The Mini-Midterms: Five Takeaways from Six Decades of House Special Elections

Races often break against president’s party; winners rarely lose next election

Facebooktwitterredditpinterestmail
Dear Readers: Join us tonight (Thursday, April 15) at 6:30 p.m. eastern for a free, virtual panel: “The Changing Face of America: Voters of Color in the 2020 Election.” Theodore Johnson of the Brennan Center for Justice will moderate, and Andra Gillespie of Emory University, Mark Hugo Lopez of the Pew Research Center, and Natalie Masuoka of UCLA will participate as panelists. You can sign up for the event here, or just tune in tonight at this direct link.

This is the third of three University of Virginia Center for Politics panels leading up to the release of A Return to Normalcy? The 2020 Election That (Almost) Broke America — the University of Virginia Center for Politics’ new look at the 2020 presidential election and its consequences. The book is now available through UVA Bookstores, IndieBound, and other online booksellers.

Crystal Ball readers can also buy the book directly from the publisher, Rowman & Littlefield, and receive a 30% discount using the code RLFANDF30.

Edited by Crystal Ball editors Larry J. Sabato, Kyle Kondik, and J. Miles Coleman, A Return to Normalcy? brings together what Booklist calls a “stellar coterie of reporters, pundits, and scholars” to “parse the 2020 election via a data-driven set of analytics displayed in useful charts and graphs, drawing conclusions that will satisfy hard-core political junkies and provide a solid foundation for everyone looking ahead to 2022 and 2024.”

The previous two panels, “A Return to Normalcy? The 2020 Election That (Almost) Broke America” featuring Kyle Kondik, Theodore Johnson, Diana Owen, and Sean Trende, and “Taking Stock: The Societal Impact of the 2020 Election” featuring Alan Abramowitz, David Byler, Grace Panetta, and Madelaine Pisani, are both available at our YouTube channel, UVACFP.

— The Editors

KEY POINTS FROM THIS ARTICLE

— There have been nearly 300 U.S. House special elections since the mid-1950s.

— These elections more often flipped against the party that holds the White House — just like what often happens to the presidents party in midterm House elections — but the president’s party has scored some noteworthy wins, too, which can cloud the predictive value of special elections.

— Special election winners rarely lose their next election, but it does happen.

Six decades of special House election trends

Almost exactly 47 years ago — April 16, 1974 — Republicans suffered what would be the fourth of five U.S. House special election losses in the first half of that year. Bob Traxler (D), who would go on to serve two decades in the U.S. House, defeated James Sparling Jr. (R) in MI-8.

This happened despite — or perhaps, because of — embattled President Richard Nixon campaigning in person for Sparling days before the election.

The New York Times’ R.W. Apple Jr., who followed Nixon’s campaigning in the district covering Michigan’s thumb as well as the cities of Saginaw and Bay City, reported, “A Sparling aide confided this morning that he would be happy if the President never mentioned the local candidate’s name.” Alas, Apple reported, the president “mentioned Mr. Sparling constantly.”

Traxler won by three points in a district that Nixon had carried by 24 points in 1972 and 13 points in 1968. “It was yet another in the string of upset Democratic victories in special elections that showed Congressmen — even Republicans — how devastatingly unpopular were Richard Nixon and his works,” wrote the authors of the 1976 Almanac of American Politics. Another one of those Democratic victories, which came earlier in the year, was that of Richard Vander Veen (D) in the Grand Rapids-based MI-5, which Gerald Ford had left behind when he became vice president in 1973.

The Democratic victories in the first half of 1974 probably represent the most influential special House elections in recent history. The GOP losses “helped convince Republicans that Nixon needed to resign,” the authors of a more recent (2016) Almanac of American Politics wrote. Perhaps the only comparably influential set of special elections were those held between the 1930 midterm election and the opening of Congress in December 1931, which allowed Democrats to capture the majority during the Great Depression after narrowly coming up short in November 1930.

The 1974 specials also provide perhaps the best example of how special elections — themselves their own form of a “midterm” election — can operate as a miniature version of regular midterm House elections, which often deliver setbacks to the president’s party.

But they do not always operate this way, and special elections can sometimes be deceptive bellwethers for the approaching regular election.

With one House special election already in the books this year — Rep. Julia Letlow’s (R, LA-5) victory in the race to replace her late husband — and several more on the horizon, we thought it would be a good time to take a look at the modern history of special elections, dating back to 1957, the start of Dwight D. Eisenhower’s second term as president. We picked 1957 because we used Bloomberg congressional expert Greg Giroux’s excellent compilation of House special elections as our guide for this article, and Greg’s list goes back to that year.

We’ve identified five big-picture takeaways from this election history. But first, a few numbers:

There have been 289 House special elections since 1957. That includes Letlow’s victory last month, but it does not include a looming runoff between Democrats Troy Carter and Karen Carter Peterson in LA-2, coming up on April 24. This list also includes a couple of do-over elections, where the November results were wracked by problems: 2018’s NC-9 election, which was re-run in September 2019, and 1974’s LA-6 election, re-run in early January 1975.

Just to put in context how relatively few special elections there are, remember that there are 435 individual House elections every two years. So the whole history of special House elections since 1957 only adds up to roughly two-thirds of the number of races in a single, regular November House election.

There has been at least one House special election in every calendar year since 1957, with the exception of 2000. This averages out to roughly nine House special elections in every two-year election cycle, some of which end up being contested on the same day as a regular, biennial November federal election.

Despite occurring relatively infrequently, House specials are still a regular part of the election calendar, if irregularly scheduled. Article I, Section 2, clause 4 of the Constitution mandates that “When vacancies happen in the Representation from any State, the Executive Authority thereof shall issue Writs of Election to fill such Vacancies.” This is why every House vacancy is filled through a special election, and why governors have no power to appoint temporary replacements to the U.S. House (unlike with Senate vacancies, which are handled differently based on state laws and where governors often have appointment powers).

Some of the most prominent House members in the country have first come to Washington via special elections. That includes the top two House Democrats, Speaker of the House Nancy Pelosi (D, CA-12) and House Majority Leader Steny Hoyer (D, MD-5), as well as the second-ranking House Republican, House Minority Whip Steve Scalise (R, LA-1).

With that, here are our five takeaways from nearly 65 years of House special elections:

1. Special House elections more often break against the party in power

Just as the non-presidential party is likelier to net seats in midterms, so too is the non-presidential party likelier to capture House seats in special elections than the presidential party.

Of the 289 House specials since 1957, 55 were won by the party that did not hold the seat prior to the vacancy. Of these 55 flips, 39 were won by the non-presidential party.

The Democrats’ 1974 victories are an extreme example of this dynamic — five flips over the course of just four months — but the non-White House party enjoyed other longer, successful runs at various other times, too.

Democrats won three Republican-held seats in special elections over a half-year span in Nixon’s first year in office, 1969. And Republicans flipped six Democratic-held seats over the course of Jimmy Carter’s first three years in office.

2. They can be a preview of the upcoming November general election

In the case of 1974, the Republican losses ended up being a preview of the November campaign — even with Nixon gone, the Democrats still won 48 more seats in 1974 than they had won in 1972. In other words, the Democrats’ “special” strength manifested itself in the regular election.

This also ended up being the case, to a lesser extent, in the other aforementioned examples. Democrats netted a dozen seats in the 1970 cycle, Nixon’s first midterm, and Republicans netted 15 in 1978 and 34 in 1980.

There are other examples when special success proved to be a harbinger of things to come.

Republicans flipped two Democratic-held seats in Oklahoma and Kentucky in May 1994 — seats formerly held by long-tenured Democrats that George H.W. Bush had carried in his reelection loss to Bill Clinton in 1992. One of those victors, Rep. Frank Lucas (R, OK-3), is still in the House today. Following the Kentucky loss, the New York Times’ Richard Berke wrote, “The Kentucky results only add to the fears of Democrats that they will lose many seats, particularly those of several conservative Democrats in the South who are retiring this year.” Later that year, Republicans won a majority of Southern House seats for the first time since Reconstruction, a regional majority they have largely augmented over the past three decades.

More recently, Democrats flipped three Republican-held seats in the first half of 2008, including the Illinois seat of the former Speaker of the House, the now-disgraced Dennis Hastert, and two Republican-leaning seats in Louisiana and Mississippi. Democrats ended up augmenting their majority that November.

In a 2010 study covering House special elections conducted from 1900-2008, political scientists David R. Smith and Thomas L. Brunell found that when “one party takes seats away from the other party in special elections, the gaining party generally fares reasonably well in the general election.”

In the 2018 cycle, Democrats only flipped one Republican-held seat in a special election — PA-18, won by Rep. Conor Lamb, who now occupies a different, redrawn district, PA-17 — but that performance combined with a series of strong Democratic showings in heavily Republican districts that cycle attracted the attention of analysts. Writing in early 2018, Daniel Donner of the liberal elections site Daily Kos Elections used historical special election results from state and federal legislative races to identify a strong Democratic environment, which would manifest itself that November.

Unexpected close calls in seemingly safe seats have at other times provided a warning for the president’s party in advance of bad elections. In both 1981 and 2005, Ohio Republicans only barely held districts that, just the previous year, had each voted Republican for president by roughly 30 points apiece. Republicans suffered significant House losses in 1982 and 2006. So don’t just pay attention to the winners of these races — pay attention to the margins, too.

3. But sometimes specials are not a preview

History is also dotted with examples of special election results that break in favor of the White House party and/or do not reflect what would happen in the November election. As noted above, 39 of the 55 special election party flips were victories by the non-presidential party. But that also means that the presidential party flipped 16 seats from the opposition party.

Democrats flipped a Republican-held seat, NY-23, in November 2009, giving the party hope that it was a harbinger of 2010, as opposed to the twin losses the party suffered in the New Jersey and Virginia gubernatorial races on the same day (as it turned out, the gubernatorial races provided more of a preview of the GOP’s huge, 64-seat net House gain that cycle). Democrats also held difficult swing seats in New York and Pennsylvania during that special election cycle.

In the first half of 2004, Democrats flipped red seats in Kentucky and South Dakota; in 1963, Republicans flipped two Democratic seats in California. The string of Republican special election House flips during the Carter administration was ended by a Democratic victory in a GOP-held seat in Louisiana in May 1980. None of these results pointed to what happened in the 1964, 1980, and 2004 elections.

Sometimes the special elections provide mixed signals. In May 2011, now-Lt. Gov. Kathy Hochul (D-NY) flipped a usually Republican western New York district, boosting Democrats. Then, in September, Bob Turner (R) flipped a usually Democratic New York City district, boosting Republicans. The following November, Barack Obama won a competitive reelection while Republicans held the House. So perhaps the mixed 2011 special election results in New York did provide a preview of a mixed election, but that wasn’t apparent at the time.

4. Special election winners typically win their next election, but not always

In the wake of now-former Rep. Karen Handel’s (R) victory in the closely-watched 2017 GA-6 runoff, a number of prognosticators (including us) initially gave her the benefit of the doubt for the regular election based on the belief that special election winners usually win their next election. That is the case, but there are exceptions — as Handel herself ended up becoming when she lost to now-Rep. Lucy McBath (D, GA-6) in November 2018.

Just 20 of the 288 special election winners since 1957 lost their next House election, although some did not run again (Letlow is excluded from this group because she hasn’t had the chance to run for another term).

A few of these feature technicalities. In 1972, William Conover (R) won a special election for PA-27, a district that was being eliminated in redistricting. Conover lost a primary for another seat that same day. More recently, Shelley Sekula-Gibbs (R) won a November 2006 special election to win the remainder of former House Majority Leader Tom DeLay’s (R) unexpired term. However, she had to run as a write-in for the election to the term starting in 2007, and she lost to Nick Lampson (D). (Lampson would lose the then-heavily Republican seat to Republican Pete Olson in 2008.) In 2018, Brenda Jones (D) won a special election primary to a safe Democratic seat in Detroit, but lost the primary for the regular election on the same day to now-Rep. Rashida Tlaib (D, MI-13). Something similar happened to Neil Abercrombie (D) in Hawaii in 1986 — he won a special election the same day he lost a primary for the full term. He would return to the House in a 1990 election and later become governor of Hawaii.

So these four didn’t really lose their next election — rather, they lost at the same time they were winning.

Some of the special election winners whose victories were suggestive of the November results ended up losing despite the good political environment for their party. Tom Luken (D) was one of the five 1974 special winners, but he lost a rematch with Bill Gradison (R) in the Cincinnati-based OH-1 later that year (Luken won a different Cincinnati-based seat in 1976).

Dan Cazayoux (D), a surprising 2008 victor in the LA-6 special referenced above, lost that November to now-Sen. Bill Cassidy (R-LA). Louisiana was not using its familiar, top-two jungle primary system that year, and Cazayoux was likely hurt by a Black state representative running as a third party Democratic alternative who won 12% of the vote (Cazayoux lost by eight).

Hochul, the 2011 special election winner noted at the end of the last section, is another special election victor who lost her next election, although her district became more Republican in redistricting and she only lost narrowly. Turner, the other New York 2011 special election winner, saw his district eliminated altogether — instead he sought, and lost, the 2012 Republican U.S. Senate nomination in New York.

The last person to flip a seat in a special election, Rep. Mike Garcia (R, CA-25), would be on this list were it not for his narrow, 333-vote victory last November following a 10-point win in a May 2020 special election. It’s a credit to Garcia that he isn’t on this list — Joe Biden won his district by 10 points in November, making Garcia a major overachiever — but it does suggest, along with other examples in this section, that special election winners are not necessarily untouchable in their next election.

In 2009, Scott Murphy (D) somewhat surprisingly held the swing seat now-Sen. Kirsten Gillibrand (D-NY) left behind after she was appointed to the Senate to replace Hillary Clinton. But he ended up losing in the poor Democratic environment of 2010. The same thing happened to Peter Barca (D), who narrowly held WI-1 against Mark Neumann (R) in a 1993 special only to lose another close race to Neumann in the 1994 GOP wave. Four years later, Neumann unsuccessfully ran for Senate — allowing his southeast Wisconsin seat to be won by Paul Ryan (R), the future House speaker.

Ron Paul — yes, that Ron Paul — won his first U.S House victory in a 1976 special, defeating Bob Gammage (D). Gammage came back and beat Paul in the 1976 general election, and then Paul beat him in November 1978.

Paul served through 1984, when he ran for Senate and lost the primary to Phil Gramm, who had won a special House election a year earlier to fill a vacancy created by… Gramm’s own resignation.

Gramm resigned after switching parties, and won his seat back under his new Republican label (Albert Watson of South Carolina did the same thing in 1965 — both Gramm and Watson’s victories are counted as two of the 55 overall party flips in this era).

Paul’s first retirement from the House — he would come back in the 1996 election and serve through 2012 — opened TX-22 to be won by the aforementioned Tom DeLay.

One positive sign for Republicans in advance of their 2010 national victory was Charles Djou’s (R) victory in the HI-1 special, although he was aided by an all-party special election format that split the Democratic vote in a heavily Democratic seat. Djou lost in November. Something similar happened in the heavily Democratic NM-3 in 1997, in which Bill Redmond (R) won in part because a sizable chunk of the Democratic vote broke toward a Green Party candidate. Redmond lost to future Sen. Tom Udall (D-NM) in November 1998.

One other special election winner who immediately lost his next election: the “kissing congressman,” Vance McAllister (R). He won a 2013 LA-5 special election, but then the married congressman was seen on video kissing a staffer. He initially said he wouldn’t run for a full term, but he backtracked, finishing fourth in the November 2014 all-party jungle primary.

This is the same seat that the most recent special election winner, Julia Letlow, won last month.

5. Special elections can keep seats “in the family”

Speaking of Letlow, she is just the latest widow to successfully hold a seat left behind by her husband. Luke Letlow (R) died after contracting COVID-19 in December before he could be sworn into his first term in the House.

Many of those widows opted not to run again, although Julia Letlow plans to in 2022 after her impressive victory last month.

Some widows who won special elections went on to long House careers in their own right. Perhaps the most famous is Rep. Lindy Boggs (D), who served for close to 20 years after winning a 1973 special election in a New Orleans-based seat to replace her husband, House Majority Leader Hale Boggs (D), who was presumed dead in an apparent plane crash. Rep. Nick Begich (D) of Alaska was onboard the same plane: His presumed death prompted a 1973 special won by Don Young (R, AK-AL), who is still in the House and is its longest-serving current member. A separate plane crash also took the life of Chicago Rep. George Collins (D) in 1972; his wife, Cardiss Collins (D), won the special election to replace him and served in the House for nearly a quarter century. Rep. Doris Matsui (D, CA-6) won a special election to replace her late husband, Bob Matsui (D), in 2005, and she remains in the House.

House seats sometimes pass to family members other than spouses in special elections: for instance, current Rep. Donald M. Payne Jr. (D, NJ-10) won a 2012 special to replace his late father. Another current House member, André Carson (D, IN-7), won a special election in 2008 to replace his late grandmother, Julia Carson (D). Mo Udall (D-AZ) served for three decades in the House after winning a special election to replace Stewart Udall, his brother, who had become Secretary of the Interior.

Overall, a little more than two-dozen of the special election winners since 1957 had some family relation to the representative whose departure necessitated the election.

Conclusion

As we look ahead to 2021’s House specials, we know that Democrats will hold the heavily Democratic LA-2, where two Democrats are competing in a runoff later this month. The eventual Democratic nominee will also hold OH-11 in Northeast Ohio this November. There will be a special election at some point to replace the late Rep. Alcee Hastings (D, FL-20), who died last week. But his seat is also overwhelmingly Democratic.

Potentially more interesting are races in TX-6 in the Dallas-Ft. Worth Metroplex, where an all-party primary featuring nearly two-dozen candidates is scheduled for May 1, and Albuquerque’s NM-1, where party-selected nominees state Rep. Melanie Stansbury (D) and state Sen. Mark Moores (R) will face off on June 1.

These are both races that the incumbent party — Republicans in TX-6 and Democrats in NM-1 — should hold. TX-6 only voted for Donald Trump for president by three points, but it is more Republican down the ballot. One of the top Republican candidates is Susan Wright, the widow of the late Rep. Ron Wright, whose death due to complications from COVID-19 led to the vacancy. So Susan Wright could be the latest widow to win a House special election. Meanwhile, NM-1 voted for Joe Biden by 23 points, but the seat’s previous occupant, now-Secretary of the Interior Deb Haaland (D), won by a more modest 16 points in 2020, and Democrats performed well in some similarly inhospitable districts four years ago. Can Republicans do so now with the burden of holding the White House now passed to Democrats?

We’ll be watching to see if any trends emerge from these races, but there are confounding factors in both. The all-party format of TX-6 means we are not guaranteed a Democrat vs. Republican runoff, although a runoff seems almost certain given that no one is likely to eclipse 50% support in the first round of voting. In NM-1, former state Land Commissioner Aubrey Dunn is running as an independent, and the former Republican could siphon conservative votes from Moores, making it hard for him to win even if he runs a great campaign.

Speaking to Roll Call in advance of the two 2004 specials that Democrats would ultimately win — but that did not end up being a sign of better things to come in November — 2002 cycle Democratic Congressional Campaign Committee Executive Director Howard Wolfson said “Special elections can be predictive, but you don’t know that until the regular November elections.”

It’s advice worth keeping in mind. Special House elections that break in favor of one party, particularly the party that doesn’t hold the White House, can provide signs for the general election, but not always.

Crystal Ball interns Nik Popli, Kristen Sink, and Victoria Spiotto assisted with this article.