The Demand for Lottery Products and Their Distributional Consequences

Abstract

This study examines the distributional impact of three types of lottery games operated by the South Carolina Education Lottery (SCEL). We fi nd signifi cant sales variation by game type across both age and race. We also fi nd each of the three types to be regressive, but with substantial differences in the degree of regressivity across games. By estimating the determinants of lottery sales using variables that capture the distribution of income rather than simply its level, our analysis provides a more complete description of the incidence of lottery sales. Our results suggest that lotteries may not be as regressive as suggested by the earlier literature.

The Effects of Lotto Game Changes and Large Jackpots on Income Elasticities and Sales

“…Other studies of the income elasticity of lotteries include Combs and Spry (), Ghent and Grant (), Garrett (), Garrett and Coughlin (), Garrett and Marsh (), Hansen (), Hansen, Miyazaki, and Sprott (), Price and Novak ().…”

Modelling regional lottery sales: Methodological issues and a case study from Spain

“…In terms of public policy and welfare, there has been debate about whether high jackpots tend to increase or decrease the extent to which the tax implicit in lotto games is regressive. Employing spatial data, Ghent andGrant (2010), andCombs et al (2008) compare different games offered by a lottery agency and report a tendency for games with higher jackpots to be less regressive. Like us, Oster (2004), examines differences according to jackpot size in a single game and finds again that high jackpot games deliver lower regressivity.…”

“…With notable exceptions, such as Garrett and Coughlin (2009), Coughlin and Garrett (2009) and Ghent and Grant (2010), who used county-level data for particular US states, and Oster (2004) and Combs et al (2008), who focused on sales in smaller (zip code) areas in Connecticut and Minnesota respectively, there is a relative paucity of studies exploring determinants of sales across spatial units within a single lottery jurisdiction -despite the popularity of modelling lottery demand at the national and household levels. Further, studies which do consider spatial variation of sales do so with quite small geographical units of analysis.…”

The demand for gambling in Italian regions and its distributional consequences

“…However, interest in studies run at the local and regional levels is increasing in Europe (Baker et al ) due to their potential utility for policy‐making. Some researches proposed analyses at the regional or local level, using macro‐data aggregated by region, area or postal code (e.g., Clotfelter ; Vasche ; Jackson ; Hansen ; Price and Novak ; Oster ; Combs et al ; Garrett and Coughlin ; Ghent and Grant ; Leal et al ). The vast majority of the literature uses US datasets (e.g., Clotfelter ; Borg and Mason ; Jackson ; Price and Novak ) but a European literature on gambling consumption behaviour is emerging, with studies applied, for example, to Spain (e.g., Garvía and Guillén ) and Germany (e.g., Beckert and Lutter ).…”