COMPLEMETARITY AND SUBSTITUTION BETWEEN

Abstract

In line with the liberalisation of gambling laws across many countries of the world, the lottery and other gambling revenues increasingly feature prominently in many countries’ public finance planning. Apart from being a voluntary tax levied by the government, it is economical to administer, minimal to no compliance and enforcement are required and, most importantly, citizens pay it with a smile. It is also regarded as a precursor to future models of taxation where incentives can be used to sweeten citizens’ appetite to voluntarily pay tax. In the South African context Uthingo was contracted and mandated by government to run the lottery with a view to maximising revenue for good causes. In the effort to increase revenue, the lottery operator introduced a few games; this research seeks to quantify the extent to which these additional games are complements to or substitutes of the original game. The research is broken down into three periods. During the first period (2000/03/01–2001/09/29), only weekly draws were conducted on Saturdays; total revenue of R51 Million was collected per week. During the second period (2001/10/03–2003/11/22), the game frequency was increased from only Saturday to include an additional Wednesday draw; total revenue increased to R81 Million per week. During the third and last period (2003/11/26–2007/03/31), both Lotto and LottoPlus games were offered on Saturday and Wednesday, but the total revenue per week from both games plummeted by R3 Million to R78 Million. Using regression analysis with the aid of StatPro, in the first period the research found no substitutes; the complements were Sales from four weeks previously and the Sales immediately before the current draw. Furthermore, one of the two methods used showed an upward sloping demand curve while the second method showed a small drop in effective price resulting in disproportionate increase in revenue. In the second period, Sales conducted on the same day of the week as the current draw were found to be complements, whereas for the Sales on different day, the immediately past is statistically insignificant and the one before that is a substitute. For the third period, the only statistical significant ii complements are the sale of the current LottoPlus, immediate past Lotto and LottoPlus Sales. The substitutes were the Lotto and LottoPlus draws four draws ago. The research concludes that the board did things right but not the right things: the lottery board construed their mandate to maximise revenue too narrowly and subsequently targeted their games to the general population, which in the lottery context implies the poor and vulnerable. The board should have interpreted the mandate by the need to collect maximum revenue from the rich with a view to distribute to the poor. Recommendations are made on how the board could target the wealthy and affluent in their endeavours.