Accountability for Politicians: Part 1

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 13.7

Accountability is another concern some people have about sortition. Accountability is a major selling point for elections, which are generally assumed to be an essential tool of representative democracy. Accountability refers to a principal holding an agent to account. In a representative system, the principal is the public (or perhaps a particular constituency), and the agent is either an individual representative, the entire legislative branch, or the entire government. In this political context, I will focus on two sorts of accountability: corruption or ethics accountability on the one hand, and policy or representational accountability on the other. 

Any system of representation, whether using election or lottery, needs corruption accountability procedures, which can detect, punish, and preferably prevent, bribery and the like. Policy accountability is far more convoluted, however. This deals with the honest misdirection of policy away from the interests of a constituency or the general population. It is generally presumed that democracy requires procedures for removing even honest public officials through periodic elections as a means of assuring policy accountability. For those offices filled by random selection this accountability tool is entirely absent. Voters wouldn’t be able to remove a representative, and there is no obvious way to hold an entire citizens’ assembly to account for bad decisions. The belief is that without the punishment threat of removal,¹ there is no accountability, and members of a randomly selected mini-public are free to run wild. As I will explain below, elections in the real world fail to provide meaningful accountability. After addressing the myth of accountability under an electoral system, I will examine the concept of accountability under sortition.

Can elections actually hold representatives accountable?

 Jane Mansbridge of Harvard, who became president of the American Political Science Association and in 1983 authored the book Beyond Adversary Democracy, points out two approaches to accountability. The first is the “sanction” model of accountability, where the threat of retaliation at the ballot box keeps the representatives in line, or holds them to account. This could theoretically have some effectiveness in countries with multi-party proportional representation, where voters have a choice of parties with somewhat similar platforms. Politicians might worry that failing their core constituency, or getting bad press may prompt some voters to switch parties, and depending on countless other variables, could result in their losing an opportunity to be part of a governing coalition.

However, such sanctioning is all but nonexistent in the United States, with its plurality voting, two-party system. In the US very few incumbents ever face viable challengers in their party primary. Indeed, a series of studies by FairVote: The Center for Voting and Democracy determined that in most election cycles, an incumbent member of Congress is more likely to die in office, than be defeated in a primary.² In the overwhelming majority of general election contests (after the primary) incumbents are re-elected by landslide margins. As Todd Donovan and Shaun Bowler point out in their book Reforming the Republic: Democratic Institutions for the New America, echoing the trend in primaries,

“Turnover in the House, as with the Supreme Court, is largely driven by incumbent deaths and retirements, rather than competitive elections.”

The notion that nearly every incumbent has reliably served their constituents’ true interests and behaved in fully ethical ways, and thus doesn't deserve any accountability punishment, is ludicrous.

To explain the re-election rate well above ninety percent, it is cliché among media pundits to assert that “while voters say they dislike Congress as a whole, they like their own representative.” Polling indicates that this isn’t true. In 2014 polls of likely voters, seventy-two percent said that most members of Congress should be defeated in the next election. But stunningly, only twenty-nine percent thought that their own member deserved re-election. In 2024 surveys the percentage who thought their own Representative should be re-elected only reached thirty-six percent. This seems to contradict the common saying that people hate Congress but love their own member of Congress as an explanation for the stunning rate of reelection. This disconnect can be harmonized by recognizing that although most voters think their own representative is bad, they think that the candidate from the other major party is even worse. 

With only two major parties, and the dearth of public awareness of their representative’s behavior (due to rational ignorance), election sanctions are an astonishingly ineffective accountability tool. In later writings Mansbridge further points out that

“the close races with little incumbency that are necessary to make the sanctions model work may also drive away the most public-spirited agents.”

The re-election of Representative Charles Rangel in 2010 with eighty percent of the vote, while under the cloud of confirmed serious ethics violations, showcased the non-efficacy of the elective sanction model of accountability for corruption. He went on to win re-election in 2012 and 2014, before retiring. In a polarized partisan environment, researchers have confirmed a common perception that voters readily forgive corruption and ethical failings among politicians within their partisan “tribe,” and only care about accountability among politicians not in their tribe. This tendency may be less pronounced, but still holds in jurisdictions using multi-party proportional representation.

Electoral accountability is an inherently retrospective tool. Some say it’s like driving while using only the rear-view mirror, since the only sanction that constituents can muster is removing the errant representative after the fact. But the threat of sanction is forward facing. However, the fact that elected representatives deal with a vast range of distinct issues makes this accountability tool incredibly problematic for their constituents to use as a means of guiding their representative’s policy behavior. If my representative votes as I wish on some issues, but contrary to my wishes on other issues, what should I do in the next election? What if the only viable opponent facing my current representative is likely to vote contrary to my wishes on more issues? How can I use the election process to hold my representative to account? How can I use the blunt weapon of the threat of removal to guide my representative?³ Is it only functional in extreme cases?

The second sort of accountability Mansbridge discusses is the selection of representatives who naturally, without external incentives, seek to represent the interests of constituents because they are congruous with their own. Accountability, with regards to policy, rather than corruption, is only an issue when the interests and beliefs of the representatives are different from those of the population they are representing. Accountability through elections is like a leash made out of elastic that has little possibility of actually giving much guidance to where representatives’ go. The fact that unaccountability is a widely discussed problem in modern electoral democracies is evidence that elected representatives do not accurately represent the population. Joel Parker puts it this way: 

“It may be that we tend to think of accountability almost exclusively in terms of electoral sanctions not because elections are particularly effective at controlling the actions of representatives, but because our reliance on elections makes the presence of sanctions that much more necessary.”

The next post will examine how such things as rational ignorance further negate the fantasy of electoral accountability.

1. This is an ineffective “threat” in sortition, as it assumes the representative wants to maintain that role, which would often not be the case in a jury-like situation. If there is no threat, then removing the representative after the act of wrong-doing is also ineffective at guiding behavior.

2. The exceedingly rare exceptions, such as the primary defeat of House Majority Leader Eric Cantor to a little known “Tea Party '' challenger in June of 2014, which garnered massive media attention, distort the impression of the overall trend.

3. In some extreme cases, independent or third-party candidates may surface, and “spoil” the re-election chances of the current representative in the case of plurality winner election rules. The incumbent has been punished, but this can lead to even worse representation for the aggrieved constituents.