Act "only prohibit[s] discrimination between railroad property and other forms of commercial and industrial property which are also taxed"). (Emphasis added.)
4-7. If Congress allows exempting certain property, it is unlikely that Congress intended to include in the comparison class property that escapes taxation as a result of taxpayer evasion or administrative inefficiencies. One may safely assume there will always be some taxpayers who fail to comply with the reporting requirements and that no tax system will perfectly capture all taxable property. Further, to include untaxed property one would have to estimate its value, an estimate which introduces great uncertainty. Uncertainty would arise because any "statistical sampling" would have to try to measure an unknown, i.e., the amount of personal property which has been manufactured or brought into the state and is escaping taxation.
Even if some estimate of the amount of untaxed property could be found, to include it in the comparison class would discriminate in favor of the airlines. If property that is not taxed is included in the comparison class, the comparison class will always have an assessment ratio of less than 100 percent. If such a ratio is then used to reduce the airlines' ratio to a comparable percentage, the level of taxation of the airlines' property will be less than that of other property that is actually taxed. It is unlikely that Congress intended this result. Because the Bahl study calculated ratios using the value of all property, including unreported property and property that is not *Page 368 assessed, it is of little value in determining the airlines' claim.
In addition, the court does not find the data used in the study persuasive. As the department's witness, Dr. Gloudeman, testified, definitions of equipment and structures are critical. No definitions are shown in the Bahl study, and it appears to the court there are probably mixtures of real property or "structures" involved in some of the classes. The uncertainty and lack of clarity of the classifications causes the court to question the results.
In addition, use of the perpetual inventory method raises questions about the accuracy of the figures. The BEA report states:
"Wherever possible, the industry investment estimates are based on capital expenditures data collected from the industry; where capital expenditures data are not available, investment estimates are derived from industry balance sheet data as the change in net stocks plus depreciation. For years where neither capital expenditures data nor balance sheet data are available for a particular industry, the investment estimates are interpolated and extrapolated using related series; wherever possible, these related series consist of expenditures for the types of capital goods purchased by the industry."
In summary, the macroeconomic data used by economists for various projections and studies may be adequate for those purposes, but they provide an imprecise basis to grant tax relief. They lack the capacity to be verified and have only a tenuous relationship to the specific properties and assessed values with which the court is concerned.
8. The department's ratio study was likewise accompanied by problems. Some of the audit work papers were not properly completed by the auditors, raising questions about the accuracy of the audits. For example, Exhibit 17 showed an aggregation of three items totaling $6,830,618 original cost but only $91,747 was reported. The audit report gave no explanation as to why there were only three items audited, and it provided no indication of whether the auditor checked the assets. *Page 369
The audit disclosed that taxpayers routinely failed to report noninventory supplies and misclassified that type of property. In one instance, the audit formula resulted in a $1,500 error being translated into a $500,000 error. Likewise, in some instances, the auditor wrongly included assets from Sample 1 in Sample 2. Finally, the department's own witness conceded that the calculated confidence levels were not what they appeared to indicate.
The airlines claim that the department's ratio study supports their position. In making this claim, the airlines attack the ratio study as flawed in its application of statistical principles. The airlines assert that the study needs to obtain the total dollar value of a population and that the ratio of aggregates is the only method to do that. This claim is made because, as a measure of central tendency, the ratio of aggregates is mathematically verifiable.
Despite the use of precisely defined terms in a mathematical framework, the world of statistics appears a world of shadows and mirrors. It is not clear to the court that the ratio study needs to obtain a total dollar value of a population. The study seeks to determine the "general level of assessment." To do this, it needs to obtain the total dollar value of samples, not the total value of the population.
9. Moreover, the choice of which measure of central tendency is best is a matter of judgment. There are strengths and weaknesses associated with each measure. While the ratio of aggregates may be subject to precise mathematical proof, it is susceptible to sampling error. The department's experts chose to use the median or midpoint. While this measure of central tendency is less precise, it is more forgiving of sampling error.
*Page 370"The median has several advantages in ratio studies. It is easy to compute and interpret. Because it discounts the effects of extreme ratios, the median is little affected by data errors, unlike other measures of central tendency. The median is also the base from which the coefficient of dispersion, the primary measure of appraisal uniformity, is calculated. Finally, the sample median provides an unbiased estimate of the population median."
10. In this instance, the sample size (123 accounts out of 23,707 accounts) was relatively small. More importantly, the ratio study did not have the benefit of market sales. The determination of real market value made by the appraisers was based on their individual judgment and knowledge of costs, not on actual sales of property. In short, the ratio study was not a comparison of assessed values with sale prices, but a comparison of assessed values with estimated values. This scenario is more susceptible to errors. Consequently, it is reasonable for the expert to choose the median over the ratio of aggregates.
11. Account No. 83 was properly excluded from the department's study. If that account was included in the calculation of the ratio of aggregates, it would distort the result. The sample design was not intended to exclude accounts like 83 but it well could have been. Account No. 83 was a situation where a taxpayer failed to file a personal property tax return. Being aware of the property but not of its real market value, the assessor placed a minimal value of $3,050 on the roll. The assessor's practice of placing an account on the roll at the minimal value may be of some benefit to the administrators, but it is not a true assessment. That practice also highlights a weakness in the personal property tax system. Although the county administrators are not convinced that adding personnel is cost effective in the sense of producing more revenue, it appears necessary in Multnomah County for timely completion of the rolls and assuring compliance with the filing requirements.
The airlines also contend that the department's ratio study actually contains three samples instead of two. That is an error only if the study must include unreported property. The court has determined that section 1513(d) does not require the inclusion of unreported property and, therefore, any error is of no consequence.
The department's ratio study calculates a central tendency using three measures: the mean, the median, and the ratio of aggregates. In its calculations with "appraiser judgment removed," the overall ratios range from 98.27 percent to 101.50 percent. *Page 371
Based on the evidence, the court finds that the airlines failed to show that their air carrier transportation property is assessed at a higher ratio to its real market value than the ratio of other commercial and industrial personal property. Accordingly, they failed to show a violation of 49 U.S.C. § 1513(d). This determination shall be incorporated into the judgments entered in each of the above-entitled matters. Costs to neither party.
1 This section was recodified as 49 U.S.C. § 40116(2)(A).
2 Although the federal statute uses "true market value," the court will use Oregon's constitutional standard of "real market value" because they are essentially the same and it will avoid confusion.
3 The allegations in the many complaints vary as to the ratio of assessment for each airline, but all of the complaints claim that the ratio used for other commercial and industrial property is only 25 percent.
4 Because Dr. Bahl testified and claimed responsibility for the study, the court will refer to the study as the Bahl study. *Page 372