Polarization Means Dislike and Distrust, Not Just Policy Divergence

Earlier this year my family decided to add a dog to the household.  A lovely woman whose aging knees were a bit overmatched by the rigors of puppy-parenting was searching for a good family to adopt her dog, Riley.  As my wife, Suzanne, prepared to pick Riley up, she decided it would be best to peel the Obama bumper sticker off her car.  Without knowing anything about the dog owner’s politics, Suzanne worried that she might not want to give her beloved Riley to a Democrat because many in our dark red county perceive that Democrats are not good people. 
Recent survey results suggest Suzanne might have been right to peel the sticker off before picking up the dog.  The American National Election Study has been asking people since the 1970s to place different groups on what they call feeling thermometers.  People can rate groups they like as high as 100 degrees and groups they do not like as low as 0 degrees.  As the graph below shows, Democrats’ feelings about Republicans and Republicans’ feelings about Democrats have grown much more negative of late.  During the Reagan years, the average score that a Republican rated the Democratic party was 45 degrees, and the average score that a Democrat rated the Republican party was 44 degrees.  In the Bush II years, those average scores had plummeted to near freezing temperatures, 36 and 33 degrees.  To put those scores in perspective, the average score that all Americans gave to “illegal immigrants” – no one’s favorite group – during this period was 40 degrees, significantly warmer than partisans’ ratings of the opposite party.

When scholars talk about polarization, policy disagreements have received most of the attention.  As evidence of polarization, they are looking for a picture of opinion in the electorate that looks like the picture below.  It would require the distribution of opinions to be what social scientists call “bimodal” (two humps) with most opinions clustering toward the extremes and very little overlap.

That kind of analysis is a rigged game.  Since a high percentage of Americans don't care about politics, responses to policy items will inevitably cluster toward the middle, reflecting the public’s lack of understanding.   Moreover, even well informed partisans do not want to think about themselves as ideologically extreme.  Rather it is those on the other side who, in their minds, are extreme.  Hence analysis of policy preferences will continue to produce significant overlap and muted differences between the parties.  To be sure, there is somewhat more distance and less overlap between partisans today than a few decades ago, but separation remains very far from complete
But let’s step back and think about exactly what polarization is, and why we should care about it.  At the heart of polarization is when people believe that those on the other side are bad people with dangerous ideas.  When people believe that, compromise is impossible.  You don't make compromises with the devil.  You don't look for common ground.  Instead you do whatever is necessary to minimize the danger that the other side presents.  That means all political measures are in play.  Partisans, both elites and masses, can shade the truth or change the rules without feeling too badly.  Argument quality ceases to matter.  In a polarized system, anything required to stop the other side can be justified.  Such feelings won’t show up in an analysis of policy preferences.
Whatever form of public opinion we choose to examine instead of policy preferences should capture these characteristics and be relatively easy for political experts and non-experts alike to make sense of.  How much people trust the government has potential.  People do not have to be experts to know how much they trust something.  Most people have a lot of experience with forming judgments of trustworthiness of others in their everyday lives.  Trust in government is just an extension of a familiar task.  Moreover, scholars have shown that political trust is a very meaningful attitude, strongly linked to how much people want the government to do at any given point in time.  This is especially true when a government policy requires people to make a sacrifice.  For people to make sacrifices for others, they have to trust the institutions that will carry out the policies to treat them fairly.
In 2010, I included a trust in government question on a survey.  It asked “how much of the time do you trust the government to do what is right:  just about always, most of the time, some of the time, or never.”  In the graph below, I have broken down the distribution of responses by party.  By the traditional definition of polarization (bimodal distribution with peaks toward the poles), we see far too much overlap to suggest polarization.  Indeed, 58 percent of Democrats say they trust the government only “some of the time”.  From the traditional perspective, then, the story would be over.  No polarization. 

But, wait a minute.  When we consider what polarization really means, might it not be significant that fully 52 percent of Republicans say they never trust the government in Washington to do what is right?  And a mere 2 percent say they trust government even “most of the time”.  Has there ever been a time in the history of polling when this has been the case?  (The answer is not even remotely so). This is evidence of polarization of the sort I described above.  It says that one side is completely unwilling to even consider engaging the other side’s ideas.  People do not make compromises with those whom they find completely untrustworthy.
One could object to my interpretation of these results by suggesting that they merely reflect the GOP’s long standing antipathy toward government.  Except that isn’t true.  In the graph below, I break down the amount of trust in government by party and presidential administration going back to Lyndon Johnson’s presidency when the trust question started to be asked regularly by the American National Election Study.  It shows that Republicans have often trusted the government to do what is right quite a lot.  Indeed Republicans have consistently expressed more trust than Democrats when Republicans have occupied the White House.  In fact, during George W. Bush’s presidency, when the GOP also mostly enjoyed majorities in both houses of Congress, Republicans in the electorate trusted the government more than any time since the 1960s when pretty much everyone trusted the government.  Republicans in the Bush years trusted the government more than they did during the Reagan years.

What is perhaps most striking about this graph is just how wide the Republicans’ swings are, especially lately.  Since trust collapsed after Vietnam and Watergate, Democrats’ average trust scores have fluctuated within a relatively narrow 10 percentage point band between 30 and 40 percent.  For Republicans, that band is nearly five times as wide.  In fact, trust in government among Republicans dropped an incredible 50 percentage points between the Bush II years, when their party controlled government, to the first two years of the Obama administration, with Democrats firmly in control.   
Also remarkable is how much the gap has grown between Republicans’ and Democrats’ trust in government at given points in time.  Of course, there has always been a tendency for people to trust government more when their side has been  in office, but the differences have generally not been that large.  The average party difference during the first six presidencies in the time series was about 9 percentage points.  That party difference nearly doubled to 17.75 points in the George W. Bush years, and then tripled to 28 points during Obama’s first term.  Something fundamental has changed.  Never big fans of being governed by Democrats, Republicans are now completely rejecting the notion.
Why is this important for the debate about polarization?  These data may help explain why the federal government all but ground to a halt in 2010.  Gridlock does not require Republicans and Democrats to disagree fundamentally about specific policies.  Indeed, in his recent New Yorker piece, Ezra Klein ably notes just how many policies Democrats pursued in Obama’s first term (e.g. individual health care mandate, cap and trade) that were very recently Republican ideas.  Furthermore Ronald Reagan himself was something of a Keynesian in his approach to reviving a flagging economy in the early 1980s, and George W. Bush had no fear of deficit spending either.  More important than specific policy disagreements is how one side regards the other.  Do they dislike them?  Do they distrust them?
The design of American political institutions requires compromise.  Compromise requires both sides to make sacrifices.  That is, for public consensus to emerge and to motivate policy change, a sizable number of conservatives or liberals must be willing to sacrifice their general ideological principles to allow their ideological opponents leeway to pursue their policy goals.  Those who trust government more are most willing to make such sacrifices.  With next to no Republicans willing to make such sacrifices these days, it makes the country far more difficult to govern. .  That captures the essence of polarization and helps us understand why Washington is so dysfunctional today.

– Marc Hetherington, Vanderbilt University

Read Ahead