Closed primaries help parties, provide clearer direction

Fascinating even to elected officials is just what impact Louisiana’s odd nonpartisan blanket primary system has on partisan politics and elections, which has an impact, perhaps salutary, on subsequent policy.

The one in charge of elections, Sec. of State Tom Schedler, as do many observers, note the decline in registrations for Democrats and growth among Republicans and even more in the catchall Other classification, which includes all registrants of the minor recognized parties Greens, Libertarian, and Reform, all registrants that put down another label of an unrecognized party, and those that indicated no partisanship.

He offered two theories as to why the Other category has grown the fastest – it’s gone from 18 to 26 percent of the electorate since 2000, while GOP registrants have increased only from 22 to 28 percent and Democrats fallen from 60 to 46 percent – which are somewhat complementary. One is “frustration with the major parties” and the other the participation opportunity the electoral system presents; each merits some discussion.

The frustration hypothesis actually serves as a corollary to a larger concept, called dealignment. Scholars have argued that for at least six decades, and probably longer since survey research that could detect the phenomenon only came into existence around the World War II period, that voters increasingly have deemphasized partisan appeals in organizing their political worlds and basing their decisions within it. A number of sociological and technological changes have spurred this, leaving a more rootless electorate than ever not housed fairly permanently in one major party. As parties become less needed as intermediaries to understand and act in the political world, it becomes easier to question them and their utility for making vote choices, and thereby reduced willingness to affiliate with them.

Louisiana’s nonpartisan blanket primary system magnifies this. By not insisting, except for presidential primaries, that voters express some kind of membership in a party in order to pick a nominee for that party – indeed, dispensing with the notion of primaries at all to present nominees for the general election and reducing candidate’s party labels to a mere data point for each instead of employing these additionally to organize and conduct a nomination process – voters receive no encouragement or threat of sanction to adopt a particular or any party label. If party labels chosen at registration discriminate only when comes to a voter’s participation in presidential preference primaries and for no other election, thereby in every other contest permitting regardless of label participation in choosing a candidate for “nomination” and election simultaneously, then the labels won’t mean much especially for those who care little about politics and only as statements of personal ideology and affect for parties for those who have much more interest.

In theory, were closed primaries – where only members who choose to affiliate with a party may participate in nomination processes for its candidates – instituted for all elections in Louisiana, while many of the disinterested Other voters would continue to express no party preference, those who found politics salient would want to influence at least one party’s nomination process as almost all of them in ideological terms would lean towards one and thus would abandon the Other category to pick (almost always) Democrat or Republican labels. Had such a system existed throughout the 21^st century, Democrat registrations would have declined less, Republican registrations would have increased more, and the rise in Other registrations would have been less pronounced (given the prevailing dealignment trend).

This can be tested, because of the fact every four years a closed primary appears, and also because of the statutory change in effect for the 2008 and 2010 federal elections that created semi-closed primaries, where the state parties could choose whether to restrict participation in nomination primaries. As it happened, the state GOP made completely closed primaries for these two cycles, while Democrats chose to allow a restricted form of open primary where party registrants and no party registrants could participate in their congressional primary elections.

Over time, change in proportion of registrants for each system category – years with no closed primary elections, presidential election years, and the two years of closed and partially-closed primaries – compared to the previous year in that quadrennial cycle (otherwise, comparisons would be contaminated across cycles) can reveal the impact of electoral system. White Democrat (because they would be expected to abandon the party as dynamics changed), Republican, and Other registrants can be compared for change across these three system categories.

As it is, using data from 2000 on (October registrants every year except 2005, when July’s had to be used as October records were not retained), among white Democrats change (annualized average) for blanket years was a decrease of 1.00 percent, for presidential years of 1.10 percent, and for closed years of 1.03 percent. For Republicans the respective figures were increases of 0.40, 0.52, and 0.38 percent, and for Others increases of 0.54, 0.57, and 0.42, respectively.

This pattern fits the expectation that electoral system matters. With presidential nominations at stake, among white Democrats more existing registrants seeking to abandon the party or new registrants part of generational change less favorable to the party or from out of state reduce the party rolls because to participate in selecting a GOP presidential candidate they must declare themselves as Republicans, hence the faster increase in Republican registrants during this period. The two year interregnum of the federal closed primary experiment also shows this, at a slightly lower level probably because of substantially lower turnout rates (why switch or register if you don’t plan on voting, thereby diluting the trend) and that Democrats did not enforce closed primaries in this period. The periods of much less Democrat reduction and Republican addition occurred were in the years of no closed primary elections for any offices.

While the major party figures appear to reflect the effects of system, the Other figures seem to capture the dealignment phenomenon as a function of registration and turnout. Presidential election years bring the greatest efforts for the former and the highest of the latter in Louisiana, so the trend in favor of non-major-party registering would be most pronounced in these years. That the blanket primary years showed almost as much increase shows more the impact of system, for despite much more emphasis on registration and turnout change in presidential years in essence get dampened almost to the blanket primary level because of the closed primary imperative.

This surface investigation (and hampered with only 12 data points) points to primarily the long-term, secular effects of dealignment and secondarily to partisan realignment as the main drivers in Democrats sloughing off to the benefit of Republicans and especially Other category. But were closed primaries instituted across the board, Democrat defections would slow with more stunted growth among Others and faster GOP expansion.

Such shifts would strengthen all state-recognized political parties organizationally. Likely it would enable the two major parties to have their candidates present clearer and more coherent ideological choices for voters. In turn, a more definitive sense of direction may make it easier for Louisiana to tackle chronic fiscal problems resulting in a schizophrenia from simultaneously arguing for lower taxes yet tolerating higher spending.