No, the Party can’t “Do Something” about Trump
The nature of US parties means that Trump more or less is the GOP at the moment, and hence the GOP will do nothing about Trump.
As Doug Mataconis noted in a post earlier today, there is talk about the RNC somehow removing Trump or, else, at least that party leaders would stage some sort of “intervention” to straighten out the campaign. But here’s the thing, people who are talking like this (just like those who thought that “the party” could somehow stop Trump from getting nominated going into the convention) are missing a fundamental fact: Trump is now the leader of the Republican Party for all practical purposes. Reince Priebus works for him, not the other way around (he does not literally pay his salary, but Trump can fire Priebus, while Priebus can in no way fire Trump–see, Wasserman-Schultz, Debbie if you need to bone up on the relative significant of party chairs versus candidates).
The structure of US parties is such that since the highest electoral prize is the presidency, parties are heavily shaped the pursuit of that office and by their nominees. This is what David Samuels and Matthew Shugart call the “presidentialization” of political parties. This is a function of separation of powers wherein the chief executive is elected separately from the legislature and where the survival in office of the two branches is likewise decoupled. That is: early elections are not possible and members of the legislature do not directly rely on the executive candidate for election and vice versa. This is to be contrasted with parliamentary systems wherein for a party’s to win the executive it must first win a majority of seats in the legislature (or be part of a coalition of parties that controls a majority of seats).* In these cases, the executive is an agent of the legislative party and serves at the pleasure of that party and can be removed if it no longer serves the interest of the party (or, they might resign as we saw with Prime Minister Cameron recently). One thing is for sure in parliamentary systems: parties are designed to win legislative elections only, which changes the way those parties behave (as well as, typically, how they control their labels, nominate candidates, etc.).
In presidential systems, however, the executive is not beholden to the legislative party in the same way. Indeed, unlike in parliamentary systems, a party can be in the clear minority in the legislature, but still hold the executive (e.g., the House is controlled by Republicans, but the White House is controlled by the Democrats). Indeed, the focus of parties in such systems is fundamentally focused on winning the executive branch.
Samuels and Shugart (in a 2006 conference paper)** describes how the executive candidates can deviate from what others in the party might want to do (it should sound rather familiar–bolded text mine):
Adverse selection problems result from the likelihood that agents possess hidden information. Individuals who seek leadership positions have incentives to overstate their experience and qualifications or misrepresent their true preferences, particularly if those preferences clash to some extent with the organization’s goals. By engaging in misrepresentation, upon appointment or election to the leadership post, agents can make Madison’s worst nightmare come true and pursue their own goals or their own vision of what the party’s “true” goals ought to be.
A party seeking to place its candidate in the chief executive’s office must select a leader who will be competitive in a national election contest. Yet the pool of candidates who can appeal to voters directly and the pool of candidates who can implement the collective “will” of the party organization may only weakly overlap. Thus parties may have to settle for suboptimal agents at the candidate-selection stage: the best potential agents from the party’s point of view may be incapable of winning a presidential election, while candidates who can win such an election may not share the party’s goals fully.
The problem of moral hazard arises from the possibility of hidden action. After the principal signs a “contract” with an agent, the principal cannot observe every action the agent takes, much less have perfect confidence that all those actions conform precisely to the terms of the contract. For a political party, the danger is that leaders might use their authority to advance their own personal goals rather work toward the principal’s collective goals.(19-20).
Now, Trump is an extreme example of this, but the broader point is still well illustrated: the ability of party elites to control the executive nominee is limited at best and may not always fully overlap with all the members of the party. The institutional party is therefore weaker in such a context than in a parliamentary system. It is why, for example, Paul Ryan is not the leader of the GOP at the moment even though he is the highest elected office-holder in the party. Ryan could more or less be viewed as leader of the party prior to the selection of a candidate, but ever since Donald Trump became the presumptive nominee, that role pretty much went away and will only return if the Republicans win the House but lose the White House in November. And one thing is for certain: the GOP as an institution worries far more about who its presidential candidate will be than it does about who the Speaker of the House will be.
The thing to remember is that the party orients itself, and its internal structure (such as primary and convention rules), to the selection and maintenance of the presidential candidate and once that person is chosen, the party is largely subservient to him (or her). The thing is, candidates (even bad ones) are not normally like Trump. The GOP is currently being hoisted on its presidentialized petard and so talk of the party leadership “doing something” assumes potential actions that are not designed into the system. The bottom line being that a presidential candidate (or elected president) can do severe damage to the party without the party, as an institution, being able to do all that much about it. The US system is especially susceptible since the primary process selects the candidate with very little influence by party elites (note that the reason the Democrats have superdelegates is an attempt to ameliorate this potentiality to some degree). In short one could say that “The Party” either is not in control or is, for all practical purposes, a create of Trump (and not the other way around).
Now, on the one hand, Trump may not fully represent all Republicans (as evidenced by those who have declared they will not vote for him) but he is the leader of the party at the moment and the notion that there is some entity called “The Republican Party” that is separable from Trump and can act against him is incorrect and represents a misunderstanding of how US parties function.
None of this is to absolve the Republican Party for Trump (the ways in which the party leaders help bring us Trump is a wholly different discussion). All I am addressing here is trying to demonstrate how the nature of our political institutions influence party behavior and development and how it limits what can be done by the party in regards to Trump.
Granted: the truly remarkable could happen in terms of some sort of internal party revolt against Trump, but this seems radically unlikely. A more likely move would be for an attempted mass defection of establishment Republican types to support Gary Johnson, but this is rather unlikely as well. More likely than not we will only see individual defections rather than something more dramatic.
*Minority government is also possible, but that does not undercut the basic point: even in those cases, legislative electoral victories are needed to win the executive.
**I’d quote the linked book, but it is at the office.