Moneyball, meet politics: Could VAR settle arguments about candidate strength? – Roll Call

In the era of data and metrics and models in political analysis, at least one question still remains: How do we quantify the strength of individual candidates?

Arguing over whether a candidate or incumbent is good or bad is an age-old tradition in the political media and among party operatives. Typically, candidate strength is measured by fundraising or the margin of a win or loss. But that can fail to account for the particular election cycle or the possibility that any candidate running on a particular party’s line in a particular year or state would do just as well.

But with some inspiration from baseball statistics, there is a way to quantify candidate strength and the value of an incumbent.

Before it’s possible to analyze individual performance, however, there must be an established baseline for generic partisan performance to compare against an individual candidate’s performance. In other words, answer the question of what would happen to any candidate with that party’s label.

Inside Elections’ Baseline metric captures a clear picture of a congressional district or state’s political performance by accounting for all federal and state executive election results. Baseline calculates the trimmed mean (the average after throwing out the highest and lowest results) of each party’s performance in partisan, contested statewide elections. It includes the four most recent general-election cycles in order to prevent a single election from dominating the metric.

Baseline party strength

For example, Florida has a reputation for being a perennial battleground, and the Republican baseline advantage over Democrats is 51.4 percent to 46.92 percent. Wisconsin is living up to its name as the top battleground, considering Republicans had a narrow 48.89 percent to 48.84 percent Baseline advantage.

Baseline means a typical Republican candidate running statewide should receive 48.89 percent in the Badger State after the 2018 elections, while a typical Democrat should receive 48.84 percent.

The second statistic, Vote Above Replacement (VAR), measures the strength of a political candidate relative to a typical candidate from their party within the same state. It’s a political version of baseball’s Wins Above Replacement (WAR), which measures a player’s value to a team compared to a replacement level player.

VAR calculates the difference between a candidate’s vote share in an election and the corresponding Baseline performance for that candidate’s party.

Klobuchar buries Warren

In the presidential race, DFL Sen. Amy Klobuchar of Minnesota has staked most of the rationale for her candidacy on being the most electable candidate in the field because she’s won three Senate elections, often carrying places where other Democrats have struggled. Indeed, her VAR score of 8.08 percent proves her point.

In comparison, Massachusetts Sen. Elizabeth Warren’s VAR is 0.37 percent, which means she performed marginally better than a typical Democrat in her most recent reelection race. Calculating a score for Vermont Sen. Bernie Sander is a little complicated considering he runs as an independent. But he was also the Democratic Senate nominee in 2018 and would have a VAR score of 8 percent.

Clearly, presidential primary voters are not prioritizing prior electoral performance. Montana Gov. Steve Bullock’s VAR is 6.05 percent, for example, yet he’s struggled to gain traction in a field of more than a dozen Democrats. Former Texas Rep. Beto O’Rourke’s VAR was 7.22 percent after his 2018 Senate run — and he dropped out of the Senate race last week.

To round out the field, Colorado Sen. Michael Bennet’s VAR is 2.98 percent and New Jersey Sen. Cory Booker’s VAR is 2.05 percent. California Sen. Kamala Harris doesn’t have a VAR score because she faced a fellow Democrat, Loretta Sanchez, in the 2016 general election.

In the Senate, VAR scores can shed some light on some hot races and key candidates running in 2020.

Maine Republican Susan Collins, who is up for re-election, has the highest VAR score of any senator. Her 25.24 percent VAR is more than 10 points ahead of any of her colleagues and is key evidence in the case for her to seek a fifth term in a state where she’s fighting partisan headwinds. If she decides to retire, it would send shockwaves throughout the country and put the GOP majority in peril.

In comparison, West Virginia Sen. Joe Manchin has a reputation for having one of the best personal brands, but his VAR is a modest 4.17 percent.

Republicans are high on John James’ chances of defeating Democratic Sen. Gary Peters in Michigan. Even though James is not an incumbent, he has a VAR score because he ran for the Senate in 2018 against Democrat Debbie Stabenow. James entered the 2020 race with high expectations, even though his -0.65 VAR demonstrates that he underperformed a typical statewide Republican in Michigan a year ago. In comparison, Peters’ VAR is 5.27.

Bullock vs. Daines

Armed with that 6.05 VAR, Bullock is considered a Democratic dream challenger to Republican Steve Daines in Montana. But the GOP senator’s VAR is 8.54, so he won’t be easy to defeat.

Colorado Sen. Cory Gardner (0.63 percent) and Iowa Sen, Joni Ernst (0.97 percent) have reputations for being extraordinary candidates, yet both have modest VAR scores. Gardner is likely to face former Democratic Gov. John Hickenlooper, who has a VAR of 1.61 percent. North Carolina Sen. Thom Tillis (0.55 percent) is about the same and Majority Leader Mitch McConnell’s is a bit better (2.63 percent).

Two of the most vulnerable GOP senators up for reelection this cycle with low VAR scores include David Perdue (-2.35 percent) in Georgia and Martha McSally (-3.94 percent) in Arizona. McSally, of course, lost her 2018 bid for Senate, then was appointed to the vacancy created by Sen. John McCain’s death.

Democrat Doug Jones has an impressive of 10.1 VAR after his special election victory, but he has to run for a full term in Alabama, where Republicans have more than a 20-point Baseline advantage (60.2 percent to 39.5 percent).

So even though Jones has shown an ability to overperform a typical Democratic candidate in Alabama, his VAR score might not be enough to overcome Republicans’ significant partisan Baseline advantage in the state over the last four election cycles.