Sabato's Crystal Ball

For Republicans, a 2016 Tie Is a Win

GOP virtually guaranteed to hold edge in event of deadlocked Electoral College

Kyle Kondik, Managing Editor, Sabato's Crystal Ball January 29th, 2015

Republican National Committee Chairman Reince Priebus, elected to a third term earlier this month, often notes that the presidential cycle is harder for his party than midterms because the electorate is more diverse and Democratic. “For us to win a presidential election, we have to be just about perfect, and the Democrats have to be good,” he told Kyle Cheney of Politico.

Priebus’s observation certainly squares with post-2012 commentary arguing Democrats might have an edge in the Electoral College, though like any good coach, the chairman is also trying to guard against overconfidence amongst his fellow Republicans.

The coming presidential election will provide an interesting test as to whether Democrats do have a durable Electoral College advantage, particularly if political science forecasting models suggest that the Republican nominee should win the election but he or she fails to do so.

That said, Republicans are not without at least one Electoral College edge in 2016. While the Democrat needs 270 electoral votes to win the presidency — a majority of the 538 available — the Republican only needs a 269-269 tie to win.

That’s because, in the event of a tie, the newly elected House will pick the president. Each state gets a single vote, so the party that controls a state’s House delegation will almost certainly cast its state’s vote for its candidate. (The Senate would select the vice president, and every senator gets a vote.)

Republicans dominate the states’ individual U.S. House delegations, and there’s no reason to suspect they will lose this advantage in 2016.

The GOP holds a majority of the House seats in 33 states, while Democrats only have a majority in 14. The remaining three states — Maine, New Jersey, and New Hampshire — are tied. Map 1 shows the percentage of seats controlled by each party in each state, and Table 1 provides a more detailed, state-by-state breakdown.

It’s no shock that the GOP has such a lopsided edge in state U.S. House delegations. After all, the party just won its biggest majority since 1928. But what is striking is that Republicans control the House delegations of nearly every presidential swing state.

The states where Democrats hold a majority of House seats all at least lean to the Democrats in presidential elections. Barack Obama easily won all 14 of these states in 2012. Of them, only Minnesota went for Obama by fewer than 10 points, and Mitt Romney never really contested it (Obama won the North Star State by close to eight points). The three states that are tied — Maine, New Hampshire, and New Jersey — have all backed Democratic presidential candidates in at least the last three elections.

Of the 33 Republican-controlled delegations, nine of them are in states Obama carried in 2012: Colorado, Florida, Iowa, Michigan, Nevada, Ohio, Pennsylvania, Virginia, and Wisconsin.

In many of these states, the Republican House edge is overwhelming. Take the trio of Florida, Ohio, and Virginia, which were Obama’s three closest Electoral College wins in 2012 and are arguably the three most vital swing states in the Electoral College. Republicans have a combined House edge in these states of 37-17. Throw in North Carolina (10-3 Republican), which was Obama’s closest loss, and the edge is 47-20. So in the only four states decided by less than five percentage points in the 2012 election — three won by Obama, one carried by Romney — Republicans control 70% of the House seats. And if you think this is just a function of the recent Republican-dominated midterm, think again: The GOP only netted one extra seat (NC-7) out of these states in 2014. The GOP’s swing state House edge was built in the 2010 midterm, reinforced in decennial redistricting conducted after that election, and solidified in 2012 and 2014.

The same is true in other swing states that are traditionally more Democratic, like Michigan (9-5 Republican), Pennsylvania (13-5), and Wisconsin (5-3). The Republicans didn’t net a single extra seat out of these states in 2014. Of the remaining three Obama states with GOP-controlled House delegations, Iowa and Nevada were tied prior to GOP gains in 2014, and Republicans have held a narrow 4-3 edge in Colorado’s delegation since 2010.

Just for the sake of argument, I looked at what would happen to the makeup of House delegations if Democrats had an incredibly strong Election Night. Under a very rosy scenario, would Democrats win a majority of House delegations?

Surprisingly, no.

Let’s say that Democrats in 2016 net even more than the 30 seats they need to flip control of the House: 36, or every single Republican-held seat we currently list as at least hypothetically competitive in our Crystal Ball ratings (in other words, every GOP seat we don’t rate as Safe Republican). Those seats are listed below.

Such an improbable victory would give the Democrats a 224-211 House edge: Not a big majority, but a majority nonetheless. And yet, Republicans would still control a bare majority of House delegations and retain tiebreaking power in the Electoral College.

Here’s the math:

Remember, Democrats are starting with 14 delegations. They would gain the three currently tied delegations in Maine, New Hampshire, and New Jersey. That gives them 17, to the Republicans’ 33. Then, Democrats flip seven additional delegations: Alaska, Arizona, Colorado, Iowa, Michigan, Nevada, and Virginia. The Democrats would then control 24 delegations, but the Republicans would still have 26 (there are no tied states assuming these results). In other words, even under a highly implausible pro-Democratic scenario, the party still would lack the votes to win a tied Electoral College.

Back to reality, and a vital caveat: A 269-269 tie would mean a very close presidential election nationally. That would rule out the Democrats getting the requisite national wave they would need to capture the House, let alone a majority of House delegations. If the House is in danger of flipping to the Democrats next year, it seems reasonable to assume that the Republican will handily lose the presidential race, rendering this discussion irrelevant.

This is all a long way of reiterating that the Republicans’ magic number to win the White House next year is 269, not 270.

Granted, this edge is almost certainly not going to matter. In our current Republican vs. Democratic political alignment, which dates back about a century and a half, the House has never decided a presidential election. There have been a few Electoral College misfires in this time: In 1888 and 2000, the winner of the popular vote lost the Electoral College. In 1876, Samuel Tilden (D) won the popular vote over Rutherford B. Hayes (R) and also seemed poised to win the Electoral College, but a dispute over the results led to a deal in which Hayes won in return for an effective end to post-Civil War Reconstruction. In none of these elections did the House have to pick a winner: 1800 and 1824 remain the only elections when the U.S. House selected the president after the Electoral College failed to produce a victor.

So there’s no modern precedent for a tied Electoral College, and the chances of it happening are beyond remote. However, one can visit our friends at 270ToWin.com and play around with the maps to create many plausible tie scenarios. Here’s one that’s not incredibly hard to imagine:

Map 2: 269-269 tie in Electoral College

This map assumes the GOP nominee wins all of the states Mitt Romney carried in 2012 while peeling off four 2012 Obama states: Florida, Iowa, Ohio, and Wisconsin.

Imagining the Republican winning Florida and Ohio is easy: Obama won the former by less than a point and the latter by about three points. These were his closest wins. Then we’re giving the GOP nominee Iowa and Wisconsin, two Midwestern swing states that are typically at least slightly more Democratic than the nation as a whole in presidential elections. These are heavily white states, though, and perhaps Democratic performance among working-class whites erodes even further, putting these states more in play for the GOP. This 269-all tie would represent something of a stall for the “emerging Democratic majority” thesis, because Democrats hold swing states with significant minority populations like Colorado, Nevada, and Virginia but can’t get over the finish line because of problems in the whiter Midwest.

However, and assuming a relatively uniform vote swing from 2012 to 2016 across the country, one would expect the GOP nominee to win Colorado and Virginia before Iowa and Wisconsin because Romney did better in the former group than the latter in 2012. Swapping out Iowa and Wisconsin for Colorado and Virginia on this map would deliver the Republicans an outright win.

Overall, an Electoral College tie is not likely. It’s very unlikely. But if a 269-269 split somehow occurs, it would not really be a tie. Rather, after much gnashing of teeth in the House and a national constitutional crash course, it would be a Republican victory.

Updated 2020 Reapportionment Projections

Geoffrey Skelley, Associate Editor, Sabato's Crystal Ball January 29th, 2015

At the end of 2014, the Census Bureau released its 2014 population estimates. This provides new fodder for political geeks when looking ahead to the 2020 census and its major political impact: congressional reapportionment.

Unfortunately, the Census Bureau doesn’t release state-level population projections anymore, just national estimates up to the year 2060, leaving it to others to extrapolate where the 50 states will be on April 1, 2020. The well-regarded firm Election Data Services — you may have seen some of their big election maps — put together a set of projections and found that, if trends continue, somewhere between 14 and 17 states may gain or lose U.S. House seats after the 2020 census, as shown below in Map 1.

Map 1: Potential reapportionment change projections for 2020

Source: Election Data Services

Based on EDS’s projections, California (+1), Colorado (+1), Florida (+1), North Carolina (+1), and Texas (+3) are in line to gain at least one seat. In some cases, EDS found that Arizona, Oregon, and/or Virginia may also pick up an additional seat.

On the other side of the ledger, Alabama (-1), Illinois (-1), Michigan (-1), Minnesota (-1), Ohio (-1), Pennsylvania (-1), Rhode Island (-1), and West Virginia (-1) are positioned to lose a seat, and in at least one projection New York also lost one.

These changes are a broad continuation of the same patterns we have seen over the last few decades. States in the South and West, mainly in the Sun Belt, continue to see larger increases in population relative to states in the Northeast and Midwest. As a result, successive reapportionments keep shifting more and more seats to Southern and Western states at the expense of Northeastern and Midwestern states, as a Crystal Ball analysis from last year explored.

The gains or losses are worth exploring. The biggest winner for the second straight cycle will probably be Texas, with projections showing it likely to gain three seats after it gained four following the 2010 census. Should it gain three, the Lone Star State will have 39 House members, consolidating its position as the second-largest state in the U.S. Looking further ahead, if it gains at least one more seat after the 2030 census — a seemingly likely outcome given recent trends — the country will have two states, California and Texas, with 40 or more representatives in Congress for the first time in its history, though the 435-seat House has only existed since 1910. Having two 40-plus states almost happened after the 1970 reapportionment but just missed: While California crossed the 40-seat threshold to 43 seats, New York fell from 41 to 39.

Until 2010, California had gained at least one seat in every reapportionment since it became a state in 1850, but the latest estimates suggest that it may again add a seat after 2020. In December 2014, the Census Bureau announced that Florida had surpassed New York to become the third-largest state in the union. It appears that after 2020, the size of the Sunshine State’s House delegation and its influence in the Electoral College will reflect that status as it will move ahead of New York by one, maybe two seats and electoral votes. The Empire State has actually lost seats in every reapportionment since 1950, falling all the way from 45 seats after 1940 to its current 27. It may well lose another seat after 2020, though it will probably be a borderline case.

A seeming outlier in the geographical trend is Alabama, which may lose a seat in 2020 despite being a Southern state. However, growth in the South has been very geographically inconsistent, with states like Florida and Texas accounting for most of the region’s gains over the last few decades while the rest of the South’s gains and losses have largely evened out. Elsewhere, the Midwest and Rust Belt look likely to continue their downward trend. Illinois, Michigan, Ohio, and Pennsylvania are all set to probably lose one seat each; the last time that quartet didn’t each lose at least one seat in reapportionment was 1970, when both Illinois and Michigan preserved the 1960 status quo.

The two smallest states in this discussion, Rhode Island and West Virginia, may see their House delegations shrink after 2020. In the initial apportionment of 1789, the Ocean State had only one seat, but after the 1790 census, it moved to two seats. Except for a 20-year stint with three seats from 1910 to 1930, it has had two seats for almost its entire history. However, the state looks in danger of dropping to the ranks of the one-seat states, and if it does, Rhode Island will become the eighth state with just one representative in the House, a new all-time high. Meanwhile, the Mountain State looks like a fairly safe bet to see its delegation fall from three members to just two: It is the only state that saw negative growth between 2010 and 2014, based on the Census Bureau estimates.

These changes will have many consequences. Last year, from projections based on the 2013 estimates, the Crystal Ball’s Kyle Kondik took a look at the future shape of the House, and most of those observations remain true in the face of new numbers one year later. Dave Wasserman of Cook Political Report recently explored (paywall) the House picture as well, including some hypothesizing about where new districts might go.

As for the impact on the Electoral College, the two major parties’ bedrock states, California for Democrats and Texas for Republicans, both stand to gain electoral votes, particularly the latter. Recently competitive states such as Colorado, Florida, and North Carolina are all positioned to gain, raising the stakes even further in those places. Meanwhile, the historically key swing state of Ohio will see its position further reduced, and if Virginia gains a seat, the two swing states will be closer in electoral vote strength (from Ohio having an 18-13 edge to just 17-14). Other potentially competitive states such as Michigan, Minnesota, and Pennsylvania are positioned to lose an electoral vote.

Obviously, the next census is still over five years away, so these estimates could still be off to some degree. But projections only get more certain as we get closer to 2020, and yet again there will be winners and losers in the reapportionment process. Changes in the level of representation for different states and in their relative strengths in the Electoral College will, as always, be politically consequential.

Obama’s Approval Rating and the Outlook for the 2016 Presidential Election

Alan I. Abramowitz, Senior Columnist, Sabato's Crystal Ball January 22nd, 2015

With the 2014 midterm election in the rearview mirror, the attention of pundits and political prognosticators has quickly shifted to the outlook for the 2016 presidential election. On the Democratic side, former Secretary of State, First Lady, and U.S. Sen. Hillary Clinton appears to be the prohibitive favorite to emerge as the nominee. On the Republican side, however, there is no clear frontrunner, and early maneuvering by prospective candidates has intensified with the announcement by former Florida Gov. Jeb Bush that he is seriously considering a run for the White House. In addition to Bush, several prominent current and former Republican officeholders have already signaled their interest in running, including 2012 Republican nominee Mitt Romney, former Arkansas Gov. Mike Huckabee, former Texas Gov. Rick Perry, Wisconsin Gov. Scott Walker, Texas Sen. Ted Cruz, and Kentucky Sen. Rand Paul.

While the identities of the 2016 Democratic and Republican candidates will not be known for more than a year, one thing that is already known is that there will not be an incumbent in the race because President Barack Obama will be completing his second term in the White House. Even though he will not be on the ballot, however, evidence from past open-seat presidential contests indicates that the public’s evaluation of Obama’s performance in office will have a substantial impact on the outcome of the election.

Figure 1 displays the relationship between the incumbent president’s approval rating in the final Gallup Poll before the election and the share of the major-party vote won by the candidate of the president’s party in the six open-seat elections since World War II: 1952, 1960, 1968, 1988, 2000, and 2008. The line shown on this graph is the regression line for predicting the incumbent party vote share from the incumbent president’s approval rating.

Figure 1: Vote for incumbent party candidate by incumbent president’s approval rating in open-seat presidential elections since World War II

Sources: Gallup and Dave Leip’s Election Atlas

Despite the small number of open-seat elections, a clear pattern emerges from the data: The fate of the incumbent party’s candidate is strongly influenced by the popularity of the outgoing president. In fact, the incumbent president’s approval rating explains over half of the variance in the vote share of his party’s nominee. All three candidates seeking to succeed presidents with approval ratings below 50% were defeated, and the two seeking to succeed presidents with approval ratings below 40% were decisively defeated. In contrast, two of the three candidates seeking to succeed presidents with approval ratings above 50% won the popular vote, although one of those candidates, Al Gore in 2000, ended up losing the electoral vote.

The data in Figure 1 indicate that while the incumbent president’s approval rating has a clear influence on the outcomes of open-seat presidential elections, the results are also influenced by the candidates and their campaigns. These factors almost certainly explain why the points for the 1952 and 1988 elections are far below and far above the regression line respectively. In 1952, Democratic nominee Adlai Stevenson was handicapped not only by a Democratic president, Harry Truman, with a 33% approval rating, but a Republican opponent, Dwight Eisenhower, who was a national hero after serving as Supreme Allied Commander in Europe during World War II. In contrast, the Republican nominee in 1988, Vice President George H.W. Bush, was aided not only by President Reagan’s 51% approval rating but by a Democratic opponent, Massachusetts Gov. Michael Dukakis, who ran what many political observers consider one of the most inept campaigns in modern presidential history.

Implications

While the outcome of the 2016 presidential election will depend on the popularity of the Democratic and Republican candidates and the quality of their campaigns, evidence from recent open seat elections indicates that the public’s evaluation of Obama’s performance will also have a substantial influence on the results. From that standpoint, the president’s rising approval rating in recent polls is good news for Hillary Clinton or whomever the Democratic Party eventually chooses as its nominee.

Table 1: Predicted Democratic share of major-party vote in 2016 by Obama approval rating prior to election

Source: Data compiled by author

In the three most recent weekly averages of the Gallup daily tracking poll, the president’s approval rating has stood at 46%. This represents an increase of about three points from his average approval rating over the previous several months. Other recent national polls have found a similar increase in Obama’s popularity, most likely reflecting improving economic conditions. The data in Table 1 show that based on the results of the six open-seat presidential elections since World War II, an increase of three points in the incumbent’s approval rating can be expected to increase his party’s vote share by about half a percentage point.

The data in Table 1 suggest that barring a dramatic change in President Obama’s approval rating in the next two years, the outcome of the 2016 presidential election is likely to be very close. If the president’s approval rating remains below 50% in the fall of 2016, the Republican nominee should be a slight favorite; if the president’s approval rating is above 50% by the fall of 2016, the Democratic nominee should be a slight favorite. However, the results of recent open-seat races also indicate that candidates and campaigns matter: The prediction of a close election could be upset if either party nominates an unusually appealing or unappealing candidate or runs an exceptionally effective or inept campaign.