Marshall: Turnout and representation

Marshall. 1978. Turnout and representation: Caucuses versus primaries. American Journal of Political Science 22:169-182.

RESEARCH QUESTION: Are caucuses less representative than primaries?

DESIGN: Minnesota uses caucuses for presidential nominees, primaries for everything else. Marshall compares both groups to one another and to the general election voters. He wants to establish that caucuses are no less representative than primaries.

FINDINGS: No. They are not. Demographically, caucuses are somewhat worse. But ideologically, caucuses perform about as well [poorly, actually] as primaries. Both primaries and caucuses attract higher educated, wealther, and generally higher status people.

1. Caucuses: Overwhelmingly male (people who like to argue), but overwhelmingly pro-women's rights. Non-attenders are overwhelmingly female and overwhelmingly pro-male.
2. In about a third of policy areas, caucus attenders are statistically significant from the general electorate--primary voters aren't. He dismisses the differences as too small to matter, though. (Table 3).
3. Caucus-goers have stronger (more extreme) views about incumbents.

DATA

Using survey data, he looks for significant differences between (1) (Democratic) caucus attendees and Democratic general voters and (2) (Democratic) primary voters and Democratic general election voters.

CONCERN

Although the findings are interesting, Marshall downplays the differences between caucuses and primaries. His own data suggest that caucuses are less representative than primaries.