State‐sponsored lotto games, because they are pari‐mutuel and because jackpots with no winner are rolled over into the next drawing, present an excellent opportunity to test for market efficiency. Using data from Massachusetts, Kentucky, and Ohio, we investigate bettors' responses and test for weak‐form efficiency. Lotto bets do not have positive net expected returns, thus weak‐form efficiency exists. To evaluate strong‐form efficiency we utilize the concept of a rational expectations equilibrium. We find that in general lotto bettors' decisions to play generate a level of sales that conform to their original forecasts of expected value.

This paper examines sales patterns in the first three years of the UK National Lottery in order to estimate price elasticity of demand. Our long-run estimate is very close to the value of minus one required for revenue maximization. We find that the UK Government has succeeded in setting a framework for the National Lottery that maximizes turnover and therefore the amount of money available for both Exchequer revenue and "Good Causes."

The concept of rational expectations has typically been assumed, without testing, in the analysis of consumer demand and market efficiency in betting markets, including betting on lottery games. Lottery games offer an excellent opportunity to test how participants process the information that is available to them. Using the UK National Lottery as our particular case, we find that participants, in general, efficiently process available information. Specifically, they act as if they can, on average, forecast the level of sales for a given drawing.

Virtually all lottery agencies offer a variety of games to suit the tastes of players in an attempt to maximize revenue to the government. Using the UK National Lottery, which offers a variety of on-line and scratchcard games, the extent to which there is substitution or complementarity between games is evaluated Employing weekly data from the three UKNL lottery games offered over the sample period, it is found that own-game characteristics have, by far, the largest influence on sales. Some evidence is found suggesting that the lotto and scratchcard games are partial substitutes for one another. Thunderball sales appear independent of the other two games. Some evidence is also found that the Wednesday and Saturday drawings of the lotto game are substitutes. The overall conclusion is that Camelot has successfully designed and marketed three games that each appeal to bettors in different ways. Thus, sales from one game do not seem to seriously cannibalize the sales of the other games, with the exceptions noted above. Further, the introduction of another, temporary game (Big Draw 2000) contributed to net sales. These results also suggest that the games do not appear to be complements to each other, indicating that the various arguments as to why the games may be so (transactions costs, brand awareness, and the portfolio effect) do not appear to be very strong.

No abstract