Expert Answer :This study examines the long-term effects of

Solved by verified expert :r 2008 Public Financial Publications, Inc.Does Earmarked Revenue Provide Property TaxRelief ? Long-Term Budgetary Effects ofGeorgia’s Local Option Sales Tax*ZHIRONG JERRY ZHAO and CHANGHOON JUNGThis study examines the long-term effects of the 1% General-purpose LocalOption Sales Tax (LOST) on the level of property tax in Georgia counties with apooled interrupted time-series analysis. The LOST has been earmarked forproperty tax relief in Georgia counties since 1976, but debates remain onwhether the proceeds have been used as additional revenues. We find that theadoption of LOST brought short-term property tax relief but not long-termproperty tax reduction. The result suggests that long-term property tax reliefwould not be realized by earmarked revenue without careful policy design tosafeguard fungibility.INTRODUCTIONIn the wake of the ‘‘property taxpayer revolts’’ that have swept the nation since the late1970s, local officials find it more difficult to increase property taxes to meet the growingservice demands. In order to respond to growing service needs, an increasing number oflocal governments have adopted local sales taxes and other revenue sources (e.g., localincome taxes, user charges and fees, and impact fees). Since the mid-1970s, a growingnumber of states have authorized the levying of a local sales tax by at least some of theirmunicipalities and counties. By 2003, local sales taxes have been used in thousands oflocal governments in 33 states.1 In most states, the adoption of local sales taxes wasintended to provide a means of funding additional service delivery and/or to diversify*We appreciate the helpful comments of two anonymous reviewers on various aspects of this manuscript.Zhirong Jerry Zhao is an Assistant Professor, Hubert H. Humphrey Institute of Public Affairs, Universityof Minnesota. He can be reached at [email protected]Changhoon Jung is an Associate Professor, Department of Political Science, Auburn University. He can bereached at [email protected]1. Council of State Governments, The Book of the States 2003 (Lexington, KY: Council of StateGovernments, 2003), 350.52Public Budgeting & Finance / Winter 2008 local revenue systems.2 However, in a few states, local sales taxes were also authorized toprovide either a direct or indirect means of property tax relief. To use the local sales tax,some states mandated that part or all of the sales tax proceeds be used to reduce theproperty tax in the form of a limit on tax rates, a limit on assessment increases, a limit onthe tax levy, or a combination of two or more of these limits.3 For example, in Georgia,the proceeds of the 1 percent general-purpose local option sales tax (LOST) are requiredto be used for the reduction of property tax in the second year of the tax and allsubsequent years.4 Likewise, state law in Wisconsin mandates that the full amount ofLOST proceeds be used to roll back the property tax during the next year.5 In SouthCarolina, 63 percent of first-year LOST revenues must be used for a property tax rollback, and the proportion increases in subsequent years.6 Moreover, in Iowa, mostcounties earmark a specific portion of their LOST proceeds for property tax relief,although mandatory reduction of the property tax is not required by the state.7 Unlikethe aforementioned states, many states do not require a direct relief of the property taxburden in authorizing their local sales tax. However, the use of sales taxes, even in thesestates, may provide indirect property tax relief by preempting the need for an increase inproperty tax revenue to finance local service delivery.8Although thousands of local governments in several states are using local sales taxeswith mandatory property tax reduction requirements, only a few studies have systematically studied the effects of local sales taxes on the property tax relief to date, andalmost no studies have analyzed the long-term effects of such taxes on property tax relief.While a local sales tax earmarked for property tax relief is currently levied in Georgia,South Carolina, and Wisconsin, Georgia is one of a few states that adopted the localsales tax dating back in the late 1970s, and thus, the state provides an excellent case toinvestigate the long-term effects of local sales taxes on property tax relief. Based onprevious studies regarding earmarking behavior, this paper investigates the long-termeffects of local sales taxes on property tax relief in Georgia counties during the years1984–2002 by employing a pooled interrupted time series research design. In so doing,this study tests competing scenarios concerning the fungibility of earmarked revenuesources to explain the long-term budgetary effects of earmarked revenue sources.2. Robert L. Bland, A Revenue Guide for Local Government, 2nd ed. (Washington, DC: ICMA, 2005):105–120.3. Daniel Mullins and Sharon Cox, Tax and Expenditure Limits on Local Governments (Washington,DC: Advisory Commission on Intergovernmental Relations, 1995).4. Changhoon Jung, ‘‘Does the Local Option Sales Tax Provide Property Tax Relief? The GeorgiaCase,’’ Public Budgeting & Finance 21, no. 1 (2001): 73–86.5. Wisconsin Taxpayers Alliance, Do County Sales Taxes Lower Property Taxes? vol. 70 (Madison, WI:Wisconsin Taxpayers Alliance, 2002).6. Holley H. Ulbrich, ‘‘Local Option Sales Taxes and Municipal Finance in South Carolina: A Look atthe First Few Years,’’ paper prepared for the Municipal Association of South Carolina, 1996.7. See Iowa Department of Revenue and Finance (2003). 2003 County Financial Overview; availablefrom: http://ww.iowacounties.org/fiscalinfo/cfo/cfo2003.htm: Retrieved 22 April 20048. Steven Gold, Property Tax Relief (Lexington, MA: Lexington Books, 1979), 207.Zhao and Jung / Does Earmarked Revenue Provide Property Tax Relief53 Understanding the long-run budgetary effects of LOST on property tax relief will greatlyenhance our knowledge concerning the local fiscal behavior of earmarking practices.This paper is organized as follows. The next section briefly provides backgroundinformation on the use of LOST in Georgia counties. The third section reviews literature,while the fourth section lays out the conceptual framework and research hypotheses ofthe study. The fifth section discusses the research method, and the sixth section presentsthe research findings. The final section concludes the paper with a discussion of theimplication and future direction of the study.BACKGROUND INFORMATION ON THE LOST IN GEORGIA COUNTIESThe General-purpose Local Option Sales Tax (LOST) Act was first passed in 1975 andamended in 1976 in the Georgia General Assembly. Upon the approval of voters, the Actpermitted counties to enact a one-percent general-purpose sales and use tax which was tobe shared between the county and the cities within the county.9 The tax can be collectedunless it is repealed by a referendum or by a county and its qualified municipalitiespursuant to an agreement negotiated by the county government and these cities based oncriteria established by general law.10 Since the adoption of the tax, no counties haverepealed the tax to date, and thus, the tax has almost become an institutionalized revenuesource for local governments in Georgia.In order to achieve the goal of property tax relief, the LOST Act mandated that allLOST revenue collected in the second year of the tax and all subsequent years be used fora reduction of property taxes. The LOST revenues are used to reduce county millagerates across the board on all taxable property within the county, including propertywithin municipalities located in the county. If a city located in a county collected its ownLOST independent from the county, taxpayers who own property within the incorporated city also have their city millage rate reduced for all taxable property within the city.The Act required that the tax bill of all property taxpayers show the reduced county andcity millage rates resulting from the receipt of sales tax revenue from the previous year, aswell as the reduced dollar amount of the property tax resulting from the receipt of suchrevenue.11 Since the law specified that all LOST revenues were to be used for a dollar-fordollar property tax relief in the second year of collection, a property tax rollback appearsto be certain in the second year of the collection. However, after the second year, thelegislation is ‘‘less explicit’’ about the use of LOST revenue for property tax rollbacks,although it requires local governments to show the millage rates and the amount of the9. In counties in which the county government did not enact the LOST, cities could, with voter approval, impose their own sales tax. However, no city has ever adopted the LOST independently.10. Betty J. Clements and Devereux Weeks, County and Municipal Revenue Sources in Georgia (Athens,GA: Carl Vinson Institute of Government, 1990), 4.11. Clements and Week, 5. City governments, however, were not required to roll back property taxesafter the second year.54Public Budgeting & Finance / Winter 2008 property tax rollback resulting from the collection of the LOST.12 Because it is verydifficult to determine what the property tax rate (or dollar amount) would have been inthe absence of LOST revenue, a ‘‘paper rollback’’ is possible, and as a result, a portion ofthe LOST may be a new revenue source for at least some counties.13The LOST has been a very popular revenue source for most local governments, andthe number of counties adopting it has gradually increased since its inception in 1975.Thirteen of the 159 Georgia counties adopted the tax in 1976 and by 1979, the numberhad increased to 82 counties. It further increased to 104 by 1981, 131 by 1984, 144 by1988, 148 by 1992, and 153 by 1997. By January 2004, all but five Georgia counties hadadopted the tax.14 Additionally, LOST proceeds now make up 7–35 percent of generalrevenue sources in counties, which makes the tax the second largest source of revenuefollowing property taxes.15LITERATURENonproperty taxes, such as local sales taxes or local income taxes could provide propertytax relief in two ways.16 First, when the adoption of nonproperty taxes is earmarked forproperty tax relief (e.g., dollar-for-dollar reduction requirement), a direct provision ofproperty tax relief would be highly likely at least in the short-term if the enacting legislation requires a strict implementation of the mandatory relief. However, if there is norigorous requirement of continued or subsequent reduction of the property tax burdenby local governments, there may be high probabilities that the true long-term reductionof the property tax burden would not be realized in the midst of growing service needs oflocal governments. Few states (Georgia, South Carolina, and Wisconsin) require aproperty tax relief with LOST, and the review section of this literature primarily examines the effects of such direct earmarking on property tax relief. Second, although theadoption of a nonproperty tax does not require property tax relief, the collection ofadditional revenue sources could provide property tax relief indirectly because the additional available revenue may make it possible for local governments to rely less heavilyon the property taxes. Efforts to diversify the local revenue structure in many states mayhave resulted in such indirect property tax relief.Despite the theoretical and empirical importance of the effectiveness of earmarking,few studies have examined the effects of earmarked local sales taxes on the property tax12. Dan Durning, Distributing Georgia’s General Purpose Local Option Sales Tax Revenues: An Examination of the Present Policy and Some Options (Athens, GA: Carl Vinson Institute of Government,1992), 6–7.13. Durning, 29.14. Georgia Department of Revenue, Historic Rate Chart (Atlanta, GA: Georgia Department ofRevenue, 2004).15. Jung, 76.16. Steven Gold, Property Tax Relief (Lexington, MA: Lexington Books, 1979), 207; Jung, 76.Zhao and Jung / Does Earmarked Revenue Provide Property Tax Relief55 burden to date. Examining the use of the LOST in South Carolina cities, Ulbrich foundthat the property tax growth rates in LOST-cities (1.5%) was less than one-tenth of thatin non-LOST-cities (17.9%) after comparing the revenue structures in LOST cities andnon-LOST cities.17 Moreover, she found that the per capita property tax between FY1991 and FY 1994 increased by about $7 in LOST-cities, compared with about $32 innon-LOST-cities. Thus, Ulbrich concluded that LOST has been ‘‘almost exclusively asubstitute for rather than an addition to the property tax’’ in South Carolina.18 Cautionshould be exercised in interpreting Ulbrich’s finding, however, because her study simplycompared property tax levels in LOST and non-LOST cities without controlling forother factors that may affect the property tax level. In addition, the study covered onlyseveral years, and thus it does not adequately capture the long-term effects of the earmarking. It should be noted that South Carolina authorized the use of LOST to its localgovernments in 1990 with the requirement that the property tax be rolled back by anamount equal to 63% of the LOST proceeds received in the first year. In addition, thispercentage increases in subsequent years.Employing a fixed effects model of a pooled time-series research design, Jung examined the effects of LOST on the level of property tax burden (per capita property tax andmillage rate) and total spending in 136 Georgia counties during a 13-year period (1984–1997).19 His findings show that counties with LOST on average have a $12 (or 1.8 mill)lower property tax than those counties without LOST. However, while an extra dollar ofLOST revenue provides about 28 cents in property tax relief, it also leads to about a 48cent increase in total spending. Thus, Jung concluded that although LOST achieved theobjectives of property tax relief, ‘‘the use of LOST is more of an augmentation of than aneffective substitute for property taxes in Georgia counties.’’20 Although Jung’s studyadvanced our understanding of the effects of earmarked LOST on the degree of propertytax relief in general, the study was rather limited in explaining how the degree of theproperty tax relief changes in the long-term in counties because the study did not employvariables to capture a long-term change in the study. However, this limitation can beovercome by employing a pooled interrupted time series research design in this study.The Wisconsin Taxpayers Alliance study provides additional insights into the budgetary effects of LOST. Analyzing fiscal data of 56 Wisconsin counties between 1986 and2001, the study compared the actual property tax changes to the property tax trend(calculated by the previous five-year patterns before LOST was adopted).21 The analysisshows that the property tax relief effect of LOST phases out over time. That is, the taxrelief was relatively high during the first few years, but it began to wane in the sixth andseventh years. On average, the study also found that only 28 cents per dollar of LOST17.18.19.20.21.56Ulbrich, 15–17.Ulbrich, 16.Jung, 83.Jung, 86.Wisconsin Taxpayers Alliance, 12–20.Public Budgeting & Finance / Winter 2008 have been used to reduce property taxes. Although the Wisconsin case is a more advanced longitudinal study in investigating the long-term effects of LOST on property taxrelief, using a longer time period in a study would help to better capture the changingeffects of earmarking behavior over a long-term period. As with the Wisconsin case, thebudgetary effects of LOST in Georgia may also vary over time. With a pooled interrupted time series research design, this paper investigates the long-term effects of LOSTon property tax relief in Georgia counties, which will improve our understanding aboutlong-term fiscal behavior of earmarked revenue sources.CONCEPTUAL FRAMEWORK AND RESEARCH HYPOTHESESThe LOST Act in Georgia mandates that local governments use LOST revenues to rollback property taxes during the second year and in ‘‘all subsequent years.’’ In so doing,LOST collecting counties are required to show: (1) the reduced millage rate resultingfrom the receipt of LOST revenue from the previous year, (2) the amount of property taxthat would have been levied without LOST (PTt à ), (3) the amount of LOST proceeds ofthe preceding year (LTt À 1), and (4) the resulting net property tax for the calendar year(PTt). Mathematically, the rollback can be illustrated in the following equation: PTt ¼PTt à À LTtÀ1 :Despite these requirements, the extent of the actual rollback becomes less certain afterthe second yearFbecause it is difficult to assess the amount of property tax that wouldhave been levied without LOST (PTt à ). Actually, there is not a binding limit on theproperty tax to be levied (PTt), given the amount of LOST proceeds of the precedingyear (LTt À 1). Thus, in fact, local governments might calculate the rollback in a reverseway. At first, the net property tax millage rate to be levied is assessed and is then addedwith the millage rollback provided by the LOST receipts of the preceding year, whichresults in the ‘‘would-be’’ millage rate without the LOST.22 In this case, the property taxrelief requirement has little influence on the actual budgetary decision, and the actualeffect would be less than the ‘‘paper rollback.’’23 Without a ‘‘binding’’ limitation on theproperty tax levy, local governments could receive more revenue from a combination ofsales tax and property tax than they would have received without LOST.Conceptually, in explaining the long-term effects of LOST (or other earmarked revenues) on property relief, a couple scenarios could be employed.24 Figures 1–3 depictthree commonly assumed scenarios regarding the effects of LOST on property tax relief.22. This is one of the authors’ interpretations from a personal interview with Chris Caldwell in theOffice of Budget and Finance, Athens-Clarke County, on December 22, 2003.23. Durning, 29.24. There are two assumptions underlying these scenarios. First, there is a stable trend of property taxgrowth if the LOST was not levied. Second, LOST proceeds in these counties remain unchanged over time.The first assumption is critical for the model, while the second one is for the purpose of simple interpretation.Zhao and Jung / Does Earmarked Revenue Provide Property Tax Relief57 FIGURE 1Scenario IFLOST is a substitute for property taxFor example, counties may use all LOST proceeds to provide property tax relief (Scenario IFFigure 1). In this case, the sum of property tax and LOST revenue after thecollection of the LOST should not exceed the projected property tax level, and thusLOST works as a full substitute for the property tax. Alternatively, the counties may useall LOST proceeds as additional revenues (Scenario IIFFigure 2), which results in noproperty tax relief at all. Under this scenario, the projected property tax level followsthe pre-LOST path, even after the collection of LOST. In many cases, however, LOSTproceeds may be used for both purposes (Scenario IIIFFigure 3) in the long term.Under this scenario, part of LOST revenue is used to provide property tax relief and partprovides an additional revenue source. Accordingly, the property tax level is loweredafter the collection of LOST, but the sum of property tax and LOST revenue after thecollection of LOST exceeds the projected property tax level of the pre-LOST period.However, it remains lower than the sum of property tax and LOST revenue than inScenario II. This scenario can be described by the concept of ‘‘fungibility.’’ Basically, ifthe earmarked revenue was not fully used for the earmarked function and part of it wereused for other purposes, the earmarked revenue is fungible. Literature on earmarkedrevenue sources reports this fungibility problem. For example, several studies investigating the effects of earmarked state lotteries for educational spending found that theFIGURE 2Scenario IIFLOST is an additional revenue source58Public Budgeting & Finance / Winter 2008 FIGURE 3Scenario IIIFLOST is a substitute for property tax and additional revenueearmarking did not increase the level of total education spending as states were likely todecrease their education appropriation in anticipation of lottery revenue.25In the case of Georgia, the LOST is basically adopted to provide property tax relief.However, if it did not result in a substantial property tax relief but more of an increase inthe total expenditure level, the LOST might have been fungible. In the case of Georgia,scenario III has received some support from Jung’s study.26 However, because Jung didnot employ variables to capture the long-term effects of the property tax relief in hisstudy, his finding presents an incomplete picture because it is highly possible that thebudgetary effects of the LOST on the property tax level would vary over time as seen inthe case of Wisconsin local governments.The above three scenarios (scenario I, II, and III) basically assumed a linear andparallel trend in the growth (or reduction) of the property tax level before and after theadoption of the sales tax. That is, there is no change in the slope of a pre-LOST propertytax trend, even in the post-LOST period. However, if the average growth rate of theproperty tax level will increase faster in the post-LOST period than in the pre-LOST era(a higher slope of property tax trend in the post-LOST period than in the pre-LOST era),this increasing growth rate of the property tax will eventually offset the short-term effectsof the property tax rollback and will lead to an even higher property tax level in the longterm. Since the LOST Act requires property tax relief starting in the second year ofLOST collection and continuing in subsequent years, an initial rollback is quite certain.However, because it is difficult to assess the amount of property tax that wouldhave been levied without LOST (PTt à ), counties may increase their property taxes overa number of years, and this increasing growth rate in the property tax during the25. Mary O. Borg and Paul M. Mason, ‘‘Earmarked Lottery Revenues: Positive Windfalls or ConcealedRedistribution Mechanisms?’’ Journal of Education Finance 52 (April 1989): 75–85; Mary O. Borg and PaulM. Mason, ‘‘The Budgetary Incidence of a Lottery to Support Education,’’ National Tax Journal 41, no. 1(March 1988): 75–86; John Mikesell and Kurt Zorn, ‘‘State Lotteries as Fiscal Savior or Fiscal Fraud: ALook at the Evidence,’’ Public Administration Review 46 (July/Aug. 1986): 311–320; Charles Spindler, ‘‘TheLottery and Education: Robbing Peter to Pay Paul?’’ Public Budgeting & Finance 15 (Fall 1995): 54–61.26. Jung, 83.Zhao and Jung / Does Earmarked Revenue Provide Property Tax Relief59 FIGURE 4Scenario IVFLOST serves as a source of fiscal illusionpost-LOST era may offset the short-term effects of the property tax rollback. Thus, itmay eventually result in an even higher property tax level in the long run. As illustratedin Figure 4, this pattern could be termed as a ‘‘fiscal illusion’’ scenario.A plausible explanation for this situation is the ‘‘fiscal illusion’’ theory which holdsthat taxpayers’ perceptions of the cost of government can be obscured by the ways inwhich government taxes are raised, and thus taxpayers are led to pay (accept) even highertax levels eventually.27 There are several sources pertaining to perception errors. First,more invisible taxes, such as LOST (or lottery revenue), may make it harder for averagetaxpayers to assess accurately the true cost of public service. Thus, the public may fail torealize that they are paying for a higher level of government spending after the adoptionof these less visible taxes. Second, income-elastic taxes (e.g., a sales tax) will automaticallyraise more revenues as income increases over time without increasing the tax rate, whicheventually results in more total government spending. Finally, the ‘‘paper rollback’’28 suchas is found in Georgia counties camouflages the nonbinding nature of the property tax reliefrequirement, and the illusion that property tax has been rolled back may, paradoxically,indulge governments to levy an even higher property tax in the long run.Given that the LOST Act mandates property tax relief with LOST revenues in thesecond year of LOST collection and subsequent years, it is almost sure that the adoptionof the LOST will bring property tax relief in the second year of LOST collection (shortterm relief ). However, because the LOST Act does not require binding limitation on theproperty tax levy or assessment, it will not bring a long-term property tax…