New lotteries

Three questions about gambling with public policy implications:

• Does a new gambling product, such as a lottery ticket, crowd-out spending on other forms of gambling or does it reduce non-gambling spending? Does spending on necessities fall?
• Are there specific effects of new instant lottery products such as 'scratch-off' tickets?
• Do people make informed choices of lottery products? If consumers make informed choices their purchases are welfare-increasing.

A recent paper by Melissa Kearney provides an analysis of these issues using State-based household expenditure data for the US 1982-1998. Her main findings:

1. Introducing a state lottery reduces non-gambling spending by $24 per adult per month compared to the average cost of lottery sales of $18 per month. For the average household spending on lottery tickets is financed completely (and permanently) by reduced non-gambling spending. People do not have a fixed total gambling demand, merely substituting new gambling products for old.

2. The poorest 1/3 of households have the biggest response to a new state lottery. Their non-gambling spending falls by 2.5% (3.1% when the lottery offers instant games). These households spend less on food eaten in the home (2.8%) and on housing (5.8%). These are averages - many households do not buy lottery tickets - so actual figures for those who gamble are much higher than this.

3. Lotteries have investment and entertainment value. Assuming pecuniary and entertainment components are separate, sensible behaviour suggests purchases should be greater for products with higher expected returns and this is so when other factors are controlled for. Consumers however also respond to non-wealth creating entertainment features. Hence purchase is consistent with informed choice - deriving an entertainment value equal to the price of the lottery (cost less expected return) and then making informed investments. However it is also consistent with the belief that consumers believe that entertainment characteristics enhance their chance of winning so the evidence is not necessarily consistent with informed choice.

See: Abstract 'State lotteries and consumer behavior', Jourmal of Public Economics, 89, 2005. Earlier complete version here. Related papers on gambling here.