Donald Trump continues to win Republican primaries, tightening his grip on the word "inevitable." The problem is that a detailed examination of the actual results shows Trump to be the weakest frontrunner of either major party in modern times. If Trump can be stopped from gaining a first ballot majority in Cleveland, his electoral weakness makes denying him the nomination an easy call.
I reviewed the numbers in the Republican primaries and caucuses for 2016 and compared them against their best equivalent numbers for 2012. There are four conclusions that can be drawn: First, he may well have momentum within the media, but he has none with the voters. Second, Donald Trump is not attracting an overwhelming number of new voters. Third, his success is built on a peculiar dynamic where he is facing opponents strong enough to stay in, but not strong enough to knock each other out. Lastly, the Republican primary numbers show why Trump consistently performs so poorly against Hillary Clinton (and Bernie Sanders) in national polls.
It is very difficult to analyze political primary voting. Every Presidential primary differs in important ways. Outcomes are dependent, not independent, which is to say after Iowa, each subsequent primary or caucus is partly dependent on previous contests. With several states changing their primary dates every cycle, it is impossible to make state-to-state comparisons for every state. There are always different numbers of candidates and they drop out at different rates.
But there are some important general trends. On the Republican side, there has not been a drawn-out primary contest since 1976. By mid-March (and normally much before), the party coalesces around a nominee. More specifically, as candidates drop out, their votes tend to drift toward the frontrunner, and the stragglers lose support very quickly. Thus, a legitimate frontrunner at this point should not just be on top, but should be gaining vote share faster than his remaining rivals.
This dynamic is not happening for Trump. In his first four states, his vote share was 32.8%. After seven candidate departure, his percentage only increased to 35.4%, up just 2.6 points. Cruz went from 20.8% to 29.6%, while Rubio fell slightly from 19.8% to 19.1%. Kasich has gone up, from 8.3% to 9.0%. If you look at 2012, comparing the same first four with the later primaries, Romney went from 31.3% to 40.2%, far better than Trump. At the same time, while Santorum went up significantly (from 16.0% to 28.1%), both Gingrich and Paul collapsed. Gingrich fell from 29.2% to 22.2% and Paul’s vote nearly halved. The 2012 race was quickly coming down to two candidates, with Romney well ahead. For 2016, two of Trump’s three opponents are picking up vote share and his nearest competitor is closing the gap.
Romney was also getting majorities and near majorities in his vote totals. He won Idaho and Massachusetts outright and topped 47% in Florida and Michigan. In 2016, Trump has fared far worse, just missing a majority only in Massachusetts and Mississippi and nowhere else topping 45%. Trump is ahead, but he is not consolidating the Republican vote like past frontrunners, nor is he putting distance between himself and his remaining opponents.
Trump explains his success by claiming that he is bringing huge numbers of new voters into the Republican Party and that’s a big reason for his success and why he should be the nominee. Well, that is half right. Trump likely is attracting new voters, however, it appears that more are drawn to vote against him than for him.
Comparing equivalent primaries and caucuses between 2012 and 2016, we can see that over 2.3 million additional voters have participated. Since exit polls do not ask people if this is their first GOP primary, we have to make a guess as to how many of these voters could be Trump backers. My estimation takes the top vote-getter in each contest in 2012, subtracts that total from Trump’s total and then calculates that as a percentage of new voters. For example, in Iowa Trump got 46,427 votes, while the 2012 winner, Santorum got 29,839. However, the 2016 caucuses had increased turnout of 66,689. So, Trump gained 16,588 more votes than Santorum, but that is under 25% of the "new" voters.
How is Trump doing? Not so good. Using this proxy estimation, Trump did pretty well in Nevada, picking up 46.6% of new voters. Outside of Nevada, he only gained over 40% in one state, Tennessee. In most states he gained under a quarter of new voters. In South Carolina, he actually got fewer votes that Newt Gingrich, in spite of nearly 145,000 more votes cast.
Even worse, Trump performed poorly in swing states. Six states in the comparison set went for Obama in 2012, but are possible swing states: Michigan, Iowa, New Hampshire, Nevada, Minnesota and Virginia. Trump did best with new voters in Nevada, while gaining merely 1% in Minnesota. He gained just 24.9% in Iowa and 20.7% in Virginia (Note: in 2012, Gingrich and Santorum did not qualify for the ballot, the 2008 results were substituted). In the biggest electoral prize, Michigan, Trump gained just 20.3% of the nearly 350,000 new voters.
The conclusion? There is a huge increase in Republican turnout, but it seems to be as much to vote for Trump’s opponents (or against Trump) as it is to vote for Trump. Not only that, his biggest strength lies in states highly likely to vote Republican regardless and in lonely, utterly liberal Massachusetts.
Trump’s advantage is not with voters; it is that he is facing a kind of divided opposition that has never truly existed before in any contested Republican primary season. In a normal contested primary, candidates in third place or lower very quickly evaporate. But in this season, the third place candidate’s percentage has fallen very little and the fourth place candidate has actually gained vote share as the field has narrowed.
The competitiveness of two opponents is not Trump brilliance, it is the luck of happenstance. Given how poorly Trump has done gaining voters as the field has narrowed, in a "normal" year he would be heading for the exit by now.
So, what are the implications for the GOP going forward? Trump is clearly the weakest Republican nomination leader since the two parties started using primaries to select nominees. If Trump is denied a majority of delegates, either by the party coalescing around one candidate very soon or if the remaining candidates are able to split up the winner-take-all primaries, then the smart move is to deny Trump the nomination.
The fear of Trump is backwards. Certainly he has the power to disrupt the Republican Party. However, keeping him as the nominee would be an even worse outcome. Trump has consistently polled worse than any of the Republicans against Clinton. When you dig into the primary numbers, it becomes obvious why. Most of the new voters are not Trump supporters – they likely melt away in a general election matchup. In addition, as his Republican opponents leave the race, their supporters move toward other candidates. The resolutely anti-Trump vote is far larger than the pro-Trump vote with both habitual and new voters.
Given the facts, the Republican Party is faced with an easy decision – but one that is guaranteed to cause a lot of heartburn. Throwing Trump overboard will test the eventual Republican nominee and his leadership team – any misstep could be fatal. But shrinking in the face of Trump’s aggression would be a far worse decision. The votes lost would far outnumber the votes gained.