Case Study 19.1: The 2016 Republican Primary as Institutional Moment

What Happened

On June 16, 2015, a New York real-estate developer and reality-television personality descended an escalator in his Manhattan tower and announced his candidacy for the Republican presidential nomination. Most of the country's political-science establishment, most of the Republican Party's elected officials, most of its donor class, and most of its intellectual movement-conservative ecosystem treated the announcement as either a publicity stunt or a vanity candidacy. Within a year, that candidate — Donald J. Trump — had won the Republican nomination, defeating sixteen other candidates including the brother of one former president (Jeb Bush), one sitting senator considered a movement-conservative front-runner (Marco Rubio), one libertarian-leaning senator with a serious primary infrastructure (Ted Cruz), one Wisconsin governor with movement-conservative establishment backing (Scott Walker, who dropped out early), and a former neurosurgeon with a passionate evangelical following (Ben Carson).

In November 2016, that same candidate won the presidency.

The 2016 Republican primary is, by a substantial margin, the most severe test that the "party decides" thesis has faced since Cohen, Karol, Noel, and Zaller published the book of that title in 2008. It is also a useful case study in the difference between the party-as-organization, the party-as-label, and the party-as-government — and in what happens when those three faces of a party come apart.

The "Party Decides" Thesis as the Pre-Trump Consensus

To understand why 2016 is shocking to political scientists, you need to understand the consensus that preceded it.

The McGovern-Fraser Commission of 1969–1972 transformed the Democratic Party's nomination process from a convention-centered system controlled by party leaders into a primary-centered system in which delegates were chosen by primary voters. The Republican Party adopted a substantially similar system over the next decade. By the 1980 cycle, both parties were nominating their candidates through state-by-state primary and caucus contests rather than through convention-floor brokering.

The conventional academic interpretation of this shift, in the 1970s and early 1980s, was that the parties had been democratized — that party-leader gatekeeping was over and that primary voters now decided. Cohen, Karol, Noel, and Zaller's The Party Decides (2008) argued, against this interpretation, that the gatekeeping had not actually ended. It had moved from the convention to the "invisible primary" — the year-long period before primary voting begins, in which candidates compete for endorsements from elected officials, key constituency leaders, and major donors. The candidates who consolidated invisible-primary support went on to win the actual primary contests; the candidates who failed to consolidate invisible-primary support failed in primary contests, often before voters had really paid attention. The party, in other words, still decided. It just decided through endorsements, donor coordination, and media-narrative cultivation rather than through convention votes.

The empirical case for the thesis was strong. Through 2008, the candidate with the most pre-primary endorsements in each open contested nomination — McCain in 2008, Bush in 2000, Dole in 1996, Bush in 1988, Bush in 1980, Carter in 1976 — had won the nomination. Even Obama's 2008 victory over Hillary Clinton was, on closer examination, supportive: Obama secured a significant majority of endorsements over the course of January and February 2008, after Iowa demonstrated his viability. The thesis worked.

By 2014, The Party Decides had become standard reading in graduate political-science programs. It was widely treated as the authoritative account of how American presidential nominations actually worked.

What Trump Did

Donald Trump entered the 2016 Republican race with essentially zero invisible-primary support. He had no endorsements from current Republican senators, governors, or House members. The Republican Party's major donor class was supporting Bush, Rubio, Walker, or Cruz. The movement-conservative intellectual establishment — National Review, The Weekly Standard, the American Enterprise Institute orbit — was uniformly opposed. The talk-radio ecosystem was split: Rush Limbaugh and Sean Hannity were sympathetic to Trump's themes but had not endorsed; Mark Levin and Glenn Beck were openly hostile.

By the standard "party decides" framework, Trump's nomination should have been impossible. He had no party-coordination support and no plausible path to assembling it. The framework predicted that as primary voting began, Republican voters would coalesce around the candidate the party had most clearly anointed — most likely Bush or Rubio.

That is not what happened.

Trump won New Hampshire decisively (35% to Kasich's 16%), then won South Carolina (33% to Rubio's 22% and Cruz's 22%), then Nevada, then most of Super Tuesday. Each victory generated more media coverage; each round of media coverage generated more polling support; each round of polling support generated more primary voting; each primary victory drove additional candidates to drop out, consolidating the anti-Trump vote behind a shrinking number of alternatives but also clarifying that Trump was the only candidate with a clear plurality. By March 2016, Trump had a substantial delegate lead. By May, after Indiana, Cruz dropped out and Trump was the presumptive nominee.

Throughout this process, the formal party organization tried — and failed — to stop him. The Republican National Committee chair, Reince Priebus, met privately with senior Republicans to discuss whether the convention rules could be changed to deny Trump the nomination. The "Never Trump" movement, including major donors and elected officials, ran ads in primary states. Senator Lindsey Graham (who later became one of Trump's most loyal Senate allies) called Trump unfit. Mitt Romney, the 2012 nominee, gave a speech in March 2016 explicitly calling on Republicans to vote for any other candidate. The institutional party organized, threw its institutional weight at the problem, and lost. Trump won the nomination on the first convention ballot.

What This Means for the "Party Decides" Thesis

There are at least three ways to interpret what happened in 2016.

Interpretation 1: The thesis is dead. The party did not decide. The voters did. Trump's nomination demonstrated that in the era of small-donor internet fundraising, free social-media reach, and a cable-news ecosystem that rewards attention regardless of source, candidates with sufficient name recognition and viral capacity can win nominations against the explicit wishes of the party establishment. The era when party insiders coordinated to select acceptable nominees is over. Whatever the post-1972 system used to be, it is now genuinely democratic — and that has costs as well as benefits.

Interpretation 2: The thesis was partially right but missed Trump's specific conditions. The party usually decides, but Trump exploited a specific configuration: a large primary field (16+ candidates) that prevented the party from coordinating around a single anti-Trump alternative early enough; a fractured party establishment (the Bush-Romney wing did not actually trust the Cruz-Tea-Party wing, and vice versa); a candidate with pre-existing celebrity that gave him the universal name recognition that usually requires party-coordinated promotion; and a media environment that rewarded Trump's incendiary statements with constant coverage. Under different conditions — a smaller field, a more unified establishment, a less famous outsider candidate — the party would have decided.

Interpretation 3: The thesis is dead in the open-field case but alive when there is an obvious establishment candidate. Compare 2016 (no clear establishment alternative) to 2020 Democratic (Joe Biden as the obvious establishment alternative). The 2020 case shows the thesis still operating; the 2016 case shows it failing when no clear alternative exists. Future cycles will likely show some primaries fitting each pattern.

The honest position, in 2026, is that we cannot yet definitively choose among these interpretations. Each is supported by some evidence; each is challenged by other evidence. What we can say is that the simple version of the "party decides" thesis — that party insiders effectively coordinate to pick acceptable nominees regardless of voter preference — does not survive 2016 unmodified.

Counter-Example: 2020 Democratic Primary

The 2020 Democratic primary, examined more briefly, supports a more conditional version of the thesis. Biden entered the primary with significant invisible-primary support but was widely seen as a weak campaigner. He finished fourth in Iowa and fifth in New Hampshire — a performance that would, under standard expectations, have ended his candidacy. The early front-runner narrative shifted to Bernie Sanders, who won New Hampshire and was tied with Biden after Nevada.

What happened in late February 2020 was a coordinated party intervention. Representative James Clyburn endorsed Biden three days before the South Carolina primary, providing crucial Black-voter signaling. Biden won South Carolina decisively. Within seventy-two hours, Pete Buttigieg, Amy Klobuchar, and Beto O'Rourke (who had previously dropped out) endorsed Biden. The field consolidated. Biden won most Super Tuesday states three days later. The primary was effectively over by March 17.

This was, in the language of the "party decides" framework, a textbook case of insider coordination. Endorsements arrived in a tight cluster, signaling the consolidated preference of the party establishment to primary voters. Voters responded. The party decided.

But — and this is the critical point — the party decided under specific conditions. Biden was a known quantity who had served as vice president under a popular Democratic president. The establishment was unified in its assessment that Sanders would be a weaker general-election candidate. There was a clear Schelling point (Biden) for coordination to converge on. None of these conditions held in 2016 Republican.

What This Case Shows

The 2016 Republican primary and the 2020 Democratic primary, taken together, suggest that:

1. The "party decides" thesis is not a universal rule. It is a conditional regularity that depends on the structure of the field, the unity of the establishment, the existence of a clear alternative, and the candidate's relationship to the party-as-organization.
2. Outsider candidates can defeat the party-as-organization under specific conditions — most importantly, fragmented fields and fragmented establishments.
3. The post-1972 primary system has not eliminated the party's role in nominations, but it has made that role contestable in ways the pre-1972 system did not allow.
4. Both parties in 2026 face a primary system in which their formal organizations are weaker than in any era since the McGovern-Fraser reforms — weaker not because of any specific failure but because of the structural shift toward small-donor internet fundraising, candidate-centered campaigns, and a media environment that rewards attention.

The next test will be the 2028 cycle, when both parties will hold open primaries. Whether the Republican Party reverts to establishment patterns or extends the Trump-era pattern, and whether the Democratic Party can maintain establishment coordination under a wider open field, are open empirical questions. The 2016 and 2020 cases give us frameworks for asking them but not yet answers.

Reading and Sources

The primary academic source is Cohen, Karol, Noel, and Zaller, The Party Decides (University of Chicago Press, 2008). For the 2016 cycle specifically, see the FiveThirtyEight endorsement primary tracker (archived on the site) and the Cohen et al. 2016 Vox piece "How a Donald Trump nomination would change presidential primary politics forever." For the 2020 Democratic primary, see Hans Noel, "The 2020 Democratic Primary," in American Political Science Review commentary and David Karol's analysis in the Monkey Cage. For the structural-shift argument about why the parties are weaker in the small-donor era, see Daniel Schlozman and Sam Rosenfeld's recent work on the "hollow parties."