Periodically political enthusiasts express concern that members of a particular political party will conspire to swing the result of the opposing party’s primary election by strategically voting for a candidate who does not express the actual will of that party’s “real” voters. This form of bad-faith strategic voting is sometimes called party raiding.

Party raiding is only feasible in states with open primaries, and fear of it is sometimes used as a argument in favor closed primary systems, which only allow registered partisans to vote in their respective primaries.

Wisconsin is an open primary state. In fact, the state’s Election Commission maintains no records of party affiliation whatsoever. Every party’s primary contests share space on a single ballot. Voters choose their preferred party in the privacy of the voting booth. No state presents fewer barriers to strategic party raiding than Wisconsin.

Nonetheless, there is no evidence that this kind of voting behavior occurs at all in Wisconsin. As I mentioned, registered voters do not have the option to formally affiliate with a party in Wisconsin. We can, however, measure party identification through public opinion data.

I pooled the results of three Marquette Law School Polls preceding the 2016 presidential preference vote and three surveys preceding the 2018 partisan primary. The combined dataset includes 3,515 likely voters. Each respondent was asked if they planned to vote in either the Republican primary, the Democratic primary, or if they didn’t plan to vote at all. We also recorded answers from respondents who insisted they would vote in “both” primaries, even though this would result in a spoiled ballot if carried out.

Respondents were also asked if they “usually think of yourself as a Republican, a Democrat, or an Independent.” Those who answered “independent” were then asked, “Do you think of yourself as closer to the Republican Party or to the Democratic Party?” We consider those who answered affirmatively as “leaning” partisans.

Here is how each partisan group planned to vote in the upcoming primary.

An identical share (2%) of Republicans and Democrats planned to vote in the other party’s primary. Even if this tiny share of people were indeed “party raiding,” they cancelled each other out. But there is no good evidence suggesting they weren’t voting in good faith. In the following general elections the share of self-identified Democrats or Republicans voting for a nominee of the other party exceeded 2%, so it’s quite likely that some share of self-identified Democratic voters genuinely preferred one of the Republican primary candidates and vice versa.