The conservative case for a Romney run: a Moneyball analysis


Now that Trump has sewn up the Republican nomination, should conservatives run a different candidate, like Mitt Romney, as a third-party candidate? Based on my analysis, yes. In addition to preserving their concept of the party, such a run actually improves the likelihood of a positive outcome (and decreases the likelihood of a disaster) for traditional Republicans.

This is a non-partisan analysis. The question I’m trying to answer is, if you are a traditional conservative, how would running a candidate like Mitt Romney affect the chances of a positive outcome for traditional conservatism in the presidency and the U.S. Congress? This analysis takes no stand on whether such an outcome would be good or bad for the country.

The baseline scenario: Clinton vs. Trump

While Bernie Sanders has not conceded, it’s overwhelmingly likely that Hillary Clinton will win the Democratic nomination. To compare scenarios, we need a baseline. The baseline two-party scenario is as follows:

  • Chance of a Clinton win: 72%. Based on projections by betting markets.
  • Chance of Democratic control of the Senate: 62%. Also based on betting markets.
  • Chance of Democratic control of House: very low. Too many seats to flip.

Using some reasonable estimates of the correlation between Senate races and the presidential race, this yields the following probability matrix.

Screen Shot 2016-05-10 at 8.41.19 AM

The three-party scenario: Clinton vs. Trump vs. Romney

I compared this to a potential scenario in which Mitt Romney runs as a third-party candidate. Conservative William Kristol recently met with Mitt Romney to discuss this possibility. While Romney says he will not run, conservatives dissatisfied with Trump may find another candidate that matches their values better than Trump does.

It’s easy to understand this. Conservatives face two undesirable outcomes. If Trump loses, their movement faces a divided future after having been dragged through the mud in the primaries. If Trump wins, they face the prospect of a president whose values on trade, taxes, and social issues are not aligned with conservativism. Rather than face a no-win scenario, they would like to show that more of the country favors traditional conservatism than Trump-type protectionism. A lot of traditional Republican contributors will stay on the sidelines if Trump and Clinton are the only choices. And conservatives worry about the effect of an unpredictable Trump run on congressional and state legislative races.

My model does not depend on Romney being the nominee. Here are my assumptions in the three-party scenario:

  • Chance of Clinton winning in the electoral college: 80%. A three-party election will divide those who previously voted Republican, increasing her chances.
  • Chance of Trump winning in the electoral college: 5%.
  • Chance of Romney winning in the electoral college: 5%
  • Chance of no one winning a majority in the electoral college: 10%. If no one wins in the electoral college, then the House of Representatives decides the election. The House presidential contest gets one vote per state delegation, not one vote per representative. There are currently 33 delegations with a Republican majority, 14 with a Democratic majority, and 3 that are tied. Because of this, Republicans have a significant advantage in the House. So if the election goes to the House, I estimate the following probabilities:
    • Clinton win: 5%
    • Trump win: 20%
    • Romney win: 75%

Based on these estimates, I project the following probabilities, including both electoral college wins and wins in the case where the election goes to the House of Representatives.

  • Clinton wins presidency: 81%
  • Trump wins presidency: 7%
  • Romney wins presidency: 13%

This cuts both ways. As you can see from the matrix below, the chances of a Clinton win and a Democratic Senate increase from 47% to 54%. But the chances of the conservative dream scenario — a Romney win and a Republican Senate — are 7%, up from zero. Another benefit for conservatives is that the chance of Trump winning the election go down by a factor of four, from 28% to 7%.

Screen Shot 2016-05-10 at 9.03.33 AM

Calculating the value of these scenarios to conservatives

In addition to, arguably, winning back the soul of their party, there’s a real value of any of these scenarios to conservatives. To estimate that value, I used a scale of power in government, where -10 would indicate complete liberal control and +10 would indicate complete conservative control (similar to my Moneyball analysis of the Supreme Court nomination). I assume that a Romney presidency would be much more valuable to conservatives than a Trump presidency. Here’s my matrix of values for the election outcomes (assuming Republican control of the House in all cases):

Screen Shot 2016-05-10 at 9.08.59 AM

With these values, the weighted average of outcomes is as follows:

2-party election: -2.89, leaning liberal.

3-party election: -2.41, somewhat less liberal.

The reason for the change is that while the chances of Hillary Clinton getting elected increase somewhat, there’s also a chance for Mitt Romney to be elected, which conservatives would value more.

[tweetthis]Election model shows a 3rd-party run boosts conservatives’ chances of retaining influence in US govt.[/tweetthis]

The break-even depends on your assessment of Hillary Clinton’s chances

My original estimate that there is a 72% chance of a Clinton win could change. At what level of Clinton electability does it make sense for the conservatives to run a third-party candidate? I modeled different two-party Clinton electability assumptions. This chart shows the result:

3-party breakeven

Remember that the more negative the value, the more liberal the result. This chart implies that if you believe Hillary Clinton has a 57% or higher chance of being elected in a two-party race, then the conservatives could improve their position with a third-party run. (I hold constant the 80% probability of a Clinton electoral college win, because adding a third party makes it far more certain that Clinton will win.) If you think Trump has an even chance to beat Clinton or a little better, you don’t want to mess around by adding a third-party run.

Want to try out my model for yourself? Click here and you can test out your own assumptions. Just enter any numbers you want in the green cells.

The qualitative value of a third-party run

Assume for a moment that conservatives run Mitt Romney or someone similarly conservative and respected. The benefits to the conservative cause go beyond this election. They will be able to:

  • Demonstrate the weakness of Trump’s arguments and command of facts in a way that the crop of Republicans in the primaries couldn’t.
  • Reassure donors of the future of the conservative wing of the Republican party.
  • Attack Hillary Clinton in a more principled and credible way than Trump can.

Of course, at the end of the election, the most likely outcome is the same as a two-party run — a divided Republican party that must redefine its values, especially to the older white men that Trump had attracted. But a third-party run might be the first necessary step to do that. And it might signal the beginning of a three-party system in American politics, which would be a refreshing change for everyone.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.