Chapter 2 examines the harm associated with being a problem gambler. Problem gambling is conventionally determined by having a score in a questionnaire screen that exceeds some critical value. The UK is fortunate in having large representative sample surveys that embed such questions, and our estimate from the 2010 survey is that several hundred thousand people in the UK could be afflicted by PG. However, existing literature has not evaluated the size of the harm associated with being a problem gambler and this chapter uses this individual level survey data to evaluate the effect of problem gambling on self-reported well-being. Together with a corresponding effect of income on well-being a money-metric of the harm associated with being a problem gambler is derived. An important methodological challenge is that well-being and the harm experienced may be simultaneously determined. Nonetheless, instrumental variable estimates suggest that problem gambling imposes an even larger reduction in well-being than least squares would suggest. The role of gambling expenditures in the transmission between problem gambling and well-being is considered, distinguishing between draw-based games, such as lotto, from scratchcards, and from other forms of gambling.
Chapter 3 investigates the price elasticity of demand for the UK National Lottery – a state-licensed, draw-based lotto game. Little is known about the price elasticity of demand for gambling products because the “price” is typically hard to define. The exception is “lotto” where an economics literature has focused on the response of sales to variations in the prize distribution. Existing literature has used these responses make inferences about the price elasticity of demand, where price is defined as the cost of entry minus the expected winnings. In particular, the variation in the value of the jackpot prize pool, due to rollovers that are a feature of lotto, has been used as an instrument for price. This chapter argues that rollovers do not make valid instruments, because of their correlation with lagged sales, and propose an alternative identification strategy which exploits two arcane features of lotto. Finally, this chapter evaluates whether changes to the design of the UK National Lottery in 2013 and 2015 had a positive effect on the sales figures.
Chapter 4 investigates the extent to which the large, flat-rate tax imposed on the UK National Lottery is regressive. This chapter evaluates a Working-Leser demand model for lotto tickets using both Heckman’s selection model and Cragg’s double hurdle estimator using household-level data. A unique strategy is employed to identify these two-stage routines by exploiting exogenous differences in consumer preference arising from religious practice. The income elasticity of lottery tickets is found to be significantly lower than previous estimates, suggesting that lottery tickets are inferior goods and that the (high) flat-rate tax imposed on lotto tickets is more regressive than previously thought.
Whilst the three chapters are stand-alone essays, they are linked by the use of modern statistical techniques and the use of the best possible data. Together, they address key issues on the economics of gambling and the results are new to their respective literatures and of interest to academics and policy makers alike.