"President of what?" (Or, measuring presidential ideology and its impact is harder than you think)

by Julia Azari 

With the 2016 presidential race apparently under way, we will probably hear much more about ideology and presidential polarization soon. This conversation got going among political science bloggers (including our own Seth Masket), when Peter Wehner wrote about this a few weeks ago, selecting data about Obama to make the case that the Democrats have abandoned the center.

Based on Poole and Rosenthal’s updated roll call scores, political scientists and data-minded journalists are moving toward the opposite conclusion -that Republicans, not Democrats, are driving much of the party polarization we observe. Scholars have used these data to refute some recent claims that Obama has pulled the Democratic Party dramatically to the left. This is perfectly sensible - but it also opens up some other questions about how to measure presidential ideology.

I think this question is interesting for a bunch of reasons - as a measure of elite ideology and change over time (as John Gerring uses presidential campaign speeches in his 1998 book),  for the sake of understanding how presidents define issues (in the political time-legitimacy sense), etc. It also obviously has the potential to be of interest to people engaged in more mainstream inter branch relations research - but I think ideas and words are useful subjects of study on their own.

Coming up with a reliable and accurate measure of presidential ideology is difficult and not straightforward. The Common Space scores at Voteview.org include presidents, though with the caveat that many of the presidential scores are not based on many observations. Adam Bonica has a useful measure of presidential candidate ideology, although it doesn't capture how presidents behave in office. (And more recently, this measure using Twitter!) In a 2010 piece in Political Analysis, Shawn Treier compares a few different ways to measure the ideology of modern presidents. He found that if we use presidential position scores on bills, presidents look like committed partisans. Because presidents rarely take formal positions on less divisive bills, according to Treier, “the president loses opportunities to distinguish himself from the most extreme members of his party.” Using these measures, Democratic presidents, he observes, are close to Ted Kennedy, while Republican presidents are hard to distinguish from Jesse Helms. However, Treier finds that presidents look more moderate when we measure their positions using vetoes and signatures on passed bills.

Dan Wood takes a different approach to measuring presidential ideology in his 2009 book, The Myth of Presidential Representation. Using both human and computer coding, Wood uses presidential rhetoric to develop scores for presidential “issue liberalism,” which encompasses categories like the environment, race, urban affairs, and welfare. The main finding of the book is that presidents’ ideological positions, as measured through this extensive analysis of their rhetoric, reflect the ideologies we associate with their parties, rather than converging near the center of public opinion.

In other words, in contrast with, for example, Wehner’s use of a couple of high-salience items in the NYT, measuring presidential ideology defies a simple answer. Depending on what we look at, we get different conclusions.

Jesse Helms and Ronald Reagan

But when analyses of presidential ideology tend to have an implicit causal question buried in there somewhere. Presidents usually fill the causal role. The idea that presidents "pull" their parties in ideological directions lacks adequate evidence in most cases. Presidents do, however, affect party politics through their governing choices and how they define those choices.

Gary Jacobson and Alan Abramowitz have both made cases for Bush - specifically, the Iraq War - as a catalyst of further polarization. And polarization has climbed by most measures during Obama's presidency also. This leads us to some even more complicated questions: how much do presidents influence the nature of party and ideological conflict? And assuming that answer is different from zero, then why do some seem to exacerbate polarization, and by what mechanisms?  

Obviously, I don't know for sure. But I suspect that in addition to the the kinds of extensive quantitative measurements that Treier and Wood offer, there are contingent and qualitative factors that shape the process. It's commonly assumed that a polarizing figure would also be ideologically extreme (and vice versa) - but it's not clear that this is necessarily true.

In terms of governance, Bush garnered more support from the other party for his signature tax cuts in 2001, and worked with Ted Kennedy on the legislation that eventually became No Child Left Behind. Obama has fared worse when it comes to that kind of bipartisanship – although there are obviously many ways to attribute blame for that. In a sense, both presidents have governed to the left by expanding government. And in one sense, both presidents have governed to the right by maintaining a strong surveillance state and guarding unilateral executive power. But that doesn't really capture the political impact of their presidencies.

Both presidents have taken advantage of social issues. For Bush mobilizing socially conservative activists around the issue of gay marriage proved politically advantageous in 2004. In 2012, Obama embraced marriage equality and engaged new activists in the Democratic coalition. Because Obama chose to make healthcare a priority, the question of abortion and contraception coverage has been prominent on the issue agenda throughout his presidency. This isn’t a move to the left, exactly. Healthcare has long been a Democratic priority, and political conditions allowed it to pass while Obama was president. What the impact illustrates is the fact that governing is divisive. Just as new policies create supportive constituencies, they can also mobilize opponents.

Another factor that Obama and Bush share is a strong reactive theme in their presidencies. Responding to the attacks of September 11, 2001, changed the direction of Bush’s presidency. For Obama, the financial crash preceded his election, but not most of his campaign. The Arab Spring opened up new questions about U.S. involvement in the Middle East after a campaign that clearly emphasized limiting our role. The beginning of his second term was spent promoting gun control after the Sandy Hook shooting. They are hardly the only presidents to have to respond to issues that they hadn’t anticipated, or to have their years in office defined by something that came before. But the frequency and salience of the reactive situations have been a notable feature of both presidencies.

Here’s the connection with ideology. Because Bush and Obama have spent so much time addressing situations they did not fully anticipate, neither one has done a great job of establishing a larger narrative about their policies. Bush talked about compassionate conservatism but never clarified what that meant. He aligned himself with Reagan’s assertion that “government is the problem” but expanded federal reach into education and signed a new cabinet department into law.

For Obama, definition has been an even bigger problem (see systematic early analysis here). Part of the problem is being an “opposition” president – to whose tradition should Obama link his policies? FDR? LBJ? (Expansive liberal policy achievements, brought down by badly conceived foreign policy? Probably not the frame the administration wants.) Beyond that, Obama seems to have struggled to define the political meaning of his administration’s positions and decisions on healthcare reform, anti-poverty measures, immigration changes, or foreign policy. Paradoxically, their short-comings in self-definition have made them appear more ideological, because their political opponents have been able to cast their decisions in divisive, often hostile, terms.

When Bush and Obama did try to define their presidencies, the focus has largely been on their own personas rather than on policy. Part of what constitutes ideology in American politics - and gives it its potency - is its relationship to conflicts over national identity (links to classic works on these  debates). Running in 2000, and for the rest of his presidency, Bush presented himself as a kind of cowboy everyman. This lent a very specific ideological cast to his foreign policy decisions. Obama has drawn on his personal background to embody the national unity on which he first ran. Political opposition has arisen in response to the idea of changing demographics and changing culture. In the electoral process, this is of minimal importance. But when presidents present their own biographies in ways that are deeply symbolic and linked to ideology, they invite opponents to offer objections that are both personal and ideological. Contested interpretations of identity get tied into policy debates, changing their stakes and implications. 

Scholarship on presidential ideology reveals that it is not simple to assess. It's even harder to measure and explain a president's ideological impact. The ideological effect of a presidency may also be driven by contingency and context - not just the president's positions. That's all the more reason to be cautious with the conclusions we draw about presidents and polarization.