Skip to main content
ABC News
The Pennsylvania 18th Result Tells Us What Everything Has Been Telling Us For A While

If you skipped watching the results of the special election in Pennsylvania’s 18th Congressional District last night to catch an action movie or read a crime thriller instead, you picked the less exciting activity. Democrat Conor Lamb and Republican Rick Saccone ran neck and neck for most of the evening (at one point, they were separated by only 95 votes), and even The Associated Press went to bed without calling the race. As of 8 a.m. Wednesday, with 100 percent of precincts reporting, Lamb led by 641 votes, or 0.28 percentage points. And he and Democrats had claimed victory.

The exact margin will likely change,1 but it’s going to be very difficult for Saccone to make up that deficit. The only votes left to be counted are around 200 absentee ballots in Greene County (expected to be announced on Wednesday) as well as a handful of provisional and overseas ballots, which may take days to finalize. There may not even be 641 ballots left to count.

Nor is a recount likely to change the final result. Although it would be pretty easy for Republicans to request a recount should they want one, recounts typically don’t shift election margins by that much. That’s especially true in Pennsylvania, where most voting is done on electronic touchscreens; a recount would only reveal errors in the small population of paper ballots.

But as we’ve told you from the beginning, for those of us who don’t live in the Pennsylvania 18th, it doesn’t really matter who wins if what you’re mainly interested in is the 2018 midterms.2 The takeaway for November’s elections will be the same no matter whether Lamb wins by a fraction of a percentage point or Saccone wins by a fraction of a percentage point: Tuesday represented yet another huge Democratic overperformance in a Trump-era special election.

The Democratic overperformance in Pennsylvania 18

Special elections this cycle, by the constituency’s partisan lean and final vote margin

Year Date Seat Partisan Lean Vote margin Dem. Swing
2017 April 4 California 34th† D+69 D+87 +18
April 11 Kansas 4th R+29 R+6 +23
May 25 Montana At-Large R+21 R+6 +16
June 20 Georgia 6th R+9 R+4 +6
June 20 South Carolina 5th R+19 R+3 +16
Nov. 7 Utah 3rd R+35 R+32 +3
Dec. 12 Alabama U.S. Senate R+29 D+2 +31
2018 March 13 Pennsylvania 18th R+21 0* +22

Partisan lean is the average difference between how the constituency voted and how the country voted overall in the last two presidential elections, with 2016 weighted 75 percent and 2012 weighted 25 percent.

* Unofficial.

† Primary results used because both general-election candidates were from the same party. Primary included multiple Democratic candidates, and results show vote share for all Democratic candidates combined.

Source: Daily Kos Elections, secretaries of state

One special election is not enough to draw midterm conclusions from, but when it fits so clearly into an existing pattern, it’s safe to say something’s going on. Counting Tuesday’s result, Democratic candidates in federal special elections have now outperformed the normal partisan leans of their state or district by an average of 17 percentage points. In recent midterm election cycles, that number has tracked closely with the eventual national popular vote for the U.S. House.

Special elections have tended to predict midterm outcomes

Swing in average special congressional election from weighted presidential lean vs. the national House vote margin in the next midterm since 1994

Cycle Average Swing National House Margin Difference
2014 R+4 R+6 2
2010 R+6 R+7 1
2006 D+15 D+8 7
2002 0 R+5 5
1998 D+1 R+1 2
1994 R+9 R+7 2
Average 3

The swing compares the district’s Democratic lean in the previous two presidential elections vs. the special election result. The Democratic lean compares the district’s voting patterns to the nation. For each cycle, the previous presidential election is weighted 75 percent and the one before is weighted 25 percent.


In other words, the result in Pennsylvania 18 is just the latest indication that Republicans are in trouble. The best predictors of midterm strength we have — President Trump’s approval rating, the generic ballot, congressional retirements and special elections — all are saying the same thing. In fact, of those, special election results may suggest the rosiest future for Democrats.

Does the Pennsylvania 18 result tell us anything else about the national political environment? That gets a little tricky, and I’d be cautious. But the pattern of Lamb’s overperformance may conceal other clues. For example, it’s still up for debate which political map represents America’s true baseline: the Obama-era 2012 map or the Trump-remade 2016 map, under which the white, working-class Midwest turned red and well-educated suburbs trended more Democratic. Perhaps counterintuitively, the Pennsylvania special election suggests that the 2016 map has some staying power. Lamb outperformed Hillary Clinton’s margins by a nearly identical 19 points in white-collar Allegheny County and in Trump-loving, blue-collar Washington and Westmoreland counties.3 However, Lamb outperformed then-President Obama’s 2012 margin by 23 points in Allegheny while doing so by only 9 to 12 points in the district’s other three counties. Perhaps because Lamb assumed some socially conservative positions and cozied up to blue-collar workers, the Pennsylvania 18th is often portrayed as just another Midwestern working-class district. In fact, though, it is both wealthier and better-educated than the nation as a whole. As the numbers show, Lamb won this election not in “Trump country,” but in the Allegheny County suburbs.

Finally, Pennsylvania 18 was yet another example of a Democratic turnout advantage — the telltale sign of an enthusiasm gap — helping the party do better than expected in special elections. The number of votes Lamb received was 80 percent of the number of votes Clinton received in this district in 2016 — but Saccone got just 53 percent of the number of votes Trump got. Anecdotally, turnout was also much higher in Democratic precincts than it was in Republican precincts. But, crucially for Democrats, the results hinted that maybe that Democratic turnout advantage is transferable to the higher-turnout environment of a midterm election. The 228,177 votes cast in the special election exceeded the 214,912 votes the 18th District cast in Pennsylvania’s 2014 gubernatorial election. In other words, turnout on Tuesday was even higher than it was in the last midterm.

Now we’re digging pretty deep into the numbers, a depth at which you can fall victim to the trap of overinterpreting the results of one special election. The big-picture takeaway is this: If Democrats can win districts like Pennsylvania 18, they won’t need to stretch and scrape together a House majority. According to FiveThirtyEight’s partisan lean measure,4 118 Republican-held districts nationwide are less red than the Pennsylvania 18th is. Of course, the entire country is unlikely to shift 22 points to the left in November the way the 18th did — and the country may not even shift the full 17 points implied by the aggregate of special-election results — but Republicans should still be very worried.


  1. If you’re curious, the polls did pretty well in Pennsylvania 18. An average of the four polls taken of the race in March showed a narrow Lamb win, but one that was comfortably within the margin of error. That’s virtually how the race turned out. In fact, the polling error in Pennsylvania was smaller than it typically is for a House special election. Polls taken within three weeks of such races since 2004 have missed the final result by an average of about 5 points. This time around, the polling average gave Lamb about a 2-point lead in a race he appears to have won by about 0.3 points.

  2. OK, so maybe it matters a little. The result may buoy Democrats in other measures of midterm strength. For example, Lamb’s strong performance could convince other top Democratic recruits who were previously on the fence about running in tough districts to take the plunge. Conversely, Saccone’s poor showing could push more vulnerable Republican incumbents out the door.

  3. All references to county-level results in Pennsylvania 18 refer only to the portions of those counties within the 18th District.

  4. The average difference between how a district voted in the past two presidential elections and how the country voted overall, with 2016 results weighted 75 percent and 2012 results weighted 25 percent.

Nathaniel Rakich is a senior editor and senior elections analyst at FiveThirtyEight.