Patrick Ruffini
Aug 28, 2015 · 4 min read

Most of the commentary around public polling right now centers around the fact that August and other early polls aren’t awfully predictive of who will win the nomination, especially in the Republican field. Just ask Presidents Rudy Giuliani, Fred Thompson, Wesley Clark, Rick Perry, or Hillary Clinton.

Part of this is simple common sense. The further out from the time first votes are cast, the less predictive polls will be.

But this doesn’t tell the full story. The progression of polling in wide-open nomination races is not simply one of entropy — of the inherent unpredictability of forecasting. In nomination races, there comes a point in time where events speed up, and the public begins to lock in on the winners of the early nominating contests, and ultimately, the eventual nominee.

In the current nominating contest, that time is not likely to come for at least three more months.


Common patterns repeat themselves in the last four open nomination contests (2004 Democratic, 2008 Republican and Democratic, and 2012 Republican) and it’s instructive to review their lessons, starting in 2004.

The 2004 Democratic Race

In early December 2003, early frontrunner John Kerry was faring poorly in the Democratic race for President. So poorly that he self-loaned himself $7 million from the Kerry-Heinz fortune to finance a last-minute push in Iowa and New Hamshire.

Kerry bottomed out in the polls in December. Then something happened. He caught on in Iowa, winning the caucuses with 38% to John Edwards’s 32% (also a late-surging candidate). It was over that night. As nomination contests go, 2004 was a cakewalk, with Kerry losing only a handful of contests from that point forward. The Democratic base’s 2003 fling with Howard Dean lasted almost until the end, with the former Vermont governor peaking in December.

The takeaway is not just that Kerry surged and Iowa mattered. It’s that early “momentum” meant almost nothing. Kerry was losing support through early December, and he managed to come back and win the nomination anyway.

The 2008 Democratic Race

Barack Obama’s turnaround in the 2008 Democratic primary was less dramatic than Kerry’s, but he recorded his lowest numbers of the primary season just three months before the Iowa caucuses in early October. (As a reminder, we are currently five months from Iowa.) His support steadily ticked up throughout the fall, but only after Iowa and subsequent contests did he actually match Hillary Clinton in national polling. If you were looking at who would win the Democratic primary in August and September, you would not only say that Hillary Clinton was the odds-on favorite, but that she was actually consolidating her lead over Obama.

The 2008 Republican Race

John McCain’s story has been told many times. The early frontrunner in 2008, who was then forced to lay off the majority of his staff in July 2007, seeing declining or flatlining poll numbers for the rest of the year, emerged Lazarus-like to win the New Hampshire primary and the nomination despite being outspent by both Rudy Giuliani and Mitt Romney.

When did it first become clear McCain was back? Arguably, he began to tick up in New Hampshire polling in November, but from the chart above, there was little evidence of national momentum at any point in 2007. McCain was also at his low point in early December, just one month before voting began. Immediately after the New Hampshire primary, McCain vaulted into the lead nationally.

It’s also important to look at the candidates who led but didn’t win — the potential analogues to Donald Trump today. Rudy Giuliani was a fairly strong frontrunner until just two months before the Iowa caucuses, but it took until November for his numbers to actually decline.

The 2012 Republican Race

As the immediate precedent for 2016, 2012 polling attracts the most attention and analysis today. The process of “discovery, scrutiny, and decline” that played out with numerous candidates in 2012 is most talked about as an omen for Trump’s eventual decline. You can watch this process play out in the bell curves for Perry, Cain, and Gingrich in the chart above.

Compared to the other cycles, though, 2012 looks like more of an outlier. Romney maintained his support throughout the primary cycle (in contrast to other early frontrunners who lost altitude). Like other eventual nominees, he saw an uptick in support in the last 30 days. Donald Trump’s support also looks different than Michele Bachmann, Perry, and Cain’s in that he’s already sustained his high numbers for longer.

As in other years, the race looked much different the moment the voting actually began, though it took longer for Romney’s lead to consolidate. Despite polling no higher than 20% throughout 2011, Romney still had no problem collecting more than 50% support from primary voters — like all other recent nominees who also could have been considered Trumpian “losers” heading into the first nominating contests.


What can we learn from this?

In no case did the eventual Republican or Democratic nominee exhibit strong momentum in the spring and summer the year before the election. In most cases, they lost support, and were either written off completely or had serious doubts raised about their ability to win. Candidates who appeared strong early in the process (Dean, Clark, Giuliani, Fred Thompson) only began losing support in the late fall or winter.

Crucially, the nominees’ turnarounds came as little as 30 days before the first nominating contests, and even then, these gains could have appeared uncertain and fleeting. But their gains in December were predictive of victories in January and February. Their numbers before November and December? Not as much.

Echelon Indicators

Data, trends, and analysis from the team at Echelon Insights

Patrick Ruffini

Written by

Polling/analytics. Digital ex. Co-Founder @EchelonInsights.

Echelon Indicators

Data, trends, and analysis from the team at Echelon Insights

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade