Media coverage makes a fetish of novelty, overemphasizing change and ignoring continuity. A 10% increase in a city’s murder rate can be portrayed as a crime wave even when the overall rate remains near record lows. The power of the Latino vote is growing and will loom large in future decades. However, Latinos still do not punch their demographic weight at the ballot box. There are three reasons for this. Latinos are less likely than other groups to be citizens, are less likely to vote even if they are eligible, and are clustered in safe states.
The following five states all have outsized Latino populations. None of them will be close in a competitive 2016 election:
State |
% Latino (2012) |
Obama's 2012 Margin |
New Mexico |
46.3 |
10.2 |
California |
37.6 |
23.1 |
Texas |
37.6 |
-15.8 |
Arizona |
29.6 |
-9.1 |
New York |
17.6 |
28.2 |
Only Arizona was within 10 percent, and it was a whopping 14 points from the national tipping point. If Arizona is close, the national election will not be. Over half of the national Latino population live in the Presidential hinterlands of California, Texas, and New York.
However, the electoral college wears a Janus face. While the swing states are only 7.9% Latino, this sliver of votes is concentrated in the most important swing state of all: Florida. In 2016, Florida’s eligible voters should break down as follows:
Non-Hispanic White |
Black |
Latino |
Asian |
Other |
63.5% |
14.5% |
19.4% |
2.1% |
0.8% |
Florida is important for three reasons: (1) it is the biggest prize among swing states with 29 electoral votes; (2) it was very close in the 2012 election; and (3) with Nevada out of reach in a close election, Republicans must either win Florida or make big gains in the upper mid-West in order to win the White House.
Furthermore, Republicans depend upon Latino votes in Florida to an extent they do not elsewhere. In 2012, Romney did better among Florida Latinos than Obama did among Florida whites:
Romney’s Florida Vote Share by Race
WHITE |
BLACK |
LATNO |
ASIAN |
62 |
4 |
39 |
26 |
The path dictated by a “more white votes” strategy diverges at the Florida line. If Republicans could plausibly pursue a nativist strategy and still win Florida, then such a strategy could work without a major realignment in the electoral map. Conversely, if such a strategy would necessarily cede Florida to the Democrats, it would require Republicans to capture blue bastions in the Upper Midwest.
There Aren’t Many “Missing White Votes” in Florida
Despite winning Florida, Obama did five points worse than Kerry among white Floridians.
Kerry’s 2004 Florida Vote Share by Race
WHITE |
BLACK |
LATINO |
OTHER |
42 |
86 |
44 |
34 |
Obama won 37% of Florida's white vote in 2012. Roughly half this erosion resulted from losses among voters age 65 and over. Kerry won 48% of senior votes in 2004 while Obama won a paltry 41%. The senior vote in Florida is overwhelmingly white, so this movement reflects a change in the behavior of senior, white voters. In 2004, many of Florida’s senior voters were still part of the New Deal Coalition for whom Roosevelt was a living memory and the Great Depression justification for a strong safety net. By 2012, this generation was mostly dead or senescent, and most seniors came of age during the boom years following World War II. These more conservative citizens made the senior vote 7 points more Republican. Because Roosevelt Democrats had already evaporated by 2012, this trend will not continue.
Furthermore, Romney achieved crushing margins among culturally Southern whites. North Florida displayed the same voting patterns as Alabama. In the 27 northern-most counties, Romney clobbered Obama with 61.2% of the two-party vote. Given that 20% of voters in these counties are black and another two to three percent are Latino, Romney must have won a full 80% of white votes in North Florida. These figures include cosmopolitan Leon and Duval counties, two areas in which Obama showed some strength. In the pine barrens of the panhandle, Romney won 9 out of 10 white votes. Not much room for improvement there!
I ran a county-level regression which predicted Obama’s vote share based upon four variables: (1) the proportion of Puerto Ricans in a county; (2) the proportion of blacks; (3) the county’s latitude; and (4) the natural logarithm of the county’s population density. The overall significance of this regression was a pleasing 7.02 X 10^-22, and it explained 81.3% of the county-level variation in Obama’s vote share. (The R-squared). All four predictive variables had a high significance. Even the least significant variable, the proportion of Puerto Ricans, only had only a 6 in 10,000 chance of being statistical noise.
Variable |
Coefficient |
T Statistic |
P-Value |
% Puerto Rican |
0.007274 |
3.63 |
0.000583 |
% Black |
0.00803 |
10.85 |
5.78x10^-16 |
LN(Population Density) |
0.0309 |
5.37 |
1.24x10^-6 |
Latitude |
-0.0416 |
-8.211 |
1.71x10^-11 |
Four features of this result are unsurprising. Obama did better in counties with high concentrations of blacks and Puerto Ricans, and also in counties with high population densities. He also did much better in polyglot South Florida (counties with smaller latitudes), than in North Florida, which is more culturally Southern.
One result has received little attention elsewhere and deserves emphasis: Puerto Rican voters favored Obama almost as overwhelmingly as blacks.
Conversely, the proportion of non-Puerto Rican Latinos had no measurable effect on Obama’s vote share! This population is large enough that it would show up in the regression if they displayed even a slight preference for one candidate over the other. It follows that non-Puerto Rican Latinos split their votes more or less evenly.
With new white votes hard to find, and blacks and Puerto Ricans difficult to convert, a decline in the GOP’s share of non-Puerto Rican Latinos would be fatal to its prospects in Florida. Even without the immigration issue, the continuing arrival of Puerto Ricans and the increasing potency of more liberal, third-generation Cubans makes it uncertain whether the GOP can maintain Romney's 39 percent share of the Latino vote. Any Republican Presidential candidate who is seen as anti-immigrant will have a tough time winning Florida in the face of an already challenging demographic headwind.