With the embarrassing number of hopeful Obama appointments running into tax cheating problems (the latest being Ron Kirk), it's natural to wonder if evasion by high profile leftists is illustrative of a real world trend, or just a string of unfortunate anecdotes.
The GSS provides some relief for that wonder. It provides the results for 2,418 people queried on whether or not cheating on taxes is wrong, by political orientation. The first graphic from the GSS shows the distribution of responses. The second graph shows the mean tax compliance score, computed by designating "not wrong" as 1, "a bit wrong" as 2, "wrong" as 3, and "seriously wrong" as 4, and then averaging the responses for each of the seven categories of political orientation (click for higher resolution).
The standard deviation for the dataset is .76, so the difference between self-described conservatives and extreme liberals is nearly one full SD. Amalgamating the responses into three categories yields one-third a SD between liberals and conservatives:
Liberals do not consider cheating on taxes to be as morally problematic as conservatives do. This presents an obvious moral quandary of its own, as, putatively less surprisingly, liberals are more likely than conservatives are to favor greater amounts of taxation and wealth redistribution.
The purest question the GSS asks with regard to a respondent's philosophical position on taxation is, "If the government had a choice between reducing taxes or spending more on social programs like health care, social security, and unemployment benefits, which do you think it should do?" The GSS provides results for 970 respondents to this question by political orientation:
There is a full standard deviation difference between extreme liberals and extreme conservatives on this most standard of political issues*. Combining the shades of liberalism and conservatism into a single category, more than half a SD still separates those on the left from those on the right:
Attitude is not behavior, and I am unaware of any studies on the political persuasions of convicted tax cheats**, but as a self-described empirical paleoconservative, it is difficult not to find parodiable humor mixed with irritation in discovering that those most likely to favor increased taxation and redistributive economics are also the most likely to approve of illegally acting to avoid having to suffer on the contributive side of the equation.
GSS variables used: POLVIEWS, TAXCHEAT, TAXSPEND
* Parenthetically, this shows the presumption that the liberal-conservative spectrum as represented in the GSS is a gauge of positions on social issues rather than economic ones is a stretch at best. Optimally, the GSS will ask a couple of questions on political orientation in the future in place of the one now asked. Cliched though it may be, separately inquiring about both a respondent's social and economic liberalism or conservatism would accomplish this.
Notice, too, that the graph's parameters are set from 40% to 90%--even among self-described conservatives, nearly half of people favor more spending by government over the reduction of tax rates. Too many people have faith in Leviathan.
** However, in Freedonomics, John Lott shows that Republican criminals are as elusive as leprechauns (p184):
[Based on a Public Opinion Strategies survey] I found that felons were 36% more likely than non-felons to have voted for Kerry over Bush and 37% more likely to be registered Democratic [after controlling for socio-cultural factors like race, gender, educational attainment, etc]. ...
While not all felons may be as Democratic as those in Washington State, the survey indicates that the previous estimates understated how frequently felons vote Democratic. Remarkably, it looks as if virtually all felons are Democrats. Felons are not just like everyone else. And the fact that felons are even more likely to vote
Democratic than previously believed surely guarantees that some Democratic operatives will continue their efforts to get them to the polls.