GOP Registration Higher Than Ever, But Will It Mean More Republican Votes?
There are more registered Republicans in Louisiana today than at any earlier point in the state’s history. The Republican Party added more than 65,000 registered voters since the last presidential election, an 8 percent increase over their number in 2012. While the total of all registered voters has hit a new height, the overall growth rate (about 1 percent) falls well below the growth in the number of registered Republicans. This growth is all the more sharp when contrasted with the loss of nearly 100,000 registered Democrats, a 7 percent decline.
This latest four-year bump in registered Republicans follows a decades-long shift in the attitudes and behaviors of Louisiana voters. The change goes well beyond registration to include an even more stark shift in the voting habits of the Louisiana electorate, which once consistently elected Democrats but now usually, though not always, elects Republicans. Does the continued growth in the share of voters declaring themselves Republicans on their registrations mean Republicans stand to gain even more at the ballot box as well?
There are more Republicans everywhere, well almost everywhere
Republican registration is growing throughout Louisiana. Sixty parishes saw a net gain in the number of registered Republicans since 2012. The number slipped in only four: Orleans, St. John the Baptist, Caddo, and Madison. Yet, three of these saw an even larger drop in the number of registered Democrats allowing registered Republicans to grow their share of the local electorate despite their drop in numbers.
In fact, registered Republicans increased their share of the voter rolls in every parish except Orleans. This growth tended to be strongest in the state’s whiter, rural parishes. This is probably because they had more room for growth there than in parishes that have larger minority populations, which tend to support Democrats, and suburban parishes, which tended to already have relatively large shares of Republican voters.
Party registration is a messy measure of partisanship…
It is important to remember that party registration is an imperfect measure of partisanship and voting behavior. For one, there are many registered voters who remain formally unaffiliated but fairly consistently vote for one party over the other. This reflects a national trend in which many voters do not like to identify themselves as partisans despite behaving very much as partisans.
The other reason is more specific to Louisiana and its southern neighbors. Throughout the Deep South in particular there are a great many voters who are registered as Democrats but who tend to vote for Republicans.
Consider the chart below. The solid line plots the share of voters registered as Democrats in each parish in 2012, from the parish with the smallest share (St. Tammany at 27 percent) to the parish with the largest share (St James at 74 percent). The dashed line plots Obama’s 2012 vote share in these parishes. If Democratic registration is a good measure of Democratic voting strength, the lines would be fairly close. They are not.
Generally, registration overstates Democratic strength…by a lot. Almost everywhere, there were fewer votes for Obama than there were registered Democrats. Obama’s vote share exceeded Democratic registration in only five parishes: Orleans, East Baton Rouge, Caddo, Madison, and St John the Baptist.
What’s going on here? This is partly a function of turnout. Registered Democrats had a lower turnout rate in 2012 than registered Republicans (as they often do). So, part of the reason the solid line falls below the dashed line is that a share of registered Democrats are just not voting.
But that’s not the main reason. Even when comparing Obama’s vote share to the share of the electorate registered as Democrats and actually cast ballots, the gap remains. The reason is that many registered Democrats in Louisiana do not typically vote for Democratic candidates. These are voters who registered as Democrats decades ago when most Louisiana voters still identified with a party they saw as moderate-to-conservative, as it had been through much of the twentieth century (at least in the south).
Keep this in mind later this week when pundits begin counting the number of registered Democrats who cast early ballots in each parish in order to get a sense of how the election may go. Knowing the number of registered Democrats who vote tells very, very little about whether or not those voters are actually casting ballots for Democrats.
…but it does work better at identifying Republican voters
Republican registration is a messy measure of Republican strength too, but in a way that makes it more useful. Registration greatly understates Republican voting strength. Consider this chart, which is similar to the last but plots Republican registration against Romney’s vote share in 2012.
Again, the lines are not at all close. However, the pattern is opposite than for Democratic registration. Romney beat Republican registration everywhere. Unlike registered Democrats in Louisiana, registered Republicans are typically committed partisans. They are voting for Republican candidates… and they get a lot of help from registered Democrats and unaffiliated voters who vote Republican too.
Newly registered Republicans were probably voting Republican anyway
The voters who have registered as Republicans since 2012 are likely to vote for Republicans in the presidential, senate, and congressional races this year. But that doesn’t necessarily mean Republican candidates will do all that much better because of them.
This may seem a little counter-intuitive. Why wouldn’t more committed Republicans help Republican candidates? Because the new registered Republicans were probably already voting Republican before they updated their registration, or, if they are first-time registered voters they are replacing older voters who were already regularly voting for Republicans even though they never registered as one themselves.
The chart below shows how the increase in the share of Republican registration in each parish is related to the 2012 'Romney surplus' (i.e. the gap between the share of registered Republicans and the 2012 Romney vote in the chart above). The largest gains in Republican registration are in those parishes that had the largest 'Romney surplus,' that is, the places that had the most people who were not registered as Republicans but were voting for Republicans.
Republican vote share has leveled off in recent years
Another way to see this is to look at the trend in Republican registration compared to the trend in the Republican share of the vote in Louisiana elections for president and U.S. Senate. Republican registration has steadily increased. It rose from 24 percent to 30 percent of registered voters since 2004, but the vote share in the state for Republican presidential candidates has remained inert: 57 percent in 2004, 59 percent in 2008, and 58 percent in 2012. Republican candidates in the state’s last two U.S. Senate races received nearly the same share of the vote: 57 percent in 2010 and 56 percent in 2014.
If you prefer a more rigorous test, the 2008-to-2012 parish level gains in Republican vote share contributes almost nothing to accounting for the variance in 2012 Romney vote share after accounting for parishes’ prior registration and propensity to support Republicans. [In statistical terms, including change in Republican registration in a model of the Romney vote hardly boosts the adjusted R-squared at all, from 0.975 to 0.979.]
More registered Republicans no longer means more Republican voters. Why? Increasingly, Louisiana voters are aligning their registration to match their voting behavior, rather than the other way around.