Maximizing Social Welfare through Variations in State Lottery Redistribution Methods
Janine L’Heureux[1]
The George Washington University
Economics Department
Abstract
This study finds that consumers are indeed aware of the redistribution methods employed by state lotteries, and furthermore, react negatively to the operation of an earmarking program compared to a General Fund program, when controlling for relevant differences between the states during the period 2000-2010. Historically, state lotteries developed as sources of entertainment and as a sort of civic responsibility. After growing concerns over the transparency and the welfare maximizing ability of lotteries, they were abolished until 1963, when again, state lotteries have moved towards private sector techniques to achieve maximum revenue and social welfare. Due to problems of fungibility, earmarking does not, in reality provide the same boost in proceeds as typically advertised. Even with the high levels of emphasis on the benefits of earmarking to education, this empirical analysis finds that General Fund states receive higher revenues per capita, suggesting a movement away from the misleading marketing of additional benefits of earmarking.
I. Introduction
The Ohio Education Lottery’s motto is “Take a chance on education. Odds are, you’ll have fun!” This, like the lottery motto of many other states, specifically appeals to the jocular, socially conscious lottery player. But is targeting the altruism of risk-seeking lottery players an effective method of maximizing revenue and welfare across states? Do players react to the responsible administration of state lotteries by boosting their expenditures on tickets, and has this mechanism changed over time? Unlocking the existence of a positive feedback from well-administered state lotteries should be able to compound revenue and welfare objectives. The total expenditure on state lottery tickets in 2010 across all states that operated public lotteries was $53.1 billion or the equivalent of $171 spent per person. This massive volume reflects a willingness to participate, and the potential to align individual and state incentives for mutually beneficial results.
Originally created as sources of additional general financing to supplement the state’s budget, lotteries have evolved to take into account the inherent negative consumption externalities, generally through one of multiple choices of redistribution methods. States have traditionally advertised the distribution of revenues to either a General Fund or earmarked programs. State General Funds are the main revenue depositories which contribute to a basket of beneficiary programs, typically eduation, public health, and other state services according to statute. Ideally, if there is any degree of positive feedback from earmarked programs, that is, that lottery consumers respond to the culturally-beneficial idea of government revenue directly benefitting education over other programs, then education earmarking may be the best-practice recommendation for states to maximize revenue. However, this study finds that earmarking tends to correspond with lower levels of revenue per capita compared to using General Funds when controlling for administrative and demographic factors facross states, during the period 2000-2010. Those states that have decided to turn to education earmarking to attract the more altruistic lottery players may not in reality receive higher overall revenues than those using General Fund systems.
II. Literature Review
State lotteries join public provision with a private market activity, which introduces competing objectives of revenue maximization and social welfare optimization. Although Federal law regulates certain aspects of the inter-state lottery system, states are largely able to individually craft their framework for administrative organization, private consultation, and redistribution algorithms based on their unique demographics and market demand. The rapid rate of state lottery implementation since the 1960’s has provided a large cross section of multiple variations on the same theme of a quasi-public system. This also indicates that state have attempted to implement variations in administrative methodologies as evidenced by the incidence of system rebrandings and reforms over time. States provide manageable population delineations within the wider United States and thus relatively comparable units with similar objectives but differing constraints. This, specifically, has attracted many economists, political scientists and social scientists alike, to prepare frameworks for achieving both revenue and welfare maximization, which are not always mutually reinforcing.
As a public provider supplying private sector goods, the operational framework of state lotteries has come under severe pressure to increase efficiency. Brignall and Modell (2000) discuss the emergence of “New Public Sector” (NPS) or the increased emphasis on private sector management methods to public sector programs. Individual states have attempted an assortment of private sector-inspired initiatives, aimed at increasing revenue and decreasing administrative costs, or alternatively attempting to more accurately target programs in need. Certain states have experimented with optimizing marketing activities, changing the variety of ticket offerings, or contracting with private enterprises. A common approach to achieving improved social welfare levels lays in the designation of an earmarked versus General Fund system. Leveraging the altruism of lottery players to benefit local schools could perhaps contribute to the two objectives of maximizing both revenue and welfare together, but there is significant debate over the reality of this mechanism. Past studies concerning the efficiency of lottery systems has generally been divided into those evaluating the administrative operating strategies, and those analyzing the redistribution of lottery funds. This section follows this division.
Landry and Price (2007) attempt to isolate the effect of gambling for pure entertainment purposes, from gambling that takes into account the benefits accrued to the state through public good provision. They consider the following question: Do states that earmark for education direct higher percentages of revenues towards public provisions over those that use a General Fund framework, for states that operated lotteries from 1990-2000? The authors contend that during this time period, controlling for casino and multistate lottery participation and other state characteristics through a fixed-effects model, those that earmark tend to prompt higher lottery expenditures. However, during the period 2000-2010, or the proceeding decade, casinos and multi-state lotteries became commonplace so these factors become less powerful in explaining differences across states in lottery revenue per person. The following analysis applies a modified quantitative method and introduces an interaction term to a more recent time period to determine if the proceeds to education earmarking persist, or rather decline, over the following decade. A changing economic environment and ample reaction time to programs initiated by state administrations may provide explanations for altered behaviors. Landry and Price’s work provides consideration for the necessity of teasing out the effects of the gambling opportunity versus the utility derived from voluntarily contributing to government funds.
Morgan (2000) attempts to distinguish the portion of individual motivation that can be attributed to pure entertainment value, versus that which may be driven by cultural or altruistic impulses. The difficulty in synthesizing a state lottery system that allows for a just redistribution of funds lies in determining whether consumers are willing to increase their expenditures in response to the structure of the administration. Determining the general balance between economic and social motivations that lottery players consider could provide the basis for a compounding mechanism that both allows the derivation of high entertainment value while providing an additional outlet for charitable giving to engage in local community building. To this end, Morgan attempts to explain the amusement factor through comparing the payout rate between state lotteries and casino-run bingo and keno games that lack the cultural redistributive aspect in favor of a for-profit model. These payout rates were 50% and 80% respectively, where revenues from state lotteries was around $53 billion in 2010 and casino revenues were only around $57 million. This shows that more people play games with a lower payout rate, possibly indicating some consideration of social contribution.
Morgan (2000) acknowledges the vast number of studies contributing to the conclusion of positive response rates to education earmarking, and maintains that there is at least evidence of a link between public goods provision and ticket sales. Morgan also examines the total revenues received by states with General Funds versus those with earmarks during the period 1990-1995, as pictured in Figure 1 below.
Figure 1
Here, Morgan plots the total revenues per capita during this period, seemingly showing the declining difference in the revenues per capita between the two methods. However, there were several more earmarking states at the time, artificially indicating a faster convergence in the per capita revenues (expenditures in the graph). By comparing this total per capita revenue graph during 1990-1995 to the average revenue per capita during 2000-2010, it is possible to infer that there was likely an overtaking in primacy by the General Fund, although Morgan did not make his data available, so we cannot recreate the averages for the years 1990-1995 nor the in between years 1995-2000. These are, though, likely to be almost the same states, as Morgan used 36 states while the produced graph on the right uses 37. Using the data that will be explained in Section VIII: Quantitative Analysis, it is possible to conclude that as of 2010, there was still a significant differential in these per capita revenues, and furthermore that earmarked lotteries earned lower average revenues per capita than lotteries using a General Fund.
This analysis will focus on the time period from 2000-2010 for several reasons. The most complete data on state finances available by the U.S. Bureau of the Census spans from 2000-2010, but more importantly, the year 2000 marked a definitive shift in membership in multistate lottery games. According to Matheson and Grote (2011), prior to 2000, smaller states with lower per capita expenditures banded together to provide more attractive jackpots, while after 2000, even the states with larger balance sheets decided to become members. Although this study does not look into the effects of this phenomenon, it will take advantage of this commonality across states.
Most academics have pinpointed the disconnect between the marketed benefits to education yet the shortfall in revenues per capita, as a problem of fungibility. Consequently, there is a rich body of existing literature investigating, critiquing and attempting to remedy this fungibility problem. As it stands, after several years of operation, states are able to project lottery revenues, leading some states to shift budgetary education allocations to other programs under the assumption that incoming lottery revenue will fill the education budget gap. Spindler (1995) specifically conducts an analysis of the fungibility of lottery revenue through the change in education expenditures during the first five years after implementation of an education earmark in seven states. He finds evidence of a trend of increasing education expenditures during the first few years of earmarking, followed by an abrupt and large decline, rendering overall negative gains for education. He attributes this phenomenon to political budgetary manipulation. He notes a trend in annual budget alterations, which would support the hypothesis that even in earmarked states, additional lottery revenues may not technically go entirely to education. This temporal fluctuation in benefits has become the basis for continuing the work that Landry and Price (2007) performed to incorporate the most recent decade with available lottery reports.
Pantuosco, Seyfried and Stonebraker (2007) compare the eventual distribution of funds for education in states with K-12 education earmarked versus those with a General Fund.
“We find that lottery revenue rarely has a significant impact on K-12 spending in our sample of states that earmark the revenue for that purpose. Yet in our sample of states that do not earmark, lottery revenue typically does increase the growth of K-12 spending… A highly visible education lottery might convince ill-informed voters that education priorities are being addressed, irrespective of how actual monies flow.” [emphasis added]
Thus, states that utilize a General Fund may actually contribute more substantially to education expenditures than those legislatures that promote solely education earmarked programs.
Many academics have analyzed the eventual recipients of lottery funds, and the effectiveness of maximizing social welfare for both General Fund and earmarking programs. The process of redistribution may be a one-way route of fund redistribution away from the consumers towards a different set of beneficiaries. “Feehan and Forrest (2007) and Stranahan and Borg (2004) find that wealthy individuals and regions tend to benefit disproportionately from money earmarked towards cultural program and education, potentially exacerbating the regressivity of the revenue side of lotteries,” according to Grote and Matheson (2011). Thus, those individuals most likely to regularly play the lottery do not always see the redistributed benefits. Also, Borg and Mason (1988) find that the actual lottery player is unlikely to be the beneficiary from redistribution efforts. In operating education-earmarked lotteries, states are attempting to capitalize on charitable players, with little transparency. However, if players are not actively witnessing the redistributed education benefits, the feedback mechanism motivating them to increase expenditures for the benefit of the community may weaken. The trends of regressivity in ticket consumption paired with the higher returns to wealthier communities compounds this effect. This invites the question of whether lottery participants are responding to an incorrect interpretation that state lotteries increase the education expenditures in their own lower-income school districts.
It is pertinent to mention that not all states that earmark for education do so uniformly across or even including all levels of education. This study will not attempt to investigate the state methods of determining the recipients of lottery distribution, however it will consider the expenditures by players in reaction to their perception of the lottery. If, indeed, lottery consumers are reacting to marketed benefits to education for altruistic reasons, this additional expenditure may be in vain and not yield the expected returns to the immediate community. This would also indicate that state lottery administrators might be able to boost their revenues in the future by focusing redistribution high lottery expenditure areas in tandem with increasing transparency and reliability.
This study primarily extends the reasoning of Landry and Price (2007) to evaluate whether across lottery states, consumers continue to respond positively to redistributive benefits from earmarking in the most recent decade. States have increasingly revised their tactics to target the objectives of maximizing revenue simultaneously with welfare, and the relatively even split in the number of states with earmarked versus General Fund programs invites skepticism over the primacy of one over the other. During the eleven-year period from 2000-2010, the implementation of five additional state lotteries that earmarked for education reflects the increasing belief of this as the dominant method. If states that have run education-earmarked programs are not enjoying significant gains in capita lottery revenue, then it may be possible to conclude that over time, education earmarking is not an advisable method of achieving both the objectives of revenue and welfare maximization simultaneously.