The Endorsement Game

The Endorsement Game

If you were paying attention to the major party conventions held in the U.S. last month, you may have noticed a stark difference in the ‘quality’ of speakers at each event.

The 2016 Republican National Convention brought to light the struggles Donald Trump has had garnering support from key GOP figures. An overwhelming majority of former Republican presidents, presidential candidates and lawmakers opted out of the gathering at which everyone was supposed to rally behind Trump. Those who did make it there were not entirely helpful to the cause (looking at you, Ted Cruz). Meanwhile, his Democratic opponent has secured endorsements from across the two major parties as well as from big-name political outsiders. Even those ideologically adrift from Hillary Clinton, such as Michael Bloomberg and Elizabeth Warren, delivered rousing convention speeches on her behalf.

In this increasingly unusual election cycle, the rules for endorsing have changed. Many Republicans are hesitant to unreservedly back their own nominee, whereas others are finally speaking out about the choice they believe voters should make in November. Donald Trump’s harsh rhetoric and controversial proposals would initially lead us to assume that Republicans are merely withholding endorsements simply because he is regarded as a poor presidential pick. However, the factors potential endorsers must take into account extend well beyond his plans and policies.

The Republican’s Dilemma

Let’s begin our discussion in the Republican tent, where it’s fair to say that sitting politicians aren’t entirely sold on Trump.[i] Each has their own set of reasons why, whether it’s to do with his extreme rhetoric or his populist ideology. Nevertheless, a number of establishment-type Republicans, such as Speaker of the House Paul Ryan, have given Trump a tepid endorsement being fully aware of his pitfalls. Why would they do that?

To understand why Speaker Ryan and his colleagues may have backed Trump we must revisit the economic notion of self-interest, dictating that rational individuals will make decisions that best maximise their utility. For an elected official, utility could be determined by one’s rapport with colleagues, chances of re-election and prospects of achieving higher office e.g. the presidency. Thus, voicing any opinion on Trump will inevitably alter these determinants. Shown below is a payoff matrix, depicting a hypothetical game in which a prominent Republican politician (who dislikes Trump) must decide whether they will condemn the nominee, or endorse him regardless.


(Politician, Voters) Favour Trump Oppose Trump
Endorse -1, 5 -5, 0
Do not endorse -5, 5 0, 0


In this game, Player 1 represents our politician, stuck in a predicament as to whether they should endorse. For simplicity, assume that the politician’s utility is solely determined by their public approval. If their decision clashes with the prevailing view amongst voters, then the politician receives a large negative payoff of -5. We then define Player 2 as voters who are crucial to the politician’s re-election and prospects of achieving higher office, filtering out voters who have already made up their minds about the politician. Another somewhat realistic assumption made in this game is that voters are stubborn; they are unswayed by endorsements and will negatively judge politicians who associate themselves with the ‘wrong’ candidate. Bear in mind that in reality, there are bound to be other determinants affecting utility e.g. having to work with a potential Clinton administration. Player 2 would represent the complex amalgamation of such determinants.

The matrix suggests that the Republican politician’s utility would be maximised if they disavowed Trump and voters follow suit, in which case both the politician and voters would receive a payoff of zero. The politician gains nothing, but unfortunately the alternative strategy profiles all result in a negative payoff. Let’s say that from opinion polls, this politician predicts that the voters in his electorate actually intend to vote for Trump. Since the dominant strategy for voters is to favour Trump, the politician’s rational response is to endorse him, mitigating the loss of utility. This strategy profile corresponds to the top-left cell of values, awarding a payoff of -1 to the politician and a payoff of +5 to voters (for voting rationally). To put this in a real world context, suppose that Speaker Ryan dislikes Trump, but endorsed him anyway to maintain approval with voters (remember, not necessarily the main reason). This could be one example of a payoff matrix explaining the motive behind Ryan’s endorsement.

What about those Republicans who have shunned Trump, such as Ted Cruz? One possibility is that the Texas senator has gauged his voter base to generally be opposed to Trump, so distancing himself from this election would maximise his expected payoff e.g. aiding his future political aspirations. Of course, it is possible that Cruz and others are just behaving irrationally and against their self-interest, merely out of pure disdain towards the nominee.[ii] Lastly, retired politicians and ‘has-beens’ critical of Trump are getting an easier ride out of this mess! Republicans in this category are generally unfettered by voter perceptions and lacking in ambition. If that is the case, then they are likely to derive greater satisfaction from following their conscience and rejecting poor candidates, even if it does mean having Democrats in the White House for another four years. It could explain why many retired officials or unsuccessful presidential candidates such as Jeb Bush and Mitt Romney have chosen to speak out.

Endorsements for Hillary

Compare the turmoil and disunity on the Republican side to Hillary Clinton’s presidential run, which has accrued a remarkable number of high-profile endorsements in spite of high voter distrust and controversy. From progressives to centrists, support for Clinton has emerged from all factions of the Democratic Party. When faced with the prospect of a Trump presidency, prominent skeptics of the Democratic nominee are much more willing to fall in line. The biggest surprises, however, have come from the political right.

Take, for instance, recent endorsements made by fiscally conservative billionaires such as Michael Bloomberg and Meg Whitman. What motivated them to enthusiastically back a candidate who holds so many opposing views? If billionaires are acting out of genuine concern for the nation, then their public statements and comments on Trump address that question. But if we return to the idea of self-interest, remember that billionaires benefit considerably when markets are stable, and it is incredibly difficult to gauge how markets would behave under a Trump presidency.[iii] So in endorsing Hillary Clinton, billionaires have opted for a probable lower payoff (in the form of taxes and regulation) over Trump’s unpredictable antics.

At the end of the day, do endorsements even matter to the average voter? The question has been raised many times, and it feels like this is the year in which others’ opinions matter less than ever. Donald Trump has managed to seize the Republican nomination with little help from party figures, while a slew of endorsements for Hillary Clinton have not been incredibly effective in quelling wide distrust. As economic game theory suggests, making an ineffective endorsement may actually end up hurting the endorser’s approval with a stubborn voter base. In an election with such polarising nominees, ambitious politicians might want to think twice before opening their mouth.


[i] Graham, D.A. (2016, August). Where Republicans Stand on Donald Trump: A Cheat Sheet. The Atlantic. Retrieved from

[ii] Reuters. (2016, 22 July). Republican convention: Defiant Ted Cruz says he won’t be ‘servile puppy dog’ to Donald Trump. ABC News. Retrieved from

[iii] Chavez-Dreyfuss, G. & Randall, D. (2016, March 2). Wall Street’s big short: President Donald J. Trump. The Age. Retrieved from

Image: “Hillary Clinton – President Obama – Charlotte NC” by Scott Pelkey, licensed under CC BY-ND 2.0