On Thursday, New York City Mayor Bill de Blasio became the 23rd candidate and the third mayor competing for the Democratic presidential nomination. In recent posts, Hans Hassell, Jonathan Bernstein, and Vox’s Dylan Scott and Tara Golshan offer explanations for this rash of candidates. This post is on the next question: How could the Democrats shrink their pool of contenders? I suggest two approaches: party-centered and market-oriented.
But first: What’s wrong with 23 candidates?
Lost Senate candidates. Party insiders and observers have highlighted the opportunity cost: Several of these candidates (Steve Bullock, John Hickenlooper, Beto O’Rourke, Joaquín Castro) face long odds of winning the presidency but could be very competitive Senate challengers. The Democratic Party is worse off overall by having them swing and miss for the presidential nomination while decreasing Democrats’ prospects in 2020 Senate races.
Crowded “lanes.” The wide array of candidates may make it difficult for party members to coordinate on the best type of candidate. In 2016, there were several Republicans who were well-positioned to champion Reagan-style conservatism while following through on the Republican National Committee’s mandate to reach out to a more diverse coalition, including Marco Rubio, Jeb Bush, Ted Cruz, and John Kasich. Meanwhile, there was only a single candidate in the “virulently anti-immigrant” lane, which helped him consolidate a base of like-minded voters.
If, for the sake of argument, the best candidate for the Democrats is a center-left white male, that “lane” is crowded by Bullock, O’Rourke, Hickenlooper, Michael Bennet, Joe Biden, Jay Inslee Seth Moulton, Tim Ryan, and Eric Swalwell. Or if the Democrats think the best competition against Donald Trump would be a female candidate, there is still the coordination problem of choosing among Tulsi Gabbard, Kirsten Gillibrand, Kamala Harris, Amy Klobuchar, Elizabeth Warren, and Marianne Williamson.
Gimmick Campaigning. Finally, an overcrowded field creates incentives for candidates to engage in “spectacle” campaigning and debate tactics, such as simplistic policy proposals and reality TV-style confrontations.
Obviously, the Democratic National Committee (DNC) does not directly control who runs and who does not. But it does have some indirect influence by regulating who can participate in DNC-sanctioned presidential debates. While general election debates do not seem to influence voters very much, it is possible in the modern era for candidates to rise from the pack based on their debate appearances, as John Sides and Lynn Vavreck observed during the 2012 Republican cycle. The harder it is for candidates to qualify for Democratic debates, the more they might rethink their long-shot bids for office. How could the DNC best restrict debate participation?
The DNC’s current plan is to hold 12 debates with up to 20 candidates each. To qualify, candidates must register at least 1 percent in three national or early primary state polls or raise money from 65,000 donors, with priority going to candidates who meet both criteria.
One problem for these standards (as with the 2016 Republican debate criteria) is that public opinion polls are a terrible basis for allocating access. Early polls are primarily a measure of name recognition, so this criterion is vulnerable to an unqualified celebrity (say, a reality TV star) crashing a party debate. In particular, polls with conventional sample sizes (~1,000) typically have a margin of error around 4 percent and are thus incapable of telling the difference between zero percent and 1 percent support. Meanwhile, 65,000 donors represent 0.22 percent of all votes cast in the 2016 Democratic presidential nomination, and since there is no minimum donation size, it is a weak guarantee that a candidate can afford to run a viable campaign.
But, most importantly, these criteria outsource the role of party leadership in screening and selecting presidential candidates. As readers of this blog know well, a prominent political science study of presidential nominations, The Party Decides, posits that party elites effectively choose their party’s nominee by endorsing the candidate who can best unite the party coalition and win the general election.
Donald Trump’s success in the 2016 Republican primary was inconsistentwith the empirical predictions of the book, but his presidency has demonstrated the logic of the book: Absent effective party screening, the Republican Party is saddled with a leader who is likely to be devastating for the party’s long-term image. If the Democratic Party wishes to avoid making a similar mistake and can no longer rely on endorsement signals to coordinate on the candidate who is best for the party, the obvious solution is to directly involve party leaders in choosing the candidates who qualify for party-sanctioned debates.
How could it work?
1. Choose a set of Democratic elites — perhaps members of the DNC, or all qualified superdelegates (DNC members, governors, members of Congress, past party leaders).
2. Choose a limit on the number of debate participants — perhaps gradually decreasing from 10 to four.
3. Conduct secret ballots of Democratic elites to determine who they think are the best candidates. Note that there’s a range of voting rules the DNC could use:
a) Each voter could support a limited number of candidates (without pooling votes for a favorite candidate), e.g., a ballot that simply says, “Who are your top five presidential candidates?”
b) Borda count: Each voter could rank candidates from 1 to X, with the top candidate getting X votes and the Xth candidate getting 1 point. This is how reporters and coaches rank college football teams, with X = 25.
c) Ranked-choice voting: Again, voters rank candidates, but this time they get one vote, and last-place candidates are eliminated one at a time and their votes reallocated to remaining candidates.
d) One vote, one candidate.
If I were designing the process, I would vary the voting rules to become more restrictive as the process unfolds.
Polling Democratic leaders could benefit the party in several ways. A DNC poll would provide a clear basis for regulating access to the debate stage, based on the judgment of people with a lot of information and high stakes in the outcome to prioritize candidates who represent the party’s interest. Conversely, it would allow the DNC to screen out candidates who are underqualified or curiously Russia-adjacent.
Of course, this strategy would mean directly confronting the Sanders supporters who felt the 2016 nomination process was “rigged” and have fought to weaken the role of party leaders in screening presidential candidates or determining the outcome. What would be an alternative?
As Hans Hassell notes, several candidates seem to be running for president for the publicity that will help them audition for other jobs — Cabinet posts, TV anchor, reality TV guest. Meanwhile, the DNC lags far behind the RNC in fundraising this cycle. One solution for both problems is to raise the price of the DNC’s most valuable commodity: access to presidential debates. To paraphrase Rod Blagojevich, the DNC has these debate slots and they’re [very] golden, so the DNC should not give it up for nothing.
How can the DNC monetize access to the debates? The most straightforward mechanism would be, again, to reduce the debate slots to a more manageable number (five to 10) and then auction them off among Democratic candidates. The candidates who pledge the most campaign funds to the DNC get debate slots. The rest sit at home and rethink their choices.
It would be fair to argue that this system would put the Democratic donor class in the center of the selection process. This is a good reason to favor a superdelegate voting system instead. However, the DNC could offer candidates a “people power” path to a debate slot. The current debate eligibility requirements state that a candidate shall offer proof of 65,000 unique donors, but not that the candidate shares the donor list with the DNC. Instead, a campaign could sign a joint fundraising agreement with the DNC so that one penny of every contribution goes to the DNC, as well as the name and contact information of the donor. The DNC could credit a candidate some dollar value (say, $5) for every unique name they provide. By gaining direct access to candidate donor lists, the DNC can consolidate a database of potential donors and activists.
If this sounds preposterous, what would you say if I told you the DNC is already charging presidential candidates to use DNC resources? BuzzFeed reports that state parties and the DNC are charging candidates for access to party-owned voter databases:
To purchase the voter file in Iowa, the first caucus state, it’ll cost you $120,000. To buy it in New Hampshire, the first primary state in the nation: $100,000. And in the pair of key early-voting states that follow, Nevada and South Carolina, it’s another $100,000 — each. For voter file data covering the rest of the country, compiled by the Democratic National Committee in its own national voter file, campaigns must agree to a strict set of termsrequiring candidates to raise money for the party in emails, videos, and at events — on top of a $175,000 price tag.
Auctioning debate slots would serve the same purpose: charging candidates to make use of a party resource. The DNC could earmark all such funds for the DNC’s own 2020 campaign, or divide the funds between the national party committees (DNC, DCCC, DSCC, DLCC, DGA).
As the primary season unfolds, it seems likely that the Democratic Party will seek to focus attention on a subset of viable candidates. This post has described two strategies for selecting candidates who have the most support among party experts and/or make the biggest contributions to helping the Democratic Party succeed in the 2020 election.