It’s Time to Move Away from Primaries
Part 2: On General Elections and the Flaw in our System
Last time, I looked at how primary elections can elevate unrepresentative or unfit candidates. Today, I look at the realities of general elections in the 21st century and the implications, in conjunction with our system of nominations.
General Election Voters are Rational
The letter after the candidate name on the ballot is a sort of great equalizer, for better or worse; most voters are party loyalists in the end, when the parties are so ideologically distinct. Furthermore, strong party loyalty is quite justifiable these days. Why? Because executive branch appointees — and the advice and consent of the Senate — are so important for policy.
Throughout much of the 20th century, Supreme Court justices could be rather unpredictable before getting to the bench. Dwight D. Eisenhower regretted his appointments of Earl Warren and William Brennan. Gerald Ford appointee John Paul Stevens and George H.W. Bush appointee David Souter both turned out to be fairly reliable liberal votes on the Court. Reagan appointees Sandra Day O’Connor and Anthony Kennedy proved moderate — though Kennedy is perhaps better described as heterodox.
Presidents have learned from experience, and Court appointees are now much more predictable. Eight justices have been appointed to the Court since David Souter. Seven of the eight have been utterly predictable on the most politically-charged cases, with only Chief Justice John Roberts sometimes offering surprises. (Roberts is fundamentally an institutionalist and seems to see “custodian of the legitimacy of the Court” as part of the job description of the Chief Justice.)
Moreover, the Court weighs in on critical questions constantly. Since Kennedy’s appointment, the Court has ruled on key issues like abortion (Planned Parenthood v. Casey), same-sex marriage (Obergfell v. Hodges), campaign finance (Citizens United v. FEC), gun control (District of Columbia v. Heller), voting rights (Shelby County v. Holder), assisted suicide (Gonzales v. Oregon), warrantless searches (Riley v. California), the death penalty (Baze v. Rees), eminent domain (Kelo v. City of New London), environmental regulation (Michigan v. Environmental Protection Agency) and the result of a presidential election (Bush v. Gore).
Imagine if the Court had ruled the opposite way on all of those cases. We would live in a very different society indeed.
Surely, most cases in the judiciary are not the hard, ideological ones, but those cases really do carry an outsized significance. Take this view from a 2005 speech during the Roberts confirmation hearings:
“The problem I face — a problem that has been voiced by some of my other colleagues, both those who are voting for Mr. Roberts and those who are voting against Mr. Roberts — is that while adherence to legal precedent and rules of statutory or constitutional construction will dispose of 95 percent of the cases that come before a court, so that both a Scalia and a Ginsburg will arrive at the same place most of the time on those 95 percent of the cases — what matters on the Supreme Court is those 5 percent of cases that are truly difficult. In those cases, adherence to precedent and rules of construction and interpretation will only get you through the 25th mile of the marathon. That last mile can only be determined on the basis of one’s deepest values, one’s core concerns, one’s broader perspectives on how the world works and the depth and breadth of one’s empathy.
In those 5 percent of hard cases, the constitutional text will not be directly on point. The language of the statute will not be perfectly clear. Legal process alone will not lead you to a rule of decision. In those circumstances, your decisions about whether affirmative action is an appropriate response to the history of discrimination in this country or whether a general right of privacy encompasses a more specific right of women to control their reproductive decisions or whether the commerce clause empowers Congress to speak on those issues of broad national concern that may be only tangentially related to what is easily defined as interstate commerce, whether a person who is disabled has the right to be accommodated so they can work alongside those who are nondisabled — in those difficult cases, the critical ingredient is supplied by what is in the judge’s heart.”
That, of course, was then-U.S. Sen. Barack Obama explaining why he was voting against Roberts’ nomination. Those “5 percent of hard cases” really are relevant for American life.
To understand this is to understand, if not excuse or justify, why so many conservatives either voted for Trump or try to play defense for him. When so much is on the line, or at least appears to be on the line, in presidential elections, when the likeliness of moderation and ideological heterodoxy — let alone lasting compromise — from the other side is so slim (gone are the days of a Richard Nixon or Bill Clinton bucking their respective hardcore party bases), those feeling under a sense of siege will make compromises they would never consider in normal times when choices were less stark.
Trump has taken advantage of his Senate majority by filling as many judicial vacancies as he can. Largely outsourcing his appointments to the Federalist Society, Trump has been engaged in a concerted effort to push the judiciary towards a more conservative makeup. Some conservatives (myself included) sarcastically mutter “But Gorsuch!” anytime Trump says or does something destructive, signifying that one Supreme Court appointment was not worth the rest of the trouble. But lower court decisions matter, too, and Trump has already made an impact there.
The power of the presidency goes beyond judicial appointments. A year of the Trump administration has demonstrated the day-to-day power of the presidency in simply administering the executive branch. If you’re a certain sort of Republican voter, you have gotten a lot merely by virtue of the decisions from the Oval, particularly if you ignore the tweets. Trump has focused intensely on deregulation. He has put social conservative priorities front-and-center in his administration. He has realigned the Department of Justice on multiple issues facing Supreme Court litigation, including voting rights, immigration and labor rights. Under his administration, Immigration and Customs Enforcement (ICE) has acted much more assertively in enforcing immigration law, presumably taking their cues from the public posture of the president. (The bureaucracy responds to incentives and would most likely act differently with a different president steering policy. The last thing the bureaucracy would want to do is invite a rebuke — and additional scrutiny — from a sitting president.)
None of these things required a change in law — which is good for conservative policy preferences, because outside of tax reform, the Republican Congress has been stuck in a sort of intra-party gridlock in the absence of competent presidential leadership and guidance. Some didn’t even require a change in administrative rule — just a shifting of emphasis. The crucial point is that these decisions are consequential, and a President Hillary Clinton would have done none of what is described above.
This all may be penny-wise and pound-foolish for conservatives, as these executive branch policy decisions are mostly mutable, and their Congressional majorities are in severe jeopardy at this stage. But for now, conservative voters who bit the bullet and voted Trump have a lot to crow about.
All of this is to say that for Republican voters in the general election, voting for Donald Trump was fundamentally justifiable on these terms: the judiciary and executive are extremely powerful and significant, and losing out on the chance to make those appointments, points of emphasis and rules would result in significant negative consequences. That Trump was boorish, or inexperienced or unpredictable — these things could be balanced against the perceived significance of those judicial appointments and executive authorities. (In the interest of full disclosure: I voted for an independent candidate in the 2016 election.)
The Vulnerability at the Heart of the Madisonian Scheme
Herein lies the problem: general election voters have a strong tendency towards voting for the candidate their party puts forward. The idea of a Democrat voting for Eisenhower or Nixon or a Republican voting FDR or LBJ because “their guy” is worse is long gone. The process by which the party puts forward candidates is deeply flawed and leads to sub-optimal — and sometimes utterly unfit — nominees.
Combine a dysfunctional nomination process and strong, ideological parties, and you get our present state: narrow factions are not checked by the majority principle, as Madison surmised. Instead, narrow factions can hijack the resources of broader organizations and punch well above their ideological weight class.
In situations where both parties fail at producing a broadly-acceptable nominee, you are destined to get unacceptable outcomes. Even in situations where only one party produces a broadly-acceptable nominee, an unacceptable outcome is possible.
Again, let’s be clear: a vote for Trump in the general election was a rational, justifiable course of action, notwithstanding the candidate’s manifest unfitness for office. But a vote for Trump in the primary was unjustifiable, full stop, and never should have been an option.
There are better ways to pick nominees. Below I’ll suggest a few:
- Restore some version of the old convention model.
Local and state parties used to be relevant, pseudo-independent entities. Local parties would select delegates to attend a state convention. The state convention would then select delegates to attend the national convention, and then the delegates at the national convention would exercise their discretion in selecting a nominee. This is indirect democracy as it is designed to work: voters are assigning the responsibility of coming up with the best candidate to representatives with an interest in victory.
This process still exists, vestigially, except that delegate selection is now driven by the primaries themselves, rather than by state conventions. In theory a contested convention would use the delegate-driven model by default; delegates would be “freed” to “vote their conscience” on any ballot after the first.
Were there issues with the old model? Undoubtedly. Candidates and their allies often had to make promises to get support from their rivals, and those promises were often patronage-driven. Reinstating the old model would require certain reforms to the process, particularly now that the spoils system has been swept away. But it beats the alternative because the likelihood of a narrow faction dictating the nomination is limited by the need to get consensus from the delegates selected. Such a model may well have gone hard right and selected Ted Cruz in 2016, but it surely would not have selected Trump.
2. Mandate alternate voting schemes.
American-style democracies almost exclusively use “first-past-the-post” voting method, where every vote counts once and the only relevant factor is which person got more votes. If primary elections are baked into the cake — and notwithstanding the thrust of this piece, they may well be — then at the very least the first-past-the-post model must be changed. Baseball statistician and polymath Bill James identified the key problem years ago. James wrote,
“… a badly designed voting system will fail sometimes, no matter who votes. The Gold Glove is decided by what could be called an unconstrained plurality, meaning:
1. A voter can vote for anybody.
2. If the top vote-getter gets 15% of the vote, he wins, the same as if he had received 80%…
A voting structure like this is an open invitation to an eccentric outcome. If the United States were to use a system like this to elect the President, the absolutely certain result would be that, within a few elections, someone like David Duke, Donald Trump, or Warren Beatty would be elected President. If you can win an election with 15% of the vote, sooner or later somebody will. An unconstrained plurality vote gives an opening to someone or something who has a strong appeal to a limited number of people.”
General elections are not decided in this way, but primary elections certainly are, and if partisan voting behavior is strong enough, the primary election result determines the general election winner. But there are other voting methods that we could use to mitigate against the present pathologies.
- Approval voting would allow voters to vote for as many candidates on a single ballot as they wanted. This would punish candidates that were offensive to certain sections of an electorate, as voters would simply leave that box unchecked.
- Instant runoff voting would allow for voters to rank their preferences, making it more likely that the eventual winner would be a candidate that everyone could tolerate.
- Cumulative voting would give voters more than one vote and would allow them to either split their votes among candidates or give all of their votes to one candidate.
If we must do first-past-the-post voting, and we must have primaries, then one alternative is to use a run-off election model, as Louisiana does if no candidate gets 50 percent of the vote in the first election. This would reduce the potential for a first-place finisher that succeeded in making a factional appeal that turned off everyone else.
Ultimately, any alternate voting model must change voter incentives: the winning candidate should be one that can broadly appeal to the various interests in the party. At present, the incentives cut the other way: factional candidates that can build intense support can overwhelm the broader party, even if a majority of the party would reject the selection if given an up-or-down vote.
3. Reconsider a version of the Cost/Anderson Plan
Back in 2013, Jay Cost and Jeffrey Anderson proposed a modernized version of the old system that incorporates alternate voting and candidate filtering. The plan sometimes seemed far-fetched, but its wisdom seems obvious in the aftermath of 2016: it essentially democratizes the old convention system by creating a “nominating convention” that would pick five possible choices for president, and then uses regional primaries and ranked-preference voting to determine the consensus nominee.
Apparently, Cost and Anderson presented their plan to the RNC and were summarily dismissed. But the two writers were onto something: the current process fosters bad choices and does not serve the interests of the party’s rank-and-file members. Instead, as they put it, “five groups” are the beneficiaries of the current system: elite donors, the media, a handful of state governments (Iowa, New Hampshire, South Carolina), campaign consultants and “low-information voters.” In short, it’s no way to run a railroad, much less a political party’s candidate selection process.
Enough is Enough
The 2016 election should have made clear that candidate selection is a substantial problem in modern American politics. This is particularly urgent for the Republicans, considering how many Senate seats they have surrendered to the Democrats because of bad nominees. But the Democrats should take heed, as well: their process also produced a bad nominee — and they seem to be removing the safeties by scuttling the superdelegates.
Right now, the primaries draw legitimacy from their appearance as a “democratic” element of our system, with coverage emphasizing the glory of the campaign trail, the nobility of the retail politicking required to break through in Iowa and New Hampshire and the special responsibility that those two states have assumed. In the sense that primaries allow the candidate that gets the most votes to win, they are democratic. But in a larger sense, they are not necessarily reflective of the public will, considering that they have such limited participation relative to the overall electorate. At its core, democracy is about having the people rule, but that’s not the same as voting on everything haphazardly. Elections are a means to an end, not the end itself.
The Founders rightfully were concerned with the balancing of different forces in government to ensure the protection of liberty and to prevent factional interests from overwhelming the national interest. We often seem to disregard their skepticism of unfettered democracy, and this is no more apparent than it is in our candidate nomination process. It’s time to reconsider.
Upon completion of the first draft of this piece, a Twitter user alerted me to an academic article from Stephen Gardbaum and Richard H. Pildes titled “Populism and Democratic Institutional Design: Methods of Selecting Candidates for Chief Executive in the United States and Other Democracies.” That piece offers some related observations and provides better historical context than what I have done here. In addition, Jay Cost of The Weekly Standard has written eloquently on this issue.
Dan Scotto has written for Ordinary Times and The Federalist. He currently lives in Oregon.
