Follow the Leader

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 8.4

This post is largely focused on two-party electoral systems, like that of the US. While the specific manifestations will be different in multi-party proportional representation systems, the underlying psychological dynamics are still present.

It is widely assumed that partisan animosity is the result of deep policy disagreement. However, it is often the other way around. Partisan loyalty generates opposition to policies advocated by the other side as a signal of team loyalty. Politicians have an incentive to fan such animosity for electoral purposes, rather than policy purposes. Voters generally recognize that most public policy topics are more complex than they have time to figure out, and so reasonably defer to favored politicians who claim to understand such things. A respected group of fifteen political scientists wrote in Science magazine in 2020,

“Overall, the severity of political conflict has grown increasingly divorced from the magnitude of policy disagreement.”

In studies conducted by Geoffrey Cohen, a professor of psychology formerly at Yale and later at Stanford, partisan voters were asked to evaluate a policy proposal in a news story. Some participants read a version of the story that suggested a policy (for example, either a very generous or a very stingy welfare proposal) was favored by prominent Democrats, while other participants read a version of the story that suggested it was favored by prominent Republicans. The participants’ own evaluations of the policy closely mirrored the suggested preference of the party they generally agreed with, regardless of whether the policy would generally be deemed liberal or conservative. Other recent research confirmed that many partisan voters will align with policy positions they believe a party leader has taken independently of whether they self-identify as liberal or conservative.¹

In a fascinating study published in a 2019 issue of the highly respected Science Advances, Cornell researchers found that the formation of  “partisan” divisions on policy may evolve through chance information cascades, rather than ideology or thoughtful deliberation. They divided 2,000 Democratic and Republican adherents into ten “parallel worlds” with both Democrats and Republicans in each world, and gave them an assortment of unfamiliar policies to grapple with, on which the Democratic and Republican parties did not already have identified positions, such as whether artificial intelligence (AI) should be used to identify online criminals. When participants’ responses to the policy questions (indicating the partisan preference of each respondent) were shared within a given “world,” partisans tended to coalesce on policy positions. But in one world, for example, the Democrats adopted a particular policy while in another world that policy was adopted by Republicans and opposed by Democrats. As the researchers observed,

“Our results suggest that partisan alignments across substantively diverse issues do not necessarily reflect intrinsic preferences but may indicate instead the outcome of cascade dynamics that might have tipped in a different direction due to chance variation in the positions taken by early movers.”

Tribal solidarity trumped analysis or ideology.

As in the classic chicken or egg dilemma, it is not always obvious whether partisan leaders, or their partisan followers, (or chance information cascades, as suggested by the  research just mentioned) are responsible for changes in partisan and policy preferences over time. Research by sociologist Aaron McCright notes a dramatic shift in attitudes about environmental protection by Republicans. It was Republican President Nixon who was responsible for the creation of the Environmental Protection Agency in 1970. Surveys since 1974 indicated broad agreement among liberal and conservative voters on environmental issues. As late as 1990 about 75 percent of both Democrats and Republicans believed the government was not spending enough on environmental protection. However, a sharp divide started forming around 1992. McCright argues that following the collapse of the Soviet Union and the end of the “red scare” in 1991, environmental zealotry — a “Green scare” appeared as a replacement. A number of politicians utilized this alleged threat to mobilize voters. Over the following two decades a deep partisan divide has developed on environmental issues, especially related to energy and climate change. It seems likely that a reinforcing feedback loop between conservative voters and Republican leaders was at play, but the initiation seems to have come from politicians, rather than the voters. 

Research by Gabreil Lenz and others have confirmed that rather than voters having prior policy positions and then selecting candidates who agree with them on those policies, the relationship is more often the inverse. (Lenz 2012) When voters learn of the policy positions (for example, support for public works, military spending, nuclear power, etc.) of candidates they already favor, they come to favor those policies as well. For most policies, though not single-issue voters, when a voter learns that their preferred candidate has the opposite position from them, the voter is more likely to change their opinion about the policy than change which candidate they favor.

James Fishkin observes that modern technology

“has made it possible for elites to shape opinion and then invoke those opinions in the name of democracy.”

How often have you heard elected officials insist that “the people have spoken loud and clear” about whatever policy the candidate favors, simply based on the fact that the candidate won the last election? Even in the absence of overt attempts at manipulation, many political scientists and media experts who study public opinion formation believe much of what passes for “public opinion” can be traced to elites. John Zaller, author of The Nature and Origin of Mass Opinion, argues that elites induce the public to hold opinions they would not accept if they had the best available facts and analysis.

Whether it is by chance (information cascades), or changes in leadership, or vagaries of the rank and file, the policy preferences of major parties are highly mutable over time, as long as the parties remain opposed to one another on some issue or other. The agglomeration of policy positions held by Democrats as opposed to  Republicans seems logical to us only because we are accustomed to them. But these policy positions are surprisingly malleable. They are capable of changing sides and trading roles, in tandem, without explanation or evidence of careful reasoning. The idea that the Republican party favors "free trade" had been sacrosanct for generations  But when the leader of the party, Donald Trump, decided to jettison that seemingly foundational policy, and impose a variety of protectionist trade tariffs, hardly a Republican politician or voter made a peep of objection.

Partisan leaders have immense influence when it comes to framing public support of policies. A clear example is found in the partisan handling of the market-based approach to health care — the “individual mandate” — which required that Americans purchase health insurance or face a tax penalty. The idea was advanced originally by the conservative Heritage Foundation in 1989, as an alternative to the single-payer system, in which the government would insure all residents and pay for it with taxes. In 2006, the market-based individual-mandate system was enacted in Massachusetts but embraced by the state’s Republican governor, Mitt Romney. Indeed, this sort of reform was known as “Romneycare” before it was repackaged and came to be known as “Obamacare” on a national level. In the fruitless hope of gaining some bipartisan (that is, Republican) support for an expansion of health-insurance coverage, President Obama rejected the liberals’ single-payer proposal in favor of the “Romneycare” model. Most Democrats quickly followed. Ironically, Romney would become the Republican presidential nominee in 2012, promising to repeal the Patient Protection and Affordable Care Act (by then commonly known as “Obamacare”). Once the Democrats endorsed a private, market-based approach to health care, the Republicans turned strongly against it. As the political commentator Ezra Klein puts the point:

“A policy that once enjoyed broad support within the Republican Party suddenly faced unified opposition.”

The sudden abandonment of single-payer healthcare by many Democratic politicians’ was equally noteworthy, but liberal commentators tended not to dwell on it. The ideological underpinnings and operational details of the reform became rapidly irrelevant. Team loyalty trumped substance.

If voters are so readily led, or manipulated by their preferred partisan leaders, does this also doom the prospects of a citizen panel selected by democratic lottery? Actually, no — while electoral politics is grounded in “system one” thinking linked to tribal loyalty, research has indicated that when people are induced to take the time to earnestly engage in counter-arguments on a variety of issues — think for themselves using “system two” thinking (as in a citizens’ assembly), the influence of partisan leaders drops sharply.  For the vast majority of voters, elections are all about the “system one” mental process, relying on mental short-cuts, such as cues from leaders, rules of thumb, gut reactions, information that comes readily to mind and familiarity, rather than “system two” focused attention and genuine analysis. It is possible for people to switch systems, given sufficient time and motivation. Participating in a deliberative group, such as in the British Columbia Citizens’ Assembly, can allow “system two” thinking to take over. But it is not possible to get most people to switch to “system two” in a mass election by simply exhorting them to learn about the issues and candidates.

1. Interestingly, this research was able to take advantage of the fact that Donald Trump had taken antithetical liberal and conservative positions on many issues (such as both advocating heavily taxing hedge fund managers, but also advocating giving them tax breaks) so that survey respondents could be presented with either a liberal or conservative policy actually supported by President Trump, and asked what they thought.