Stanford, California - American politics is increasingly divided due to polarization between the policy positions of the two major parties, selective exposure to media outlets, and sometimes a personal dislike of political opponents, a Stanford researcher says.

Stanford researcher Tobias Konitzer studies political polarization, media fragmentation, online campaign trends, voter attitudes and issue formation. A doctoral candidate in political communication, he is tracking American public opinion on 11 key campaign issues weekly during this election season and was a co-author of a recent Washington Post column on the subject.

Stanford News Service interviewed Konitzer about this year's election dynamics:

Why is the American electorate increasingly polarized?

Whether or not the American electorate has become polarized is actually subject to considerable scholarly debate. On the one hand, it is clear that the dislike of the "other" party – the party one does not identify with – has consistently gone up in recent years and presidential approval ratings among voters identifying with the other party are at an all-time low. This kind of affective polarization manifests itself in daily life, for example, in purchasing patterns or even marriage.

In my research with Stanford political scientist Shanto Iyengar, we show that over 80 percent of all marriages in the country are same-party marriages and, when we control for the composition of neighborhoods, even trumps the degree of same-race marriage.

On the other hand, scholars have debated whether polarization is also as rampant when it comes to ideology, by which we mean policy preferences. Stanford political scientist Morris Fiorina and colleagues have maintained that voters instinctively take moderate positions on most issues, but recent data seems to suggest that many voters do take positions at the poles of the ideological distribution.

This suggests that staggering "elite-level polarization" has rubbed off on voters to a certain extent. (This refers to the distance between the policy positions of Republican and Democratic lawmakers.) We believe that polarization among the political elite – driven by a donor class that is ideologically more extreme and by the party's decline in influence on the nomination process – has slowly trickled down to the voters to some extent.

Other theories as to why voters might become more polarized include selective exposure – choosing media sources with a clear partisan agenda similar to one's own – and a growing desire to avoid political viewpoints that conflict with one's own. One result is a decline in swing voters, i.e., voters with changing party allegiances.

For campaigns, while persuading voters might become more and more difficult, the focus is likely to shift to mobilization and expanding the electorate that turns out at the ballot box. Campaigns can indeed increase turnout, especially when campaign messages are mediated through social networks, for example via shares on Facebook, as some of my recent co-authored work suggests.

What are the biggest issues this year?

Given the spectacular success of Donald Trump in the primaries, it would be tempting to say that, at least on the Republican side, immigration is such an issue.

However, exit polls seem to suggest otherwise, with immigration being the primary concern for only about 15 percent of voters. What's more, the same exit polls reveal that about 30 percent of Republican voters name the economy as the most important concern, while the numbers topped 50 percent in the 2012 election, again per exit polls. This decrease in importance this election cycle is reflective of more diverse party coalitions, especially on the Republican side.

To explain, the traditional base of the party, the so-called Rockefeller Republicans who are driven by concerns about redistribution and taxation, now are vying for influence with evangelicals, single-issue voters (on topics like immigration), and libertarians, among others. This fractious nature of the Republican Party has been long in the making:  Richard Nixon brought a racially conservative South to the party, Ronald Reagan brought anti-abortion rights advocates to the party, and George W. Bush rallied up evangelical voters.

In some ways, every fraction has ended up with its own candidate in this year's primaries, but the fractious nature of the Republican Party also has forced candidates to take ultra-conservative positions on each of the issues to not endanger splitting the Republican coalition. This has led to an increasing gap between Republican voters and Republican politicians, which can help explain the Trump phenomenon.

On the other hand, candidates are increasingly able to "micro-target" messages to small swaths of their base. Together with researchers at Duke University, I argue in a forthcoming study that the accessibility, usability and accuracy of "big data" about the electorate mean that campaigns are better able to individualize and disseminate messages to ever smaller and more narrowly targeted audiences across ever-growing online and mobile platforms. It is hence no surprise that digital campaign advertising is projected to be the fastest growing sector of campaign communication, and that development might help mask the fractured state of the Republican Party in the fall and decrease the risk of alienating some fractions.

On the Democratic side, health care, education, the economy and job growth are important issues. The clear message seems to be that the Democratic base is eager to see a continuation of the Obama agenda. In Mississippi, for example, 71 percent are in agreement that the next president should follow in his footsteps.

How do you see a possible Trump vs. Clinton contest?

If we believe the polls, Hillary Clinton would win this match-up by about 6 percent, but there is a good reason not to place too much value in these match-up polls, which ask voters for a conditional probability or scenario based on Hillary Clinton and Donald Trump getting the nominations. It is well-documented that people have a limited understanding of conditional probability, and in some cases that also goes for journalists reporting on these hypothetical match-up polls. Maybe a better way to assess the chances of victory is to look at prediction markets, which currently estimate the probability of Democrats clinching the presidency at 69 percent.

The actual election outcome aside, it will be interesting to see what a Trump candidacy will do to the Republican Party, given that resistance by the establishment has garnered wide attention. Will it result in a massive realignment (although the Democratic candidates are further apart from the Republican base than Trump is), or even a party split?  However, these are extremely rare in American politics.