A Conservative Windfall

British economists Nattavudh Powdthavee and Andrew Oswald suggest two explanations for people's political attitudes: (1) they are the result of “deeply ethical views,” or (2) they are motivated by self-interest. Their study is relevant to the second possibility. They looked at the effect of winning the lottery on political attitudes.

Many previous studies have found that rich people are more conservative, but this is subject to different interpretations. Higher income could cause conservatism, conservatism could cause higher income, or they could both be jointly caused by some other variable, i.e., having wealthy parents. Studying lottery winners is an interesting idea because a lottery is a truly random assignment of people to the condition of being “rich.” Since the authors had data from before and after the win, they could look at its effect on political attitudes while holding individual differences constant.

The data came from the British Household Panel Survey, an ongoing study in which the same people are contacted every year. Between 1996 and 2009, 4277 different people reported winning the British National Lottery 9003 times. 94.65% of these were small wins of less than £500 (as of this writing, $820). The remaining 5.35% were over £500 and were considered big wins. (The biggest win was £185,000. Unfortunately, they do not report an average.) Winners were compared to people who had no lottery winnings, either because they did not play or played and lost.

All respondents were asked which political party they favored. Only supporters of the Conservative and Labour parties were included in the analysis, since these parties are clearly right and left leaning. (The authors classify the Liberal Democrats as a centrist party.)  Participants were also asked to locate themselves on a 7-point scale, from a strong supporter of the Labour Party to a strong supporter of the Conservative Party. The scale measured changes in attitude strength regardless of whether the participant changed parties. These data were statistically corrected for prior household income and 12 other control variables.

By comparing the participants' answers to those of the year before, it was possible to identify people who had switched from favoring some other party (or none at all) to favoring either the Labour or Conservative parties in the year of the win. 

Approximately 13% of the non-winners switched to the Conservative Party that year, along with 14% of the small winners and 18% of the big winners. The data from the 7-point scale were consistent with these findings, showing that the greater the winnings, the greater the conservative shift. For unknown reasons, the conservative shift among winners was greater for men than for women. 

Since each lottery winner's political leanings were measured before and after their windfall, this design goes a long way toward controlling alternative explanations. However, there is a problem. It's possible that people who play the lottery are not representative of the population. For example, in the US, the lottery functions as a regressive tax, since low income people spend more per capita. In addition, men, middle-aged (45-64) people, and people with less education play more. In defense of the representativeness of their sample, the authors note that 57% of UK citizens play the lottery at least once in a given year. However, they had no measure of how often each person played. Since the more often you play, the greater your chance of winning, the big winners were probably people who played regularly.

My guess is that winning the lottery sensitizes people to the tax consequences of winning. The British income tax is more progressive than ours, and Conservative Party propaganda emphasizes keeping taxes low. People who play the lottery often may differ from non-players in a way that makes them more susceptible to a conservative change. Maybe they are particularly sensitive to variations in their income.

This is not the first study to look at the relationship between wealth and self-interest. Several American studies have found that higher income people are less helpful and more likely to cheat or engage in unethical behavior.

You may also be interested in reading:

Me First

Class Acts

Bending the Curve