Table games, slot machines and casino revenue

Abstract

The main focus of this study is to assess the influence of table games on slot machine and total casino revenues. The subjects of the analysis are 24 rierboat casinos in the midwestern states of Illinois, Iowa and Missouri and three racinos in Iowa. A racino is a parimutuel racetrack which also offers slot machine gaming to its customers. Two econometric models were developed, one for slot machine revenue and one for total (slot machine plus table game) revenue at riverboats and racinos. Of particular interest is the effect on slot machine and total revenue of adding table games to the gaming operation. Slot machine revenue was found to increase with the number of slot machines and decrease with the number of table games. Slot machine revenue was estimated to decrease by 11.5% in the presence of 40 table games, the mean number of table games over the sample. Total win was found to increase with an increase in both slot machines and table games. The elasticity of total win with respect to the number of slot machines and table games was estimated to be 0.786 and 0.219, respectively. Total revenue was estimated to increase by 24.5% for an additional 40 table games.

¹ This article was accepted for publication, subject to final edit, white Dr. Thalheimer was at the University of Louisville, Louisville, KY, USA. AS is customary, the authors assume full responsibility for the analysis and the remaining errors.

Notes

¹ This article was accepted for publication, subject to final edit, white Dr. Thalheimer was at the University of Louisville, Louisville, KY, USA. AS is customary, the authors assume full responsibility for the analysis and the remaining errors.

² Includes land-based casino gaming, riverboat casino gaming, Indian casino gaming, legal bookmaking, card rooms and charitable gaming.

³ The West Virginia State Lottery permitted Mountaineer Park, a thoroughbred racetrack in West Virginia to install a limited number of slot machines at its facility on an experimental basis in 1990. The experiment was judged a success and in 1994 the West Virginia Video Lottery Act was passed giving statutory authority to the West Virginia Lottery to conduct VLT gaming at the two thoroughbred and two greyhound parimutuel racetracks in West Virginia, subject to local referendum.

⁴ The racino was Prairie Meadows parimutuel horse racetrack. Although legislation permitted table games at all three Iowa racinos, subject to certain conditions, only Prairie Meadows installed them that year.

⁵ There were two Iowa riverboats that started up in 1996, both located in Council Bluffs. A third riverboat, the Emerald Lady/Catfish Bend riverboat in Fort Madison was excluded as a subject riverboat due to lack of information on its demand characteristics.

⁶ A fixed effects model was also considered, but was not chosen since these effects are highly correlated with the variable, ACCESS. The random effects, on the other hand, are assumed to be uncorrelated with the included demand variables and capture the between-boat effects. Inclusion of time effects in the model was also considered. These effects were found to be insignificant and were dropped from the model.

⁷ The algorithm used was XTREG as described in the Stata Corporation, College Station, TX, STATA® Reference Manual. The method used is FGLS (feasible generalized least squares).

⁸ As win varies widely across riverboats, it was suspected that residuals may be heteroscedastic. However, results from a test for heteroscedasticity, a regression of the squared residuals on the predicted value of the dependent variable, did not provide significant evidence to justify its presence.

⁹ Elasticity computed as (PCSLOTWIN)/(SLOTS) × SLOTS/PCSLOTWIN = α ₂.

¹⁰ Elasticity computed as (PCSLOTWIN/(TABLES) × TABLES/PCSLOTWIN = α ₃TABLES.

¹¹ Percent change in slot machine win computed as: [exp(α ₃TABLES) − 1]100.

¹² The algorithm used was XTREG as described in the Stata Corporation, College Station, TX, STATA® Reference Manual.

¹³ As win varies widely across riverboats, it was suspected that residuals may be heteroscedastic. However, results from a test for heteroscedasticity, a regression of the squared residuals on the predicted value of the dependent variable, did not provide significant evidence to justify its presence.

¹⁴ A fixed effects model was also considered but was not chosen since these effects are highly correlated with the variable, ACCESS. The random effects, on the other hand, are assumed to be uncorrelated with the included demand variables and capture the between-boat effects. Inclusion of time effects in the model was also considered. These effects were found to be insignificant and were dropped from the model.

¹⁵ Elasticity computed as (PCSLOTWIN)/SLOTS × SLOTS/PCIOTWIN = β ₂.

¹⁶ Elasticity computed as (PCSLOTWIN)/TABLES × TABLES/PCTOTWIN = β ₃TABLES.

¹⁷ Percent change in slot machine win computed as: [exp(β₃TABLES) − 1]100.

¹⁸ Illinois Gaming Board, 1997 Annual Report, p. 9, survey results.

¹⁹ MapInfo Professional™.