Election forecasters generally have a short-term time perspective and that is perfectly understandable. We are concerned with predicting the outcome of one election at a time. In making our forecasts, we therefore focus on factors that can contribute to an accurate prediction of each party’s share of the vote in a particular election—factors such as the incumbent president’s approval rating (Abramowitz Reference Abramowitz2012), the performance of the economy in the year of the election (Abramowitz Reference Abramowitz2013; Campbell Reference Campbell2012; Erikson and Wlezien Reference Erikson and Wlezien2012; Holbrook Reference Holbrook2012; Lewis-Beck and Tien Reference Lewis-Beck and Tien2012; Lockerbie Reference Lockerbie2012;), the length of time that the president’s party has held the White House (Abramowitz Reference Abramowitz2012), the results of trial heats between the major party candidates (Campbell Reference Campbell2012) or the performance of the candidates in their own party’s primary elections (Norpoth and Bednarczuk Reference Norpoth and Bednarczuk2012). However, by focusing on individual elections and short-term forces, forecasters may be missing long-term trends that are profoundly altering the nature of electoral competition in the United States.
During the past four decades, the American electorate has changed dramatically. As a result of these changes, we have entered a new and fundamentally different era of competition in presidential elections. The four key features of this new era are a close balance of support between the two major parties, a strongly partisan electorate, a high degree of stability in voter preferences and in the outcomes of elections at the national and state levels, and finally a growing Democratic advantage in both the national popular vote and the Electoral College. In this article, I lay out the evidence for these transformations and discuss their implications for forecasting presidential elections.
COMPETITIVE ELECTIONS
The new era of electoral politics in the United States is characterized by intense competition for control of the White House as well as both chambers of Congress. Of course there have been closely contested presidential elections throughout American history, but there have also been many landslide elections in which one candidate defeated his opponent by a margin of 10 percentage points or more in the popular vote. In fact, during most of the twentieth century, landslide elections were the rule and not the exception. Of the 17 presidential elections between 1920 and 1984,10 were won by a double-digit popular vote margin. But there has not been a landslide election since Ronald Reagan’s decisive victory over Walter Mondale in 1984.
Not only have recent presidential elections been considerably closer on average than earlier ones, their results have also been much more stable. It was not unusual in the earlier elections of the postwar era for the margin between the Democratic and Republican candidates to fluctuate widely from one election to the next. For example, the five elections between 1956 and 1972 included two Republican landslides in 1956 and 1972, one Democratic landslide in 1964, and two closely contested elections in 1960 and 1968. In contrast, the outcomes of the most recent elections have shown much less variability.
Table 1 summarizes the trends in competition in the 16 presidential elections between 1952 and 2012 by dividing them into four eras, each consisting of four elections: 1952–1964, 1968–1980, 1984–1996 and 2000–2012. For each era, the table displays the average popular vote margin of the winning candidate and the standard deviation of the margin of victory or defeat of the Democratic candidate. The data in this table show a striking pattern. The four presidential elections between 2000 and 2012 have had by far the closest average margins of victory and by far the smallest variability from election to election of any of the four eras. In fact, the four most recent presidential elections had the closest popular vote margins and the least election-to-election variability in Democratic margin of any set of four consecutive presidential elections in the past century.
A CLOSELY DIVIDED AND PARTISAN ELECTORATE
It is no coincidence that the results of recent presidential elections have been unusually close and unusually stable. This recent pattern reflects important changes in the American electorate during this time. Specifically, it reflects the emergence of a closely divided and strongly partisan electorate (Bafumi and Shapiro Reference Bafumi and Shapiro2009; Bartels Reference Bartels2000; Jacobson Reference Jacobson2012). Not only is there a closer division of support between the two parties, but there is now a much closer relationship between party identification and presidential vote choice because loyalty rates are very high for both Democratic and Republican identifiers. In 2012, overall party loyalty set a record with 92% of Democratic identifiers and 93% of Republican identifiers voting for their party’s presidential candidate according to the National Exit Poll.
Both the closer balance in support between the two major parties and the greater party loyalty of voters reflect the outcome of an ideological realignment that has transformed the American party system since the 1960s. That realignment has resulted in a much closer correspondence between voters’ party affiliations and their ideological orientations (Abramowitz Reference Abramowitz2013; Abramowitz and Saunders Reference Abramowitz and Saunders1998; Highton and Kam Reference Highton and Kam2011; Levendusky Reference Levendusky2009). Thus, in the American National Election Study (ANES) data, the correlation between the seven-point party identification scale and the seven-point liberal-conservative scale among voters has increased from .34 in 1972, the first time the ideology question was included in the survey, to .67 in 2012. This ideological realignment has also contributed to a regional realignment with more conservative states and regions trending Republican and more liberal states and regions trending Democratic. The end result of this ideological realignment is an electorate characterized by very high rates of party loyalty and very stable voting patterns.
In fact, the four most recent presidential elections had the closest popular vote margins and the least election-to-election variability in Democratic margin of any set of four consecutive presidential elections in the past century.
A STABLE ALIGNMENT
Another major feature of the recent era of electoral competition in the United States has been a very high degree of stability in the preferences of voters and, therefore, in the outcomes of presidential elections at the state and local levels as well as at the national level. Not only have the election-to-election swings in the national popular vote been much smaller than in earlier times, but the outcomes at the local and state levels have been exceptionally stable. The correlation between Democratic share of the vote in 2008 and the Democratic share of the vote in 2012 across all 50 states and the District of Columbia was a remarkable .98. This was the strongest correlation between two consecutive presidential elections in the postwar era.
Only two states—Indiana and North Carolina—switched sides between 2008 and 2012. This was the smallest number of states switching sides in two consecutive presidential elections since the end of World War II. The previous record of three states was set between 2000 and 2004. In fact, simply predicting that every state would support the same party in 2012 as in 2008 would have yielded a far more accurate prediction of the 2012 electoral vote than a forecasting model developed by political scientists based on measures of state economic performance (Berry and Bickers Reference Berry and Bickers2012). And this naïve prediction would also have outperformed Nate Silver’s highly sophisticated Electoral College forecasting model as late as mid-October, only three weeks before Election Day.
GROWING DIVERSITY AND AN EMERGING DEMOCRATIC MAJORITY
An examination of the results of recent presidential elections also suggests that the Democratic Party now enjoys a modest but significant advantage in both the popular vote and the Electoral College—an advantage that is likely to grow over the next several election cycles. Democratic presidential candidates have won the popular vote in five of the last six elections. Moreover, the success of Democratic candidates in recent presidential elections appears to be due to more than chance factors. The most important explanation for this recent run of success has almost certainly been the rapid growth in the nonwhite share of the US electorate.
Figure 1 displays the trend in the nonwhite percentage of voters in US presidential elections between 1952 and 2012 based on data from the American National Election Studies. The data in this graph show that the nonwhite share of the electorate grew rather slowly between the 1950s and 1980s but has increased much more rapidly since 1992, reaching 28% in 2012—an estimate that is identical to that in the 2012 National Exit Poll. Without this increase in the nonwhite share of the electorate, it is highly unlikely that Barack Obama would have been elected in 2008 or reelected in 2012.
Figure 1 Growth of Nonwhite Share of US Electorate Since 1952
Source: American National Election Studies.
The steady increase in the nonwhite share of the US electorate since 1992 reflects major demographic changes in the US population, especially the growth of the Latino population. Moreover, this trend is almost certain to continue for the foreseeable future. Census data show that the age cohorts that will be entering the electorate over the next several election cycles include a much larger proportion of nonwhites and a much smaller proportion of non-Hispanic whites than the oldest age groups in the population. And the vast majority of nonwhites who will be coming of voting age in the future were born in the United States. Normal generational replacement therefore guarantees that the nonwhite share of the electorate will continue to grow regardless of what happens to US immigration trends or policies.
The Democratic Party already enjoys a structural advantage in the Electoral College. Eighteen states and the District of Columbia voted for the Democratic candidate in all four presidential elections between 2000 and 2012. Another three states voted for the Democratic candidate in three of these elections including 2012. Taken together these 21 states plus the District of Columbia have a total of 257 electoral votes, only 13 less than the 270 needed for victory. Moreover, three of the five states that have divided their support evenly between the parties in the last four presidential elections—Colorado, Nevada, and Virginia—have shown a strong Democratic trend over these 12 years as a result of population shifts and demographic changes including the rapid growth of the Latino electorate.
The trends described here do not mean that the Democratic Party now has a lock on the Electoral College. We cannot predict that the Democratic nominee will be victorious in 2016 let alone 2020 or 2024.
At the state level, the 2012 election resulted in the most polarized outcome in modern history. Despite the relative closeness of the national popular vote, only four states were decided by a margin of less than five percentage points while 27 were decided by a margin of at least 15 percentage points. On the one hand, Barack Obama carried 15 states and the District of Columbia with a total of 191 electoral votes by a double-digit margin. He carried another eight states with a total of 81 electoral votes by a margin of between five and 10 points. On the other hand, Mitt Romney carried 20 states with a total of 154 electoral votes by a double-digit margin and three other states with a total of 37 electoral votes by a margin of between five and 10 points.
While both parties now start off presidential campaigns with large numbers of states and electoral votes essentially locked up, the Democratic candidate now starts off with a substantially larger base of electoral votes. In fact, the 23 states and the District of Columbia that Barack Obama carried by a margin of at least five percentage points have a total of 272 electoral votes, two more than the 270 needed for victory.
IMPLICATIONS FOR THE FUTURE
The trends described here do not mean that the Democratic Party now has a lock on the Electoral College. We cannot predict that the Democratic nominee will be victorious in 2016 let alone 2020 or 2024. Short-term forces including the state of the economy, the incumbent president’s approval rating, and the first-term incumbency advantage still matter. But the long-term trends described here are real and they are very likely to continue in 2016 and beyond. Based on these trends, we can make at least three predictions about future presidential elections.
1. The national popular vote margin is likely to be close but the large majority of states will be won by a decisive margin and voting patterns will be very stable from election to election. The Democratic nominee in 2016 is almost certain to carry every state that Barack Obama carried by at least 10 points. Similarly, the Republican nominee in 2016 is almost certain to carry every state that Mitt Romney carried by at least 10 points. The electoral vote winner will again be decided by a small number of swing states. Most of those states in 2016 will be the same ones that were swing states in 2012.
2. The nonwhite share of the electorate, and especially the Latino share of the electorate, will continue to grow. States with large and/or rapidly growing Latino populations will show continued movement in a Democratic direction as Latinos gain citizenship and are mobilized into the electorate.
3. Unless the Republican Party finds a way to substantially increase its share of the nonwhite vote, Republican candidates will be at an increasing disadvantage in future presidential elections. This means that Republican candidates will only be able to win these elections if short-term factors tilt strongly in their favor.
