EFFECTS OF A ROOM TAX ON / J. PAUL COMBS AND BARRY W. ELLEDGE*

Introduction primarily on room occupants, compensat- NCREASINGLY, local governments ing local residents for some of the external Iare expressing concern about the costs costs of the resort activity. of providing public services because citi- There are several important questions zens are becoming more and more reluc- which must be addressed by any local tant to support leaders who increase the government considering the alternative of local property taxes, the primary source an occupancy tax on hotel/ rooms. of discretionary revenue for local govern- First of all, a local government should ments. This problem can be especially consider whether the industry pays severe in counties where a substantial its own way through general sales taxes, portion of the economic activity is gen- property taxes, privilege licenses, and erated by the resort industry, because the expanding employment in the local econ- resort industry generates a substantial omy. The measurement of costs imposed flow of tourists into and out of the resort by the traveling public, especially external area, causing additional costs to be costs, presents a major problem in deter- imposed on the local population. These mining whether the travel industry gen- costs include (1) increased expenditure for erates benefits sufficient to offset the costs local public services such as puhlic safety, to the community. medical services, water and sewer sys- If the travel industry does not generate tems, and road maintenance and (2) non- benefits sufficient to cover the cost monetary externalities such as time loss imposed on the community, a whole new and frusi;ration due to traffic congestions, set of questions concerning the impact of pollution, unpleasant esthetic effects and an occupancy tax should be addressed. To other factors contributing to a decrease determine the revenue potential of the tax, in the quality of life for local citizens. one would need to measure the elasticity To the extent that these costs are un- of demand for hotel/motel rooms. What compensated, local citizens who do not size tax could he imposed without signifi- participate in the resort activity have a cantly affecting the demand for ho- legitimate concern about the increasing tel/motel rooms and related goods and cost of local government and may legiti- services? mately feel that they are subsidizing the Clearly, according to Musgrave's classi- resort industry. fication, a hotel room occupancy tax is One way in which the local government a discriminatory tax since it fits the cate- in a resort may deal with this situation gory of a selective excise tax; but, so do is by imposing an occupancy tax on ho- many other taxes.' However, one of the tel/motel rooms. Administratively, a rallying points for hotel/motel associa- room occupancy tax should be relatively tions is the claim that an occupancy tax easy to impose and collect. The impact imposes an "unfair" burden on hotel of a room occupancy tax depends upon operators, both by singling out the travel the elasticity of demand for hotel/motel industry and by singling out as rooms in the resort. If demand is elastic, one component of the travel industry. A the tax will fall primarily on hotel/motel related, perhaps more important, question owners, causing them to bear a greater concerns the incidence of the tax: Does portion of the burden of local government. it impose a higher relative tax burden on If demand is inelastic, the tax will fall low-income people than on high-income people? In this paper, we address the questions 'Department of Economics, Appalachian State Uni- raised above in an attempt to provide versity. policy guidance for local officials.

201 202 NATIONAL TAX JOURNAL [Vol. XXXII

Revenue Potential of a Hotel/Motel inputs for the following reason. Room Occupancy Tax days are obviously not a homogeneous product; a vacation consisting of a luxury Two major issues involved in consider- hotel and expensive food and entertain- ation of an occupancy tax are: (1) how ment is qualitatively different from a much revenue will the tax generate for vacation of the same duration consisting local government and (2) what will he the of lodging in a cheaper hotel, eating in impact of the tax on the motel industry fast-food , and less entertain- specifically and on the travel industry in ment. We expect families to attempt to general? maintain the same quality of life on vaca- In order to address either of the two tion as they do at home, therefore reducing questions above, one must consider the the likelihood of substitution among the price elasticity of demand for motel ac- inputs in response to a tax on motel commodations. The rental of a motel room occupancy. can be considered as an input into the Another possibility for substitution production of some composite good, such exists within the lodging category. Given as, sales for a marketing representative, time to adjust to the price increase result- or a family vacation. ing from a motel occupancy tax, the vaca- Viewing motel accommodations as an tioner may choose alternative forms of input into a production process, one can lodging or lodging in a neighboring juris- apply Marshall's rules for the elasticity diction where an occupancy tax has not of demand for a factor of production.^ The been imposed. The former problem can demand for motel accommodations in a be handled hy imposing the tax on all resort will be more inelastic (1) the more temporary lodging facilities which are essential motel accommodations are for rented for less than some fixed period, the production of resort , (2) the such as 60 days. more inelastic the demand for resort vaca- The problem of interjurisdictional spill- tions, (3) the smaller the ratio of the cost overs cannot be handled quite so simply. of motel accommodations to the total cost Mikesell found a significant negative of resort vacations, and (4) the more in- impact of interjurisdictional sales tax dif- elastic the supply of other inputs for pro- ferentials on total sales and sales tax ducing resort vacations. Marshall's elas- revenue collected in the areas with higher ticity conditions pose a set of questions taxes.^ This implies that an occupancy tax which, ultimately, can only be answered may cause some vacationers to locate in empirically, hut some intuitive consider- neighboring jurisdictions in order to avoid ations of the implications of these condi- the tax. However, if the tax were to he tions may prove fruitful. imposed across the entire resort, the spill- The major input classifications for re- over problem would be minimized because sort vacations include the family's time, of the additiohal transportation cost in- transportation, lodging, food, recreation curred hy those attempting to avoid the and entertainment. With the possible ex- tax by driving back and forth between ception of recreation and entertainment, their lodging and the resort activity. some amount of each of these would he Resort vacations may be identified as required for a resort vacation of more than a luxury good, for which demand is likely one day in duration, and very little sub- to be elastic. Families may choose a less stitution between these inputs is likely expensive resort or take their vacations to occur. Transportation is generally by at home in response to rising prices. family car from point of origin to the resort The supply of other inputs for resort area, and as such is not subject to much vacations is expected to be highly elastic variation as long as the trip is taken. Food, to the decision-making family because the lodging and recreation expenditures can family unit is a price-taker in the markets be varied more than can transportation for food, transportation, and recreation. expenditures, but we expect that there However, the easy availability of these would be little substitution between these other inputs will have little effect on the No. 2] ROOM TAX 203 demand for lodging if, as suggested, they mately 1200 motel rooms and a population are not good substitutes for lodging. of 30,000, a four percent tax on an average The fourth of Marshall's elasticity rules room rate of $25 a day would yield about relates to the proportion of vacation cost $350,000/year, assuming 80 percent accounted for by lodging. Intuitively, it occupancy. This does not include the tax seems that lodging must be a relatively yield from other short-term lodging. Table small proportion of the total cost of resort 1 below shows other possible tax yields vacations. Our intuition is supported by from 1200 motel rooms under alternative the recent BLS expenditure survey which assumptions concerning the rate of the indicates that lodging is less than 20 tax and occupancy. The figures imply no percent of vacation spending.'' Thus, even reduction in occupancy as a result of the if the elasticity of demand for resort vaca- tax, i.e., either demand or supply is almost tions is (say) 2 with no substitutions perfectly inelastic. among the inputs, a five percent increase in the price of lodging will result in only Equity Effects of a Hotel/Motel a two percent decrease in the quantity of vacations and lodging demanded. Room Occupancy Tax Given all of the above considerations, The question of tax equity involves we expect the demand for lodging in a consideration of who benefits and who resort to he inelastic with respect to price. pays the tax. If the purpose of the ho- An interesting observable phenomenon tel/motel room occupancy tax is to supporting this conclusion is the growth compensate the local population for of cut-rate, minimum service motels along tourist-related costs, a comparison is interstates and in metropolitan areas, hut needed of the local tax revenues attri- not in . A logical conclusion is that butable to tourists versus the estimated the demand for resort lodging is inelastic.® local cost attributable to tourists. These If, as we suggest, the demand for resort costs would include both the direct (tangi- lodging is inelastic, a small ad valorem ble) costs of providing local government tax imposed on motel rooms and other services plus the indirect (intangible) costs forms of temporary lodging would have of the tourist-related externalities, such very little impact on the industry and as the congestion, etc. mentioned earlier. would generate substantial revenue for Although estimates such as these are the local government. For Watauga difficult, and perhaps impossible for the County, a small mountain resort county intangible costs, such an approach exam- in Western North Carolina with approxi- ines whether or not tourists are contrib-

TABLE 1

Revenue Potential From 1200 Hotel/Motel Rooms Using Alternative Tax Rates and Occupancy Rates with $25 per day average room rental cost. (Table entries are potential tax revenues in dollars.)

Room Tax as a Percent of Room Rental Charge

Occupancy 60 65,700 131,400 197,100 262,800 328,500 394,200 Rate in Percent 70 76,650 153,300 229,950 306,600 383,250 459,900

80 87,600 175,200 262,800 350,400 438,000 525,600

90 98,550 197,100 296,650 394,200 492,750 591,300 204 NATIONAL TAX JOURNAL [Vol. XXXII uting their share of local taxes. Davies and Black have used this approach Tourists already pay local taxes to the in examining the equity effects of includ- extent that they buy local goods and ser- ing the value of housing services in the vices and the businesses are able to shift sales tax base.' Progressive/Regressive their taxes forward. Such taxes may in- indices were calculated depending upon clude: (1) a share of the property taxes the movement of the average rate of taxa- paid hy tourist-related businesses, prorat- tion with changes in the ability to pay. ed by the share of business done by tran- Following traditional notions, a tax is sients, (2), a share of local excise taxes, considered regressive if the average rate and (3) user charges for puhlic services of taxation declines as the ability to pay used hy tourists. increases. A tax is progressive if the The tangible costs of providing services average rate increases as the ability to to travelers include a share of the follow- pay increases and the tax is proportional ing costs (again prorated to tourists): safe- if the rate remains constant with changes ty and law enforcement, roads and park- in the ability to pay. ing, water and sewer, solid waste disposal, Alternative assumptions of the income health facilities, fire protection, legal and (or ability-to-pay) measure are common court costs, parks and recreation, and in tax incidence theory. Annual income general government. This last item may is the generally accepted measure of the be thought of as "social overhead," i.e., ability-to-pay; however, income after the costs of maintaining a functioning taxes or total consumption expenditures government and a going community, both can also be used. Income before taxes of which are essential to the local tourist overstates "true" or disposable income; industry.^ therefore an alternative income measure Intangible costs attributable to travel- is also used in the following analysis. ers include: the costs of added traffic Lacking a measure of permanent income, congestion, the added pollution and de- total annual consumption expenditures spoilation of the local environment, and provide a proxy for the ability-to-pay. whatever value is attached to an erosion Consumption better reflects the ability- of the local native lifestyle. These intan- to-pay over a longer period of time than gible costs cannot be estimated in dollar does annual income before taxes. values; however, local elected government Consumer expenditures for vacation- policy makers can form normative judge- related lodging over 12 income classes are ments about the magnitude of these costs available from the BLS's Consumer when considering additional taxes on Expenditure Survey (1976). Data for 1972 transients. and 1973 are shown in Table 2. Another necessary assessment of any The elasticity of vacation-related lodg- tax is to examine the incidence of the tax. ing expenditures with respect to the given The two basic philosophies traditionally income measure is used to indicate the used in distributing the tax burden have incidence of an occupancy tax on ho- been: (1) the benefits received principle, tel/motel rooms. If the elasticity of lodg- and (2) the ability-to-pay principle. The ing expenditures with respect to income ability-to-pay principle has been used to (i.e. the percentage change in expendi- justify the majority of taxes (at least in tures divided by the percentage change terms of total tax receipts). For this paper, in income) is greater than one, then the we accept the conventional wisdom that percentage change in receipts raised from taxes be levied in accordance with ability- the tax is greater than the percentage to-pay. change in income; therefore a tax on the The room occupancy tax is a consump- room occupancy is progressive. If the con- tion-based tax. To measure the incidence verse is true, then the tax would be re- of a consumption-based tax, the consump- gressive. tion (use) of the item being taxed across Elasticity coefficients for lodging income classes needs to be examined. expenditures were estimated with the foi- No. 2] ROOM TAX 205

TABLE 2

Current Consumption Expenditures and Vacation-Related Lodging Expenditures by Income Class, 1972 and 1973.

Before-Tax Current Average Annual Income Class Consumption Vacation-Related Lodging Expenditures Expenditures 1972 1973 1972 1973

Under $3,000 3,040 3,367 $ 6.86 $ 7.29 3,000 to 3,999 3,939 4,231 9.09 9.22 4,000 to 4,999 4,588 4,681 10.42 20.47 5,000 to 5,999 5,134 5,159 13.66 14.07 6,000 to 6,999 5,590 5,559 15.24 13.10 7,000 to 7,999 6,095 6,195 18.73 17.67 8,000 to 9,999 6,833 6,828 26.80 23.55 10,000 to 11,999 7,643 7,798 32.55 32.20 12,000 to 14,999 8,905 8,765 40.55 42.61 15,000 to 19,999 10,245 10,383 66.21 53.33 20,000 to 24,999 12,056 12,542 110.78 75.20 25,000 and over 15,943 16,197 172.06 130.17

Source : Consumer Expenditure Survey Series: Interview Series, 1972 and 1973. U. S. Department of Labor Statistics, Report 455-3, U. S. Government Printing Office, 1976. lowing regression model, using total an- over the range of incomes reported. Con- nual consumption as the measure of in- versely, a coefficient of greater than one come. suggests such a tax would be progressive on average over the range of incomes E; = I + b log C; + Ej (1) reported. The regression results for 1972, 1973, where E; = the average annual vaca- and both years combined are presented tion-related lodging ex- below as equations 2, 3, and 4 respectively. penditure of the ith class CJ = the average annual total log E. = -6.57 + 2.09 log C. R^ = .97 expenditures for the ith (.09976) income class a = constant to be estimated (2) b = expenditure elasticity log E; = -5.59 + 1.82 log C. R^ = .95 coefficient to be estimated (.12656) and Ej = the random disturbance term. (3) Transforming the variable into log form log Ej = -6.09 + 1.96 log C. R^ = .96 changes the coefficient b into an elasticity (.08400) coefficient. An elasticity coefficient of less than one (4) suggests that a tax levied on the consump- Figures in parentheses are standard tion of lodging is regressive on average errors. 206 NATIONAL TAX JOURNAL [Vol. XXXII Tests of significance on the elasticity such as the big-city visitors bureaus that coefficients show the following: often are partially funded by part of the tax receipts. Frequently, governments act Both Null 1972 1973 years on this implied assumption of perfectly Hypothesis Data Data combined inelastic demand as well, assuming that (t- values) tax revenue will rise in the same propor- Elasticity tion as the tax. This assumption may or coefficient is may not be realistic. Ideally, a locality zero 20.95 14.38 23.33 Elasticity considering such a tax would need to coefHcient is derive estimates of the local elasticity of one 10.93 6.48 11.43 demand for hotel/motel rooms. In prac- tice, however, estimating the demand Row one shows that the test of the equation for hotel/motel rooms is diffi- routine hypothesis that the elasticity co- cult. Theory suggests that the demand for efficient is zero is rejected at the one hotel/motel rooms is a function of own percent level of significance. A more ap- price, price and availability of substitutes propriate null hypothesis, however, may (such as camping), income, and the total be that the elasticity coefficient is one, volume of travel. Attempts by the authors since we are interested in whether a room to empirically estimate the demand func- tax is likely to be progressive. The t- values tion have been unsuccessful due to lack in the second row indicates that the elas- of adequate data. We anticipate further ticity coefficients are significantly dif- work in this direction. ferent from one at the one-percent level of significance. Policy Implications The results generally conform to expectations: a room occupancy tax is Most counties are faced with a continu- likely to be progressive, with the extent ing problem of inadequate revenues rela- of progressivity depending on the income tive to expenditures for county services. concept.* Using consumption as a measure Although some relief in recent years has of income, the elasticity coefficient is ap- been provided by federal revenue sharing, proximately 2.00, or elastic. Although county governments are forced to increase crude, these estimates suggest that a tax existing tax rates or to look for new tax on hotel/motel occupancy would fall more sources. Counties have traditionally relied heavily on those with a higher ability to heavily on property tax receipts, but face pay, at least through the range of availa- increased resistance from property owners ble data. when property taxes are raised. This paper It should be noted that, within income has examined a hotel/motel room classes, there may be wide variations in occupancy tax as a partial substitute, spending patterns. Thus, there may be albeit small, to the property tax in a resort much variation in the use of and county. The presence of a large number motels within a given income class which of tourists in a resort is likely to be related affects the individual's tax burden. to substantial external costs such as con- Moreover, extending the conclusion of the gestion and pollution that may be un- likely progressivity of a room occupancy compensated. A hotel/motel room tax may tax to a particular locality requires the provide a means of forcing the tourist assumption that local hotel and motel use industry to pay the full costs relating to is distributed over income classes similar tourist activity. to the national data. Many states have already authorized The foregoing incidence analysis as- localities to levy a hotel/motel room sumes that the elasticity of demand for occupancy tax. In a resort community, a hotel/motel rooms is perfectly inelastic room occupancy tax may have special such that the tax would be passed com- appeal as a means of compensating local pletely along to the consumer. This is the taxpayers, (especially those who do not conventional view of proponents of the tax. participate in and benefit from the tourist No. 2] ROOM TAX 207 industry) for the costs imposed by the clude both a room occupancy tax and an tourist flow. Tax incidence analysis indi- amusement tax. cates that a room occupancy tax is likely A border city (county) problem may to be progressive, with the degree of pro- result if there is a business loss to sur- gressivity dependent upon the measure of rounding lower tax (or non tax) areas ability to pay used. In addition, the elas- because of geographic tax differentials. ticity of demand for hotel/motel rooms However, given the special attractions of is likely to be small when considering a resort area, and the unlikely availability accommodations costs as a component of of good substitutes, the border problem total vacation costs. Thus a room is likely to be small for tourist-related occupancy tax may raise a significant taxes. amount of revenue for the locality. Any locality considering a hotel/motel room tax must make its own estimates FOOTNOTES of the costs and benefits associated with 'See Musgrave (1959), p. 349. . A proper balance must be struck 2See Marshall (1920), pp. 383-386. between encouraging tourism as a desired ^Mikesell (1971). local industry and taxing that industry *U.S. Department of Labor Statistics, Consumer Expenditure Survey Series (1976). sufficient to cover all related costs. ^Note that these conclusions cannot be extended Normative judgements will have to be to . Available substitutes and reasons made relative to the external costs for travelling differ from that of vacationers. Also, imposed by tourists. If it is concluded that travel is likely to appear as a large recurring expense the tourist industry is not currently pay- to a business firm. «Netzer (1966), p. 169. ing its way locally, a hotel/motel room 'Davies and Black (1975). occupancy tax may be the most appropri- 'A similar regression was done using income before ate form of additional taxation. taxes as the measure of income. However, it was felt that consumption, as reported above, is the more It should be noted that hotels and motels appropriate measure. represent only one component of the tourism industry that could be taxed: other components include restaurants, service REFERENCES stations, gift shops, and tourist-related 1. David G. Davies and David E. Black. "Equity recreation and amusement firms. Howev- Effects of Including Housing Services in a Sales er, additional taxes on restaurants and Tax Base." National Tax Journal, Volume 28, service stations are likely to fall on many March 1975, pp. 135-139. 2. John F. Due. State and Local Sales Taxation. Public local residents, as well as tourists. Taxes Administration Service. 1971. on gift shops and tourist-related recrea- 3. Alfred Marshall. Principles of Economics. 8th Edi- tion and amusements are likely to fall tion. New York: Macmillan and Company, 1920. primarily on tourists. Some localities are 4. J. L. Mikesell. "Sales Taxation and the Border County Problem." Quarterly Review of Economics already levying an amusement tax. Many and Business. Spring 1971. of the same arguments for a hotel/motel 5. R. A. Musgrave. The Theory of Public Finance. tax could be made for an amusement tax: New York. McGraw Hill, 1959. viz., the elasticity of demand is likely to 6. Dick Netzer. Economics of the Property Tax. The be low when considered as one component Brookings Institute, Washington, D.C. 1966. 7. U.S. Department of Labor Statistics. Consumer of the costs of a vacation package; plus, Expenditure Survey Series: Interview Survey, 1972 the tax incidence is likely to be progres- and 1973. Report 455-3, U.S. Government Printing sive. An optimal tax strategy might in- Office, 1976.