bq. If ’00 CA SEN nominee/ex-Rep. Tom Campbell edges out ex-HP CEO Carly Fiorina in the CA SEN race, he may have his ballot designation to thank.
So argues Felicia Sonmez at Hotline. She cites a recent poll that included the “occupations” of the candidates, which the candidates essentially choose and are listed on the ballot. In this case:
bq. Fiorina is described as “Business Executive.” DeVore is “Assemblyman/Military Reservist.” And Campbell — despite his five terms in Congress and his tenure as Gov. Arnold Schwarzenegger’s state finance director — is listed as “Economist/Business Educator.”
Did it make a difference in this case? It’s hard to know, absent an otherwise equivalent poll that did not include the occupations. However, there is some political science research that suggests the inclusion of ballot labels matters. In this article (gated), Monika McDermott examined a 1994 Los Angeles Times poll that randomly assigned respondents to receive or not receive information about occupations. McDermott focuses on a bunch of down-ballot races (Lieutenant Governor, Attorney General, Treasurer, Controller, etc.).
Respondents who saw the occupations were more likely to make a choice (i.e,. not abstain), and there were also differences in their choices. For example, in the race for Treasurer, respondents were 13 points more likely to choose Phil Angelides (“businessman, financial manager”) over Matt Fong (“appointed member, state board of equalization”) when the occupations were given.
McDermott isn’t in the business of figuring out what kinds of labels are more effective than others. So while her findings from 1994 seem sensible — “businessman” sounds better than “state board of equalization” — we really don’t have any systematic understanding of this issue.
The ultimate question is how well these findings generalize to the CA Senate primary. McDermott notes that cues like occupation are probably most valuable when voters have little other information about the candidates. It’s unclear to me whether the California primary will meet these conditions. The Republican gubernatorial candidates will have advertised a lot and the smaller primary electorate will tend to be more attentive to politics generally. Occupational information would thus matter much less.