Are Swing States Disappearing?
In January, computer scientist Randy Olson published a piece called “The Shrinking Battleground.” In it, Olson argues that the number of presidential swing states has declined in recent decades, that this decline has adverse consequences (fewer states in play leads to lower turnout in those states and presidential candidates only spending time and money on a few swing states), and concludes that the president should be elected by national popular vote rather than by the Electoral College.
Olson is not the only one who believes the Electoral College should be abolished – a lot of smart people agree with him. Doing so, however, would be exceedingly legally and politically difficult – so for the foreseeable future we’re stuck with this political relic.
But if we take a slightly different perspective on this data, we can assuage some of the national popular vote advocates’ worries. Specifically, the presidential battleground may shift, so more states may become swing states in the future. Moreover, the success of state level candidates from the minority party in partisan “strongholds” shows that Republicans in blue states and Democrats in red states do not always end up wasting their votes on pre-decided elections.
The Shifting Battleground
Olson identified nine “swing states” and 41 partisan “strongholds” using the following map:
I have a few minor disagreements with how states are categorized (New Hampshire, Pennsylvania, Wisconsin, and Arizona may have more swing potential than this map communicates, and Indiana and New Mexico probably fit better in the partisan stronghold category), and the measure used (consecutive wins) does not take into account how close those wins are. But the map mostly gets the current situation right. The red and pink states tend to vote for Republicans, the gray states swing and the blue states favor Democrats.
But that’s just the current situation.
These graphs show how the partisan index (inspired by the Cook PVI) of the partisan strongholds (colors of points correspond roughly to colors on Olson’s map) has evolved since the end of World War II. We calculated the index by subtracting the Republican candidate’s share of the two-party national popular vote from the Republican’s share of the two-party state level popular vote for each state and election. A positive value indicates that the state voted more heavily for the Republican candidate than the nation as a whole, and a negative value indicates a relative advantage for the Democrat. Zero means that the state’s two-party vote was identical to the national two-party vote. This measure controls for wave elections, so it shows a smoother version of how partisan loyalties in these states have changed over time.
It’s easy to spot the trends in these graphs. The long-standing western Republican strongholds have become steadily more Republican since the end of WWII (the red graph). The more recently Republican South (along with Montana, Arizona and Missouri – shown in pink) was Democratic, hovered around the national average but then steadily moved toward Republicans. The Democratic states are a mix – some of them have been Democratic for a number of decades and others leaned Republican into the 1990s. But most of them are now much more Democratic than the national average.
The obvious takeaway from these graphs is that stronghold states have generally pulled away from the national average and that Olson is right – the battleground has shrunk in recent years. But there is another critical detail – states can also move into the battleground.
For most of the last 50 years (with the exception of 1976 and 1980, when southerner Jimmy Carter was on the ballot), Virginia leaned pretty hard to the right. But in the last few cycles, it has become a swing state.
New Hampshire also went from being a Republican stronghold in the 1970s and 1980s to a swing state in the 1990s and 2000s. The state leaned Democratic in the last few presidential elections, but important down-ticket races suggest that the state could be in play in 2016. Republican Kelly Ayotte won the state’s 2010 Senate election by 23.5 points and Republican Scott Brown held popular incumbent Democratic Sen. Jeanne Shaheen to a 3.25-point victory in 2014. Since 2010, candidates from both parties have won and lost both of the state’s House seats.
Minnesota leans further to the left than Virginia or New Hampshire. In 2012, Obama won there by a larger margin (7.69 points) than he did in New Hampshire (5.58 points) or Virginia (3.88 points), so it might be difficult for Republicans to push Minnesota into the battleground in 2016. That being said, the state’s partisan index has moved toward the center in recent years – showing that states can move toward the battleground as well as away from it.
There is reason to believe the map might expand in coming election cycles. Republicans lost the Electoral College by more than 100 votes in 2012 and by almost 200 in 2008. They have not done well against the Obama coalition, so they can be expected to try to expand the map. The Republican nominee may attempt to appeal to Hispanics and thus make more of the Southwest competitive. Alternatively, he or she could try to win more of the Rust Belt and Midwest. Either way, the Republican Party has an electoral incentive to craft its messages and political positions in a way that could pull blue states into the swing state column.
Democrats may try to expand the map as well. Without Barack Obama on the ballot, the Democratic nominee theoretically could make a credible attempt to regain rural white voters in the South (both Hillary Clinton and Jim Webb would certainly try) or simply reverse some of the recent movement of whites toward the GOP. It’s also possible that by appealing to these groups, Democrats drive other portions of their coalition toward the Republicans.
Finally, it’s important to note that even if party coalitions do not move in the coming cycles, wave elections can temporarily push partisan strongholds into swing territory. In 1980, Ronald Reagan’s (nearly 10 point) popular vote landslide victory over then-President Jimmy Carter pushed many reliably Democratic states – including Massachusetts, Alabama, Arkansas, Mississippi and Kentucky – into single-digit margins. So even if the two parties fail to create more swing states by changing their coalitions, national conditions (e.g., the state of the economy causes a wave election) could create close races in reliably partisan states. In other words, we are in no way stuck with the current number or set of swing states.
Olson focuses on presidential politics in his analysis. He writes “voting Democratic in Alaska is about as pointless as voting Republican in California, which is why so many voters don’t bother showing up to the polls in these staunchly polarized states.” He describes how presidential candidates focus their time and money almost exclusively on a few swing states. Olson seems to be right about lower turnout and presidential candidates only focusing on a few states. But his premise that Democrats in Alaska and Republicans in California are wasting their vote only holds if one assumes that presidential elections are the only ones that matter.
Alaska Democrats mattered in 2014, however, when Independent Bill Walker (whose running mate, Byron Mallot, is a Democrat) unseated Republican Gov. Sean Parnell. California Republicans mattered when Gov. Arnold Schwarzenegger won re-election in 2006. On the state level, minority party candidates (defined here as candidates from whichever party typically does not win the electoral votes of a given stronghold state) for the governorship win with some frequency.
This map uses Olson’s stronghold and swing state designations to show where a minority party gubernatorial candidate won anytime from the 2006 election to today. Most states hold gubernatorial elections every four years during the midterms, so (for most states) this map represents three elections – 2006, 2010 and 2014. Similarly, Olson puts any state that voted for the same presidential candidate in three (or more) elections in the “stronghold” category.
A majority of stronghold states recently elected a governor from the state’s minority party. And some of these states do not simply lean or tilt toward a party – Massachusetts voted for President Obama by more than 20 points, but Republican Charlie Baker won the governorship in 2014. Similarly, former Massachusetts Gov. Mitt Romney won Arkansas by more than 20 points, but Democratic Gov. Mike Beebe won re-election by a nearly 2-1 margin in 2010.
Minority party Senate candidates also have had some success in Senate elections. In the last three regular Senate elections (2010, 2012 and 2014), minority party Senate candidates won in ten stronghold states – New Hampshire, Missouri, Illinois, Wisconsin, Montana, North Dakota, Maine, West Virginia and Pennsylvania. Most of these states also elected a minority party governor in a recent cycle.
It’s important to note that these down-ticket races have significant consequences. Both parties have recently felt the importance of Senate election results. Senate Republicans, despite being in the minority for almost all of the Obama administration, were able to derail some of Obama’s top legislative priorities by filibustering (or simply threatening to do so). Similarly, when Republicans assumed control over the House in 2011, Senate Democrats made sure that many bills that the House passed never made it to the president’s desk. Gubernatorial elections matter too. State governments pass an enormous amount of legislation on nearly every hot-button domestic issue, and those laws often become the blueprint for federal laws.
So while minority party votes in stronghold states might be wasted in presidential elections, they often are not in senatorial and gubernatorial contests. These state level successes do not address two of the big problems Olson brought up – lower turnout and presidential candidates focusing almost exclusively on swing states – but it does show that these minority party voters are not always wasting their vote.