“That system” referred to closed primaries, in which party primary elections occur to nominate candidates for a general election, wherein only voters who choose to affiliate with a party may participate in one party’s primary. Louisiana’s blanket primary system, unique among the states, technically serves as the general election instead, where all voters regardless of affiliation may vote on all candidates regardless of affiliation running together in which the winner receives a simple majority, but if not secured then the two with the most votes advance to a runoff.
(Note: some observers use the slang term “jungle primary” in reference to Louisiana’s system. Others call it a “nonpartisan primary” because its election is really a general election; thus, prior to the Supreme Court’s Foster v. Love decision it properly was a “blanket primary” because the election occurred prior to the national election day, with any runoff occurring on that day.)
Edwards claimed the blanket primary had the edge on closed primaries, or by implication open primaries as well – where candidates compete for one party’s nomination to advance to the general election in a primary election in which any voter can participate in one party’s primary – because closed primaries reputedly produce “less bipartisanship, and less rigid ideological partisanship that is run by those at the extreme of the two parties” (the last phrase is convoluted, but I think he’s trying to say that more extreme ideologues end up elected and imposing their policy preferences in governing).
This reflects the political culture in Washington, DC, Edwards indicated, thereby implying that his state’s politics appear to feature fewer ideological appeals because the blanket primary sends more moderate politicians to Baton Rouge. Of course, you couldn’t tell by his rhetoric, where Edwards has proven one of the most polarizing governors in Louisiana history in how he always casts those who disagree with his policy preferences as enemies of what’s good for the state and its people.
How his bloviating illustrates potential extremism generated by the blanket primary system that got him there serves as anecdotal evidence against his argument, but there’s empirical evidence that disputes his allegation as well. The latest research on the question on how primary form influences the direction of policy output not only provides no real evidence to back that the kind of primary affects amount of polarization, but what little significant effect it does demonstrate is that the more “open” a primary, the more ideologically extreme candidates tend to win.
Incident to Louisiana, the study (using 1992-2010 data) showed it did have one of the least polarized legislatures. But it also revealed that the blanket primary had nothing to do with that, meaning other factors such as political culture explain this.
So, the data don’t confirm Edwards’ claim and, if we believe moderation a good in its own right, refuse to convey any advantage to the blanket primary. But closed primaries do have an unqualified good: strengthening political parties that bring greater responsibility and accountability to the public. By creating incentives to judge electorally organizations that live or die by election results, this forces greater adherence to the public median ideology. In contrast, candidates less moored to party fortunes who emphasize more their personal appeal than issue preferences have increased ability to deviate in their actions from that median position.
In short, the personalistic, issue-diluted Louisiana political environment from which Edwards and other elected officials sprang would transform with stronger parties, encouraged by closed primaries that give their ideological platforms greater prominence in a vote decision for a candidate. That’s why Edwards and these others, who have greater ability to deviate from the public on issues because of this environment, never will see a “benefit” from a closed primary system, because it threatens their political careers.