Baseball fans learn early how to read the most important part of a box score. It’s a skill some N.C. electoral reformers have yet to master.
Like a baseball newcomer who pays more attention to the number of total hits or errors before turning to runs scored, the misguided reformer assigns too much importance to the wrong election numbers.
He focuses on statewide vote totals when discussing congressional or legislative elections. Emphasizing those totals can steer him toward bad ideas.
As simple as it sounds, a contested legislative or congressional election generally features two important numbers: the vote totals for the winning and losing candidates.
North Carolina has 120 seats in its state House of Representatives, another 50 in the state Senate, and 13 spots in its delegation to the U.S. House in Washington, D.C. Candidates win each of those seats by prevailing in an election contested once every two years.
This sounds like basic civics. It is. But too many observers lose sight of these 183 distinct electoral contests. They focus on statewide vote totals for state House, state Senate, and Congress that have no bearing on outcomes of individual elections.
The Nov. 22 Raleigh News and Observer offers one recent example. Among the experts quoted in a post-election analysis is Jonathan Cervas, a Carnegie Mellon University post-doctoral fellow “who has worked on court-ordered redistricting.”
“All told, more votes were cast for Democrats in the congressional races, but they still ended up in an 8-to-5 minority,” the N&O reported. “That’s closer than the current 10-3 split favoring Republicans. But Cervas said it should be even more equal based on the statewide results.”
One is most likely to reach this conclusion only by treating the state’s congressional elections as equivalent to 13 innings in one larger ballgame. Under this approach, a tally of statewide vote totals should have some impact on the overall split between Democrats and Republicans within the congressional delegation.
But that’s not the way this state’s congressional elections actually operate. Instead they consist of 13 separate contests pitting individual candidates in distinct geographic districts. Nowhere in the formal election process does the statewide congressional vote total play even a minor role. It’s a statistic that proves misleading at best.
It’s misleading because multiple factors within the 13 distinct electoral battles can skew the overall statewide vote.
Many electoral reform advocates like to point to just one factor: gerrymandering. They decry the role politics plays in the way state lawmakers draw the 13 district maps.
But gerrymandering explains just part of the picture. It does not account for the power of incumbency. Nor does it cover the major political parties’ prudent decisions to target scarce resources judiciously. Democrats and Republicans do not consider statewide vote totals when deciding how to promote their own candidates and attack their opponents.
The gerrymandering argument takes no account of pros and cons of individual candidates. There’s no role for gerrymandering, for example, when an unforeseen controversy arises.
Imagine that Cal Cunningham had pursued a U.S. House seat in 2020 rather than a position in the Senate. Cunningham’s infidelity scandal undoubtedly would have cost him votes, dropping Democrats’ statewide congressional vote total. But no one would argue that a dip in support for that single candidate would offer useful information about overall statewide support for Democratic or Republican policy priorities.
Beyond misleading. a misguided focus on statewide vote totals can prove counterproductive. It can lend support to proposed electoral reforms that miss the mark.
Cervas, the university researcher, is not alone in believing that the split in North Carolina’s congressional delegation “should be even more equal based on the statewide results.” Those who agree with him tend to tout redistricting changes that would yield outcomes more in line with a system of proportional representation.
In other words, a purple state like North Carolina ought to have a congressional delegation that’s likely to split 7-6 toward one party or the other.
But that’s a nearly impossible task.
If it were possible to draw 13 equally competitive congressional districts in North Carolina, the delegation would be more likely to swing wildly toward one party or the other as the political winds shifted. Spread one party’s 51% competitive edge over 13 districts, and it could win 13 congressional races — leaving the other party with nothing.
That’s unlikely. The state’s political geography could not produce a map with 13 toss-up districts.
More realistic is the proposal drafted in 2016 by a bipartisan group of retired N.C. Supreme Court justices and appellate judges. Their self-described “fair” election map featured six Republican-leaning districts, four Democratic districts, and three “swing” districts.
Under the judges’ map, Democrats might gain a 7-6 advantage in a good Democratic year. But normal electoral circumstances also could generate as much as a 9-4 Republican majority in a good GOP year. That’s a greater GOP advantage than the 8-5 split that causes Cervas such heartburn today.
Focusing on statewide vote totals turns reformers’ attention away from legitimate improvements to the redistricting process.
Rather than focusing on a statistic with little meaning, reformers should emphasize changes that pass a bipartisan smell test. Push compact, contiguous districts that avoid “irregular shapes.” Tout districts that follow well-defined political boundaries like county lines.
Force mapmakers to ignore any consideration of political data. That means voter registration and previous voting histories, as well as the home addresses of incumbents or likely challengers.
Once redistricting takes its proper form, then observers can turn back to the numbers that matter most in congressional elections: vote totals for the candidates vying to represent a single well-drawn, well-defined district.
Mitch Kokai is senior political analyst for the John Locke Foundation.