Seventeen years after the infamous 1995-96 shutdown, we find ourselves again beset with the consequences of fractured political parties, with Congress and the president having just emerged from a shutdown crisis and narrowly averted default on our national debt. Not surprisingly, a good deal of punditry has focused on the causes of political polarization. Gerrymandering is a favorite bête noire, as is the effect of a politically splintered media and Internet landscape. In fact, polarization is a multi-layered development that eludes one simple answer; its roots go back decades before it became a truly national phenomenon in the 1990s.
The first point to make is that we are indeed more polarized than ever. Our parties today are more distinctive and homogenous than at any time in American history, save perhaps during the late nineteenth century. The fact that fewer and fewer moderates populate either party in Congress is evident from nearly every measure of ideology employed by political scientists.
As evidence we need to look no further than DW-NOMINATE, a formula developed by Keith Poole and Howard Rosenthal based upon roll call votes taken in the House and the Senate, widely considered as the standard. Poole and Rosenthal use a single-dimension graph to show much of the variation in public expressions of ideology, and they show the dramatic rise of polarization in recent decades. As one can see here, the ideological difference between the modern-day member of Congress in both the House and the Senate far exceeds the historical average all the way back to the nineteenth century.
So the question is not whether we are truly more polarized, but how we got there. The most-favored hypotheses include gerrymandering, partisan sorting, and population shifts that have packed Democrats in cities and spread Republicans “across suburbs, exurbs, and rural areas” in the words of The Washington Post’s Dan Balz. In fact, many scholars are skeptical about the true effects of gerrymandering, even though lawmakers and the press make much of this as a cause. Research suggests that gerrymandering is, at best, only a marginal culprit for polarization. After all, both the House and Senate have become more polarized, but Senate races are, of course, statewide. Something else is afoot.
When we look at the other theories suggested, the biggest methodological problem is that that some of the critical factors leading to polarization all happen to occur at roughly the same time, so this leaves us with a chicken-and-egg conundrum. For example, technological leaps in computational power and database management have made it easier for campaigns to pinpoint the partisan leanings of neighborhood blocks, but this has coincided with the trend of voters moving to more ideologically homogenous communities—a factor documented amply by Bill Bishop in The Big Sort.
Furthermore, voters have sorted themselves out in ways other than geography, and this started in the 1970s. As lawmakers began to represent more homogenous congressional districts and congressional parties became less diverse ideologically, Republicans and Democratic voters became increasingly more conservative and liberal, respectively. Split-ticket voting has now reached record lows. As documented by the Pew Center for the People and the Press, Republicans and Democrats, on a host of issues, have aligned themselves more consistently with the ideological poles of their parties.
With the demise of conservative Southern Democrats and liberal Northeast Republicans, elites and voters alike found themselves in parties that were increasingly singing the same ideological tune across an array of issues. What had once been a raucous cacophony within the parties had become two distinct and loud opposing chants seeking to drown each other out through sheer will and persistence. This is a seismic shift since the 1960s, when political scientist Philip Converse famously wrote a seminal article about how most voters were ideologically “unconstrained”, exhibiting little interest in bringing consistency to their diverse political beliefs.
The first major crisis of polarization shook the country in the mid-1990s, when the GOP regained control of Congress and forced the 1996-96 shutdown. In the first two tumultuous years of the Clinton administration, the GOP refused to help the president on his domestic agenda, and this strategy worked: in 1994, they aggressively exploited Clinton’s rocky start to take control of Congress for the first time in 40 years. Citing what he saw as an electoral mandate, Republican House Speaker Newt Gingrich caused the first lengthy shutdown of the federal government in a bid to force the president to balance the budget and trim entitlement spending. As we now know, the result emboldened Clinton’s presidency, hurt Republican chances to retake the White House, and tarnished Gingrich’s reputation. But polarization continued unabated.
The episode prompted scholars to look more closely at the country’s growing political divide. One of the best, more recent works is Sean Theriault’s Party Polarization in Congress, which looks at 30 years of congressional history. Theriault’s sophisticated, multilayered analysis finds that gerrymandering is indeed partly to blame, as is the “great sort” of voters, both ideologically and geographically. But other factors are also at play. Theriault notes, for example, that polarization has not only increased on substantive votes in the House and Senate, but also on routine procedural motions that once had been considered routine and often passed with considerable bipartisan majorities. What was going on?
To answer that question, Theriault and others have found three factors important to understanding the perpetuation of polarization and congressional fights. The first is the rise of “narrowcasting,” or the spread of information to increasingly specialized niche audiences. The second is the psychological phenomenon known as group polarization, or voters moving with their feet, which is aided and abetted by narrowcasting. Finally, congressional majorities have smaller governing margins available to pass legislation than they used to, and this has reduced the incentive to work across party lines; party cohesion is placed at a premium. The first two factors affect voters and political elites, while the final factor results in party leaders resorting to increasingly creative methods to advance party objectives while shutting out the minority party. We need to understand all three as necessary conditions for polarization.
This framework also helps us understand the media equation. One key driver of narrowcasting is cable television and the Internet, which have created ever-smaller spaces for individuals to segment themselves by their particular interests. In the 1960s, television consisted of the three major networks and a few independent stations. Cable allowed the proliferation of stations catering to specific interests of viewers – including political coverage. Starting with the rise of conservative talk radio in the 1980s, voters now select news sources aligned with their politics. This comports with long-held documentation by psychologists that individuals try to avoid cognitive dissonance; that is, messages and information conflicting with their beliefs and predispositions.
With the decline of the centrist and nonpartisan press, voters can more easily avoid viewpoints out of step with their own. The legal scholar Cass Sunstein has argued that the rise of the Internet and the ability to create a “Daily Me” catering to one’s own beliefs and values goes so far as to threaten democratic discourse. A recent study of social networks demonstrates the consequences of this: Republicans and Democrats alike are far more likely to have friends and follow feeds aligning with their own political dispositions. What people tweet and share is highly correlated with their political inclinations, which only exacerbates the problem. There is no shared space to engage in cross-partisan, cross-ideological dialogue.
In turn, the ability to filter out views discordant with one’s own has made it increasingly likely that this chorus of like-mindedness can actually make one’s own beliefs more extreme via group polarization. As documented by the late psychologist Irving Janis, group dynamics in decision-making can lead to suboptimal results because the same dynamics moving a group to consensus can actually stifle debate. Group polarization also moves group consensus and dynamics to behavioral extremes. Self-segmented groups with little external checks on behavior tend to foster ever more extreme behavior. No doubt, this occurs among citizens who are isolated from alternative perspectives and points of view — all of which increases the tendency politically to polarize.
If elites and the public are increasingly able to protect themselves from “dangerous” views, members of Congress find themselves increasingly beholden to producing ideological victories for those groups that are the bedrock of their constituencies. The problem, of course, is that not only is divided government pervasive in the post-World War II era, but the ability to marshal a majority voting coalition has become ever more difficult. In 1940, the average House majority party had nearly a 65-seat cushion of control. This rose to almost 110 seats in the 1970s. In 1990, the decade Republicans managed to wrest the House from the Democrats, this fell to 47. The following decade, the margin shrunk again to 34, and in the last two years it has increased only slightly, to 41 seats. Since the 2012 election, House Republicans have only a 17-seat majority.
Instead of building bipartisan coalitions, the result of slimmer margins of control is that majority party leaders increasingly resort to procedural innovation to secure legislative victories that will advance party objectives. Their hope is that this will enhance the party’s electoral brand and yield more decisive electoral control. These innovations have included the increased use of “closed rules” in the House of Representatives that limit the minority’s ability to debate and amend legislation (as documented by Barbara Sinclair at UCLA), and the rise of filibustering in the Senate by the minority to prevent majority-party victories. The cumulative effect of these tactics is fewer and fewer legislative victories, more animosity between the parties, and a search for additional leverage to accomplish majority party objectives.
As I have documented with Justin Grimmer, Craig Goodman, and Frances Zlotnick in a recent paper presented at the American Political Science Associations annual conference (view the presentation here), the minority party is more inclined to resort to partisan rhetoric to excite the base and create a winning electoral narrative. So it should not have been a surprise that the GOP sought to use the debt limit, sequestration, and the shutdown of government itself to advance its goals, given that passing legislation of any substance or significance has become nearly impossible due to divided government and the GOP’s own internal divisions.
The gridlock that led to the government shutdown, in short, was the product of several forces: polarization of elites and the masses, narrower margins of control in both legislative chambers, and the willingness of party leaders to increasingly resort to creative procedural means to advance partisan agendas. Is this the new normal in Washington? Perhaps. But this leads to the question of how long voters will tolerate this tension in government – and whether realignment is in the works.
Realignment theory, developed by the scholars V.O. Key and Walter Dean Burnham beginning in the 1950s, posited that grand, sweeping political change is relatively unusual in the U.S. For dramatic political transformation to occur, an electoral realignment fueled by voter mandates via sweeping electoral victories is necessary to produce the unified government needed to move beyond gridlock. In a realigning election, not only would the governing elite itself be replaced; voters would shift their longstanding partisan allegiances to the support the new governing majority.
Although realignment theory has fallen out of favor, one question we now need to ask is whether Congress’ persistent inability to govern will create enough anger to push voters outside of their narrow ideological boxes in search of governing majorities. That may be the only way we can move the country to a new equilibrium and break the current political stalemate. Whether a return to the elusive realignment augers a more liberal or more conservative direction – or something different altogether — remains to be seen.
Realigning elections are rare, occurring only five times in the past (1800, 1828, 1860, 1896, and 1932). For one to occur in the 2016 cycle, we would expect to see the first signs in the 2014 midterm elections. First, we would witness a consistent shift in precinct vote totals advantaging one party across a wide geographic area. Second, we would see a movement in voter affiliation to the same party. This would result in the voters decisively giving one party unified control of Congress. Following the midterm, these patterns would intensify in a realigning election of 2016, leading the majority party to not only capture the presidency but to do so with overwhelming support in the popular and Electoral College vote. In the last electoral alignment, for example, Democrat Franklin Roosevelt crushed incumbent Republican Herbert Hoover, winning 42 states and 472 Electoral College votes. The New Deal realignment not only gave the Democrats huge governing majorities in the House and the Senate, but led to Democrats winning many down-ballot races across the country.
In my last piece for Symposium Magazine, I predicted that Democrats would fare poorly in the 2014 midterm elections. I stand by that prediction, with the qualification that such a result hardly is tantamount to a Republican realignment. Instead, Republican gains would be merely consistent with historical patterns going back to the 1930s. For the 2014 midterms to auger a Republican realignment, voters of all ages and backgrounds would need to abandon the Democratic Party en masse, while Democrats would sustain losses not just among vulnerable incumbents but among those in safe seats today. Given how poorly the GOP brand has fared in the aftermath of the government shutdown, this is highly unlikely. As for Democrats, the best-case scenario is that they minimize any midterm losses – but winning the House remains a long shot.
Conversely, if Democrats minimize their losses – or even narrowly retake the House – this does not augur a Democratic realignment. Something truly devastating would have to occur to push the electorate decisively and completely into the arms of either party – say, a total disaster of the Affordable Healthcare Act rollout, or a global economic catastrophe precipitated by a debt-ceiling breach. Otherwise, it is unlikely that 2014 will produce the congressional victories and voter displacement of the sort signaling an upcoming realignment in 2016. In turn, the new normal of intense partisanship and inability to govern is likely to persist for some time to come, providing Americans with ever more frustration and consternation.