Opposing the Lottery in the United States: Forces behind Individual Attitudes towards Legalisation in 1975

Abstract

In the 1970s, opposition to the lottery started to fracture in the USA. This study examines causes of the fracture and historical factors that contributed to changes in individual attitudes towards legalisation. The opponents at the time held to traditional arguments against legalised lotteries—negative economic effects, costs to others and increased crime. Unlike the past, however, there was weak religious institutional opposition to lotteries. Individuals with a strong commitment to their religious affiliation were more resistant to pro-lottery arguments, but in most cases could be convinced to support the lottery. The pre-World War II generation remained steadfast against the lottery, but there was relatively greater support among the post-World War II generation. This study has examined the 1975 survey data using a logit model to predict future legalisation in states with large population samples. As expected, analysis of 1975 attitudes shows that states with low levels of opposition are likely to legalise lotteries earlier than states with high levels of opposition.

Acknowledgements

This research was funded in part by a grant from the Carnegie Mellon Foundation. The author would like to thank the referees for their comments.

Notes

 1. In Chicago, the Catholic Church issued a letter requesting parishes to discontinue bingo nights when it was found in violation of the law (Commonwealth, 1960). In New York City, the Deputy Chief Inspector was embroiled in a controversy when he started to close down bingo nights at local churches (New York Times, 1954).

 2. See Brenner and Brenner (1990) for the list of states. They also noted that promoters in the 1930s found ways to take advantage of the increased willingness to participate in a game of chance. It became a widespread practice for companies to run ‘contests’. Because it was not a ‘pure chance’ game, i.e. as customers received something for their purchase, contests were not considered a lottery.

 3. See Walker and Barnett (1999, p. 184) for a list of researchers who have used this definition.

 4. Data to calculate the tax burden was taken from the Statistical Abstract of the United States (1971).

 5. Episcopal, Luthern, Presbyterian, and Congregationalist were grouped in the survey and were call the mainline Protestant denominations.

 6. Two variables are not interacted with the lottery—Parimutuel and Non-Gambling State—because neither type of states allowed the lottery.

 7. Marginal probabilities are estimated by examining a unit change in the independent variable, holding all other variables constant at assumed values. Continuous variables were assumed at their mean value, and dummy variables were assumed to be for the following characteristics: state characteristics -a parimutuel, non-gambling, and rural; individual characteristics—illegal gambler, high school degree and Jewish religious affiliation. For continuous variables the unit change is 10% from the mean. For indexes the unit change is in terms of the actual discrete change in the variable.

 8. The level of significance also increased suggesting that the non-religious variables were creating some measure of multicollinearity.

 9. Only the interaction variables and the religious commitment variable are provided in the table. All other coefficients had the same sign, within the same size, and had the same level of statistical significance as the original model, except for property tax burden. It was significant at the 10% level.

10. This result is likely because of the presence of multicollinearity in the model. When the Fundamentalist dummy is dropped from the model, the property tax interaction variable is statistically significant. The same result held true for the religious attendance model.

11. The score was calculated by taking the difference squared in actual rank from the predicted rank, divided by 240 (the score of the model if it completely reversed the rankings) and subtracting this percentage from one.