no \L\\\%\3 :e\s+s aw//./5 8/-811 E CRS REPORT FOR CONGRESS _ Government Publications % A Unit Aug 17 1994 Washington univensmy Mmrarnes A NATIONAL LOTTERY: ‘ OVERVIEW AND ECONOMIC ANALYSIS The continuing debate over how'to close the Federal defi- cit has produced many different revenue raising proposals, one of which is a national lottery. This report contains an over- ‘view of the State.experience‘with.lotteries and an analysis of the economics of lotteries. , "m by Gregg A. Esenwein Specialist in Public Finance and Gary W. Shorter Analyst in Business & Government Relations Economics Division October 7, 1987 issouri um 3 lof Universi mm“ uTmm 1 tlii 010- 03939936 ml The Congressional Research Service works exclusively for the Congress, conducting research, analyzing legislation, and providing information at the request of committees, Mem- bers, and their staffs. The Service makes such research available, without parti- san bias, in many forms including studies, reports, compila- tions, digests, and background briefings. Upon request, CRS assists committees in analyzing legislative proposals and issues, and in assessing the possible effects of these proposals and their alternatives. The Service’s senior specialists and subjectanalysts are also available for personal consultations in their respective fields of expertise. CRS-iii CONTENTS BACKGROUND AND OVERVIEW OF STATE LOTTERY OPERATIONS . The Use of Lotteries . . . . . . . . . . . . . . Revenue Performance of State Lotteries . . . . . Lottery Products . . . . . . . . . . . . . . . . ECONOMIC ANALYSIS OF LOTTERIES . . . . . . . . . . . Tax Elements of Government Sponsored Lotteries . Incidence of Lottery Taxation . . . . . . . . . Revenue Potential of a National Lottery . . . . O O I O O O C O O C O O O O O I O O O O I APPENDIX: STATE-SPONSORED LOTTERIES . . . . . . . . ®-JO\U1 -I-‘bOf\)I\) A NATIONAL LOTTERY: OVERVIEW AND ECONOMIC ANALYSIS The continuing debate over methods to close the Federal budget deficit has produced interest in a national lottery as a revenue raising device. Given the seeming success of various State-sponsored lotteries, a national lottery would appear to be an attractive means of raising revenue for the Federal Government. This report examines the key issues associated with a national lottery. Specifically, the report contains an overview of the State experience with lotteries and provides an analysis of the economics of lotteries. In regard to the State experience with lotteries, the following obser- vations are made: Currently, 28 States (including the District of Columbia) have or will soon have operating lotteries. The revenue yield from State-sponsored lotteries is relatively small. On average, lottery revenue accounts for only about 2 percent of total State own-source revenue in States with lotteries. The revenue yield of State-sponsored lotteries has been some- what inconsistent over time. There is some question as to whether earmarking State lottery funds actually increases spending on earmarked programs. In regard to the economics of a national lottery, the following conclu- sions are drawn: I Initiation of a national lottery by the Federal Government would produce a net increase in consumer utility by introduc- ing a new consumer good into the marketplace. The income to the Government from operation of a lottery rep- resents a tax which from an economic perspective can be viewed as either an implicit excise tax on lottery purchases or as monopoly rent from operating the lottery. The tax element of a national lottery would be very regres- sive. Indeed, the tax elements of State-sponsored lotteries are in some cases twice as regressive as State sales taxes. CRS-2 Lottery revenue is essentially a zero-sum game. Revenue raised by a national game would come at the expense of State games and vice-versa. As a consequence, States might consider enactment of a national lottery as more of an encroachment than the introduction of a national sales tax. At best, assuming no State competition and participation rates equal to those of successful State lotteries, a national lot- tery might raise something less than $14 billion per year. If the States competed or national participation rates failed to match those of successful State operations, then the revenue yield of a national lottery would be significantly lower, probably only $2 to $3 billion. In fact, in a completely com- petitive market the Federal Government probably would be un- able to raise any significant revenue through a national lottery. Based on these conclusions, a national lottery may not be as attractive a revenue-raising device as would appear at first glance. The remainder of this report provides an overview and analysis of the issues associated with a national lottery. The first section contains an overview of the States‘ experiences, the second provides an analysis of the economic issues, and the final section briefly summarizes the findings of the report. BACKGROUND AND OVERVIEW OF LOTTERY OPERATIONS This section of the paper provides background information on lottery operations. Issues covered include a brief overview of the use of lotter- ies, lottery receipts and earmarking, and lottery products. The Use of Lotteries Lotteries have a long and varied history in America. Pre-Revolutionary lotteries were used to finance projects ranging from the construction of transportation facilities to campaigns in the French andglndiranpwars. Pri- vate lotteries, in which personal possessions were raffled, were used to liquidate personal debts. The popularity of lotteries grew through the mid-19th century. By 1820, the U.S. Congress had passed more than 70 acts authorizing lotteries for the funding of State and municipal bridges, wharves, water systems, ca- nals, roads, schools, and libraries. Private lottery operations also experienced significant growth over the same period. The growth in lottery use, however, was accompanied by an increasing number of swindles, misrepresentations, and other irregularities. 'I'he stig- ma produced by these improprieties combined with the fear that lotteries were undermining the American "work ethic" convinced the States to enact bans on the use of all forms of lotteries. By the latter part of the 19th century legal lotteries had ceased to exist in the United States. CRS-3 Legalized lotteries re-ernerged in 1964 when New Hampshire inaugurated a biannual sweepstakes. New York followed New Hampshire's example and insti- tuted a monthly lottery drawing in 1966. Both of these lotteries were re- garded as fiscal disappointments. Their major failing was lottery drawings that occurred too infrequently, a drawback which made it difficult to sus- tain consumer interest. In 1970, New Jersey became the third State to institute a lottery. Em- ploying the lessons learned from New Hampshire and New York, New Jersey in- creased the frequency of its drawings and lowered ticket prices. The reve- nue yield of the New Jersey lottery exceeded expectations. The results of the New Jersey experiment, coupled with recession and mandated tax caps, prompted other States to consider using lotteries as rev- enue-raising tools. Consequently, through the 1970s and into the 1980s many States came to view lotteries as both fiscally attractive and politically palatable. By 1980, 13 States and the District of Columbia were operating lotteries. By 1986, 23 States had operational lotteries while five more had voted to establish lotteries. Thus, 28 States either have or will soon have operational lotteries. (A complete list of the States that permit lotteries is contained in the Appendix.) The use of lotteries as revenue-raising devices is not limited to the United States. Currently, 64 countries, with a total of 90 functioning lot- teries, are members of the International Association of Government Lotteries (IAGL). The nations that have publicly run lotteries include Great Britain, Ireland, Finland, Israel, Denmark, Australia, Belgium, Chile, Bolivia, Ar- gentina, Costa Rica, Ghana, Cameroon, Japan, the Ivory Coast, Mexico, Sene- gal, Sweden, Zimbabwe, Trinidad, Turkey, and Tunisia. National lotteries are usually found in those nations that have centralized governments (such as France and Spain), while provincial or State run lotteries are usually the norm in federalized nations (Canada, West Germany, Switzerland, and the United States are examples). _1_/ Revenue Performance of State Lotteries According to the U.S. Census Bureau, the 18 operational State lotteries in FY85 (the latest year for which there is complete lottery fiscal data) grossed $8.1 billion and returned roughly $3.8 billion of this to the States as net proceeds. Although large in absolute terms, net proceeds from lot- tery operations constitute a relatively minor percentage of total State rev- enue. For example, in fiscal year 1984, no State raised over 5 percent of its total own-source revenue from lottery operations. Indeed, on average, State-sponsored lotteries accounted for only 1.95 percent of the own-source revenue of lottery States. ;/ 1/ An Overview of Lotteries worldwide. Public Gaming Magazine, May 1984. _2_/ Mikesell , John L. , and Zorn, Kurt C. State Lotteries as Fiscal Sa- vior or Fiscal Fraud: A Look at the Evidence. Public Administration Review, July/August 1986. p. 311 - 320. CRS-4 In addition, the revenue yield of State-sponsored lotteries has been inconsistent over time. Over the 1978 to 1985 period, thirteen of the sev- enteen States operating lotteries experienced reductions in their lottery revenue in at least one year. Moreover, the growth in revenues from State- sponsored lotteries tends to exhibit sizable fluctuations from year to year. The evidence has led some analysts to conclude that State-sponsored lotter- ies are not stable or reliable sources of revenue over time. 1/ Although, on average, the revenue yields from lotteries are relatively small , there are significant variations among the States. Net lottery reve- nue per capita is one way of comparing the performance of the different State lotteries. In FY85, the four highest net revenue per capita yields were earned by Massachusetts ($59), Maryland ($58), District of Columbia ($57), and New Jersey ($52). The four States with the lowest net per capita revenue were Arizona ($7), New Hampshire ($4), Maine ($3), and Vermont ($2). 3/ The top States in net revenue per capita have several characteristics in common. They are relatively urbanized and high income States, they are experienced at running lotteries, they have been aggressive marketers, and they have been innovative in terms of product diversification. In addition to different revenue yields, the States also differ in their use of lottery revenue. Of the 23 State-sponsored lotteries that were operational in FY86, 12 States directed their lottery revenues into either their General Fund or broad expenditure categories such as economic develop- ment and local revenue sharing. The remaining 11 States earmarked lottery funds for specific programs. Educational programs were the most common (7 States) form of earmarking. Some other States earmarked lottery revenue for transportation projects, aid to senior citizen programs, and conservation projects. §_/ Earmarking lottery revenue is viewed as one method of bolstering poli- tical support for the establishment of a State lottery. The issue of ear- marking is somewhat controversial. The debate centers on whether earmarked lottery revenue is actually used to increase the total amount spent on ear- marked programs. Since dollars are fungible, State officials can merely transfer a similar amount of funds into other areas. (l/ An empirical test of this issue yielded inconclusive results. Z_/ Q] Ibid. l;/ Carrol, Jane. The Lure of the Lottery. State Legislatures, April 1986. p. 26 -30. 3/ Ibid. §_/ Kaplan, M. Roy. The Social and Economic Impact of State Lotteries. In Frey, James, and William Eadington, eds. Gambling: Views from the Social Sciences. The Annals, July 1984. p. 99. 1/ Mikesell and Zorn, State Lotteries as Fiscal Savior. CRS-5 Lottery Products Since their re-introduction in the mid-60s, State-sponsored lotteries have continued to evolve to better exploit the marketplace and maximize sales. The evolution has taken the form of growth in the number and geo- graphic diffusion of retail outlets for lottery sales, increases in the frequency of drawings, and product innovation and diversification. Product innovation and diversification have been of particular importance, allowing the games to reach as many demographic groups as possible and maintaining product novelty. The three major lottery products that are presently of- fered include instant games, numbers games, and lottos. Instant games, which usually cost $1, are based on games such as tic- tac-toe or poker and let the player know immediately whether the player has won. Of the three basic products, the prize winnings from this game are generally the smallest, but the odds of winning are generally the highest. In FY85, instant games accounted for approximately 14 percent of total State lottery sales. The market share for instant games has remained fairly sta- ble since the early 1980s. 8/ Numbers games involve selection of a three or four digit number that is matched against winning numbers drawn daily, weekly, or biweekly. Numbers games first appeared in the mid-1970s and were, in part, promoted as a way of curtailing the illegal numbers games. Studies show, however, that they have had little or no effect on the illegal numbers games. 9_/ Since the early 1980s, State-sponsored numbers games have lost consider- able market share. In FY81, they accounted for roughly 71 percent of total State lottery sales, but by FY85 their share had fallen to 45 percent. _lQ/ The drop in market shares appears to be due to a shift in demand from numbers games to lotto games. Lottos are a variation on an old Italian lottery game in which players . choose a group of numbers from a relatively large field, for example, six numbers from a field of 50. If the winning combination of numbers is not chosen, then the prize is rolled over into the next game and the jackpot grows. Lotto jackpots are the largest of the three lottery games, often reaching several million dollars. Conversely, the odds of winning are the smallest of all three lottery games. Lotto is the major lottery growth product of the 1980s. In FY81, the game accounted for roughly 3 percent of total State lottery sales but, by FY85 its market share was 40 percent. _l_1/ In order to compete with the more populous States which can offer larger lotto jackpots, smaller States are §/ Lottery Fact Book. Gaming and Wagering Business Magazine, February 1987. p. 33. 2/ Abt, Vicki, James Smith and Eugene Martin. The Business of Risk. Lawrence, Kansas, University Press of Kansas, 1983. p. 67-68. 1Q/ Lottery Fact Book, p. 33. 1/ Ibid. CRS-6 forming multi-State lotto games. Vermont, New Hampshire, and Maine estab- lished a multi-State lotto in 1985. ECONOMIC ANALYSIS OF LOTTERIES This section of the report provides an economic overview of a national lottery. Issues examined include the tax elements of government-sponsored lotteries, the incidence of lottery tax payments, and the revenue potential of a national lottery. Tax Elements of Government-Sponsored Lotteries Gambling is not an investment or (for most participants) a money-making activity. It is consumption. The good consumed is participation in the game while the monetary losses associated with gambling merely represent the cost of consumption. In other words, gambling has an entertainment value similar to movies, theater, or sporting events. 12/ By legalizing a lottery, the government provides a "new" consumer good, an action which, in theory, increases consumer welfare. At the same time, however, the government, in essence, assesses a tax (the difference between the amount collected and the amount paid out in winnings, which is called the "take-out" rate) on the good. As is the case with most taxes, the tax, taken alone, results in -an economic efficiency or consumer welfare loss. Overall, the welfare gain stemming from the introduction of the consumer good will not be fully offset by the welfare loss produced by the tax and there will be a net increase in consumer welfare. l§/ The ad-valorem excise tax implicit in a lottery is simply the net reve- nue portion of the purchase price as a percentage of the returned winnings and administrative costs. For example, if 50 percent of the purchase price of a lottery ticket is returned as winnings, 5 percent goes to administra- tive costs, and 45 percent is retained as government revenue, then the im- plicit excise tax on the purchase of the lottery ticket is 82 percent (45 percent profits divided by 50 percent winnings plus 5 percent administrative costs). From an economic perspective this is equivalent to the States allowing a private lottery to operate at cost (including a normal rate of return on the investment) and applying an explicit ad valorem excise tax on each pur- chase. In fact, the implicit excise taxes associated with the various State run lotteries have been estimated to range anywhere from 60 to over 90 per- cent. 14/ 2/ Suits, Daniel B. Economic Background for Gambling Policy. Journal of Social Issues, v. 35, no. 3, 1979. p. 43-61. _1§/ Brinner, Roger E., and Clotfelter, Charles T. An Economic Appraisal of State Lotteries. National Tax Journal, December 1975. p. 395-405. _1_‘;/ See for example, Brinner and Clotfelter or Heavey, Jerome F. The Incidence of State Lotteries. Public Finance Quarterly, October 1978. CRS-7 Take-out rates vary dramatically from one organized gambling activity to another. In general , casino gambling has the lowest take-out rates. For example, in roulette the take-out rate is around 5 percent: the house pays back about $.95 of every dollar wagered. Casino slot machines have a take- out rate between 5 and 25 percent. Horse racing has take-out rates of 15 to 20 percent, paying back about $0.80 to $0.85 of every dollar wagered. l_§/ Lotteries have the highest take-out rates of any form of organized gam- bling. Take-out rates range from 50 to 60 percent in most State-sponsored lotteries. That means for every dollar wagered only around $0.40 to $0.50 is returned in winnings. l§/ Incidence of Lottery Tax Payments Several observations concerning the consumption of gambling across in- come classes can be drawn from the available empirical evidence. First, the percentage of people who gamble tends to increase as incomes increase. Sec- ond, as is the case with the consumption of other consumer goods, average outlays for consumption of gambling increase as income levels increase. The empirical evidence also shows, however, that outlays for gambling, especial- ly outlays for lottery tickets, do not rise as fast as income and so they decline as a percentage ofincome as income increases. _l_7_/ Because the per- centage of income spent on lottery tickets declines as income increases, State-sponsored lotteries as revenue-raising devices place heavier tax bur- dens on lower-income households than they do on upper-income households and, hence, are regressive. l§/ _l_5_/ State Lotteries: A Legal Sucker Bet. Consumer Reports, February 1974. p. 177-179. 1_6/ Wat son, Ronald D. Lotteries: Can the Public and State Both Win? Federal Reserve Bank of Philadelphia, July 1973. p. 3-15. _1_]_/ The only exception to this general consumption pattern is outlays for casino gambling. In this instance, outlays increase as a percentage of income as income increases. This is probably due to the fact that the availa- bility of casino gambling is limited and hence, to a certain extent is availa- ble only to relatively higher-income households who can afford the cost of travel and accommodations. _l_8_/ The regressivity of lottery taxation has been confirmed in numerous theoretical and empirical studies. It should be noted that these studies ad- dress the distribution of tax payments not the distribution of the lottery tax burden. The distribution of tax payments entails determining who actually makes the tax payments while the distribution of the tax burden involves de- termining all the income changes (both on the uses and sources side) that would result from imposition of the tax. For example see; Brinner and Clot- felter; Suits, Daniel B. Gambling Taxes: Regressivityand Revenue Potential. National Tax Journal, March 1977; Spiro, Michael H. On the Tax Incidence of the Pennsylvania Lottery. National Tax Journal, March 1974; and, Clotfelter, Charles T. On the Regressivity of State-Operated "Numbers" Games. National Tax Journal, December 1979. CRS-8 Revenue Potential of a National Lottery Estimation of the possible revenue from a national lottery is diffi- cult. There are no data available on the possible national level of demand for lottery consumption. Moreover, it is not possible to project revenue accurately without first specifying the design elements of a national lot- tery. In addition to these problems, the revenue potential of a national lot- tery depends to a great extent on the response of the States. while it is possible that introduction of a national lottery might produce a slight in- crease in overall demand for the product, the revenue potential of lotteries is essentially a zero-sum game. That is, most of the revenue raised by a national game would probably come at the expense of State-sponsored games, while revenue raised by State-sponsored games would reduce that available to the national game. l2/ The contrast between the zero-sum nature of revenue from lotteries and the nature of revenue from sales taxes is worth noting. If the Federal Gov- ernment enacted a national sales tax, State sales tax revenues would proba- bly decline because of a contraction in overall consumption. with this ex- ception, however, revenue raised by the Federal sales tax would not come at the expense of State sales tax revenue. The two taxes could exist side-by- side with neither encroaching on the other‘ s revenue pool. In contrast, revenue raised by a national lottery would come at the expense of revenue raised by State lotteries and vice-versa. As a consequence, States might consider enactment of a national lottery as more of an encroachment than the introduction of a national sales tax. Faced with introduction of a national game, the States could respond in one of three ways. They could drop out of the lottery market, make arrange- ments (either explicit or implicit) with the Federal Government to share the market, or compete with the national game for market shares. Obviously, a national game would earn the most revenue if the States dropped out of the market and would earn the least amount of revenue if there were competition. Revenue under the market-sharing arrangement would fall somewhere between these two points. As a rough approximation of the potential revenue from a national lot- tery, upper and lower revenue limits are estimated. The upper limit assumes that States with lottery operations drop out of the market, while the lower limit assumes that the States compete with the national game. State data on lottery proceeds per capita are used as proxies for a national lottery. Data for a successful State lottery are used to project the upper revenue limit, while averages for State lotto operations are used to produce the lower revenue limit. Q/ If the introduction of a national lottery produced a large enough shift in the demand for lottery consumption, then the revenue raised by the national game would not necessarily all come at the expense of State-sponsored games. Once the new level of demand is reached, however, the situation would again revert back to a zero-sum game at a higher level of total revenue. CRS-9 For the estimate of the upper revenue limit, the Maryland State opera- tion is used as a proxy for the national game. The Maryland game was chosen because of its relative experience in operating a lottery (it was introduced in 1973), its innovative marketing, and because of the fact that it consis- tently has one of the highest net proceeds per capita, $58, of any State-run lottery. _2_9_/ Extrapolating this level of net proceeds per capita to the United States as a whole indicates that the potential net revenue from a na- tional lottery might be as high as $14 billion (population of 241 million multiplied by $58 per capita net proceeds). This estimate assumes that the national lottery would have the same product mix as the Maryland lottery (instant lottery, daily numbers game, and weekly lotto), that it would have the same take-out rates as the Mary- land games (on average, approximately 44 percent), and that participation rates in the United States as a whole would mirror those in Maryland.g_1_/ Finally, it assumes that the States drop out of the market. Obviously, the nature of these assumptions is such that the $14 billion estimate should be considered as only a very rough guide to the upper reve- nue bound of a national lottery. There are several reasons why this esti- mate probably constitutes a substantial overstatement of the revenue poten- tial of a national lottery. For one, it is doubtful that a national lottery would consist of the same product mix as the Maryland lottery. Some types of games are more suitable for large scale operations than others. The Federal Government, for example, might be able to run a more profitable lotto game than a State, but might encounter greater costs in trying to run a daily numbers game. Hence, its revenue maximizing mix of games may well prove to be different than the mix of Maryland State games. Another reason the revenue estimate is probably overstated is that it assumes average participation rates in the Nation as a whole will mirror those found in Maryland. Maryland is on the east coast and is essentially an urbanized State with higher than average per capita income. It is doubt- ful that participation rates in mid-western States with large rural popula- tions or southern States with below average per capita incomes would match those of Maryland. In addition, the per capita net proceeds data for Mary- land represents participation not only by Maryland residents but also parti- cipation by residents of other States. Use of these data for estimating na- tional revenue, therefore, produces some double counting and results in an overstatement of revenue potential. Finally, the upper revenue limit assumes that the States would drop out of the lottery market in the presence of a national game. A more likely 22/ Carrol, The Lure of the Lottery. The only other State with higher net proceeds per capita is Massachusetts at $59. 21/ It is not clear that this take-out rate represents the profit maxi- mizing rate for lottery taxation. If it is not the profit maximizing rate, then the national lottery could raise more revenue than estimated here by ad- justing the take-out rate to maximize profits. CRS-10 scenario is that the States would compete for market shares with the national game. They could compete by either lowering the price (take-out rate) of their games or by banding together to produce regional lotteries with larger prizes. The potential effects of this competition on revenue are quite signifi- cant. A government that runs a lottery has in essence, set itself up as the sole supplier of a good by legislatively restricting other suppliers from entering the market. This allows the government to exert monopolistic con- trol over the supply of the product. As would be the case in any similar market situation, the government is able to charge a price for the good that is higher than it could charge in a competitive market. The difference be- tween the price set by the government and that which would exist in a compe- titive environment is monopoly rent and accrues to the government in the form of tax revenue. gg/ The introduction of competition into the legalized lottery market could significantly decrease this monopoly rent and, hence, dramatically reduce the expected revenue to both State-sponsored and national games. At the ex- treme, competition could eliminate all monopoly rent and, hence, eliminate any significant lottery revenue to both the State and Federal governments. The estimate of the lower revenue limit of a national lottery attempts to address some of the difficulties associated with the upper revenue esti- mate. First, it assumes that the national game would be a lotto type game rather than a mix of games such as those found in Maryland and that the take- out rate would be 50 percent. g_3_/ Second, it assumes that gross sales for the national game would match average per capita gross sales of States that currently operate lottos. Finally, it assumes that because of competition, the national game would not derive any revenue in those States that currently operate lottos. In other words, the national game will only be able to raise revenue in States that do not operate lotto games. The average per capita gross sales of lotto tickets for all State lot- tos in fiscal year 1985 was $39.33. The population of those States which did not have operational lottos in FY85 was 149 million. Hence, under these assumptions, the gross sales of a national lotto game wouldbe $5.6 _2__2_/ For a discussion of the monopoly rent approach to public lotteries see: Aronson, Richard J., Andrew Weintraub, and Cornelius Walsh. Revenue Po- tential of State and Local Public Lotteries. Growth and Change, April 1972. 2_3_/ It is probable that a national lottery would consist of some type of lotto game. A national lotto game would have certain economies of scale ad- vantages over its State-sponsored counterparts. That is, given the size of the participation pool it could draw on, a national lotto could over substan- tially larger prizes than State-sponsored games. There is some evidence to suggest that the size of the prize is a major determinate of demand for lotto consumption. Thus, a national lotto might have a competitive advantage over State-sponsored games. CRS-ll billion. Assuming a take-out rate of 50 percent, Federal revenue would be approximately $2.9 billion. gg/ These two estimates represent the upper and lower revenue limits of a national lottery. Rather than dropping out of the market or competing, the States could decide to share lottery revenue with the Federal Government. In this case, the revenue for the Federal game would fall somewhere between these two limits. Because the national game could not be priced differently in different States, under this scenario all States and the Federal Govern- ment would have to agree on a specific pricing strategy. As is often the case with cartels, however, this agreement could be fragile and difficult to maintain. If any State decided to drop out of the arrangement, then compe- tition would re-emerge with resultant reductions in revenue. CONCLUSIONS A national lottery would introduce a "new" consumer good which would, in theory, increase consumer welfare. The good, however, is used dispropor- tionally by lower-income households. The government, would exert monopoly control over the good and impose an extremely high rate of tax on its use. It would be a very regressive tax. To be successful, the Federal Government would have to advertise ag- gressively and market its lottery products, putting it in the somewhat awk- ward position of actively promoting behavior that some find morally objec- tionable. On the other hand, in the current environment, a national lottery might be viewed as more politically palatable than other forms of tax in- creases. While a national lottery would raise some revenue, the amount of revenue raised would depend on game structure, participation rates, and the reaction of the States. Revenue raised by a national lottery would primarily come at the expense of the revenue raised by State lotteries. If the States responded ” to the introduction of a national lottery by competing, the revenue yields of both the Federal and State games would be significantly reduced. Q/ It should be noted that if the introduction of a national lottery breaks down anti-lottery resistance in those States currently without lot- teries then even more States might enter into competition with the national game. If all States had lotteries and competed against the national game, X then the Federal Government would be unable to raise any significant revenue from a national game. gae CRS-12 APPENDIX STATE~SPONSORED LOTTERIES States with Lotteries Arizona California Colorado Connecticut District of Columbia Delaware Illinois Iowa Maine Maryland Massachusetts Michigan Missouri New Hampshire New Jersey New York Ohio Oregon Pennsylvania Rhode Island Vermont Washington West Virginia State Lotteries Approved in Nov. 1986 Voter Referendums Florida Idaho Kansas Montana South Dakota LIBRARY % V or % WASHINGTON umvsnsrrv ST». Louis - Mo‘.