If Crowds Are Wise, Why Isn’t Congress?

We’ve all heard of the “wisdom of crowds” especially after James Surowiecki’s 2004 best-selling book by that name and Scott Page’s 2007 “The Difference.” A recent entry into the English dictionary, dating from around 2006, is indeed “crowdsource.” In the last decade, every manner of organization has begun using crowdsourcing technology to do everything from write software toproofread text to design armored vehicles. Facebook famously crowdsources the translation of its pages. Remarkably, Iceland recently crowdsourced amending its constitution.

So why does the US Congress, a crowd of 535, seem so remarkably un-wise?

The reason: Group wisdom flourishes under certain conditions, diversity of thought and independence of judgment, virtually impossible under our current plurality voting and single–seat, winner-takes-all (often gerrymandered) Congressional districts.

The good news: This is relatively easy to fix–without overthrowing the government. In order for crowds to be wise, says Surowiecki, individuals in a group should be independent, decentralized, and diverse. The most important of these conditions, it appears, is diversity–diversity in opinion, skill, outlook, background, depending on the issue in question. Surprisingly, Surowiecki, Lou Hong, Scott Page, and others argue, it is more important than a group’s average intelligence and usually more important than the group’s subject matter expertise.

Diversity’s effect flows from the inclusion of varied perspectives that not only bring in more relevant information, but, just as critically, cancel out individual errors. The varied perspectives furnish the group with different decision-making heuristics (rules of thumb) each more likely to grasp certain aspects of a situation. Moreover, well-known and broad diversity makes it more likely that each individual will honestly voice his/her opinion free of social pressure.

In Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many, political scientist Helene Landemore extends the idea of crowd wisdom to political theory. In order for a group’s reasoning to be wisdom–and not its shadow double groupthink–individuals must independently come to their own judgments as to the merits and demerits of a proposition. She also argues that “cognitive diversity” makes crowds more intelligent than individuals and democracy a wiser form of government than the alternatives. She cites Surowiecki, Hong, Page, and the Marquis de Condorcet, for many of the (mathematical) arguments on group wisdom.

If democracies have shown themselves more successful, or more peaceable, it is due to their tendency to take in more information and take account of more opinions, all thanks to their cognitive diversity. Drawing a distinction between aggregative and deliberative group intelligence, she points out that these two nuances suggest different procedures. If the object is to predict a measurable result, such as the number of jelly beans in a jar, then simple, secret majority rule would yield the best result. If the objective is to solve a complex problem, then deliberation among a sufficiently large number of equally powerful individuals would work better, for example, as in a jury.

“Democracy [stems from] a reverent awareness of human folly,” says Paul Woodruff, philosopher and classicist at UT-Austin, in First Democracy:The Challenge of an Ancient Idea, published in 2005. Unlike many books about Athens that wax nostalgic or sentimental, this one tells the history of politics during the 200 year reign of democracy in a sober yet energetic manner. Democracy emerged the way it did, according to Woodruff, not only because the economic and social conditions favored egalitarianism, but because it built a feeling of group solidarity (essential for maintaining a non-professional army) and at the same time ended class warfare between aristocrats and the masses.

With respect to institutions, he also draws a distinction between the majoritarian and deliberative aspects of democracy. The Assembly, open to the first 6000 male citizens, irrespective of wealth, who showed up on a particular day served the majority rule function. Since deliberation was impossible in such a large, noisy crowd, a Council of around 400 was chosen by lot to deliberate different problems. It was the Council that would draw up measures to be “yayed or nayed” by the Assembly at large. Mob rule was avoided through a liberal judicial system, open to all, that was based on a constitution or “the basic laws.” Anyone in the assembly could sue another for proposing a measure violating the constitution and a judge and jury, selected by lot, would rule on the matter.

Athens, then, ensured “cognitive diversity” through the use of lots and the openness of both its courts and Assembly. Of course, certain public posts had to be based on merit, such as in the treasury or the military. It promoted accountability, through euthunai, whereby all persons upon leaving public office, went through a jury-led settling of accounts. If the person was found to have violated public trust while in office they could be punished or ostracized entirely from the community. Remarkably, even experts had to undergo this open, public test that ensured a wide range of perspectives.

What About Congress?

Leaving aside the more entangled issues of independence and decentralization for a moment, the case regarding lack of cognitive diversity is clear.

In the United States, members of both Houses are elected through large, single-member districts based on a first-past-the-post (FPTP) plurality method where the candidate with the most (not usually majority) votes wins. Moreover, districts are gerrymandered (with explicit approval of the Supreme Court) by the party in power, creating “safe districts” for itself. Single-seat, plurality districts means that a party could win millions of votes and still get zero seats, if none of its candidates finishes ahead in any single district.

This type of voting mathematically favors a single or dual-party system. It also happens to virtually shut out minority parties and, more importantly, minority opinions. In other words, our electoral system virtually eliminates cognitive diversity–what may be not just a prerequisite for a well-functioning legislature, but the very justification for democracy as a form of government to begin with.

The disadvantages of plurality voting are so well known in political science that Congress is one of the rare representative bodies in the world that uses single-member plurality districts. Ever wonder why modern democracies are characterized by multiple parties and “governing coalitions?”

Americans Know This Too

The State of Illinois in 1870 adopted cumulative voting for multi-seat districts to deal with recurring deadlock between Democrats and Republicans in its General Assembly. Cumulative voting in a 3-member district, for example, means that a voter could “pool” her/his votes for one candidate and ignore the other two seats. What’s remarkable about the Illinois experience is how quickly (within a couple of elections) diversity came in. This lasted for over one hundred years, and a proposal backed by then State Senator Barack Obama to revive a similar system was recently made.

Several California counties have approved another method that increases access to minority parties and independents: instant runoff voting, or I.R.V. On election day, each voter ranks the candidates (from most to least desirable) on his/her ballot. When no single candidate receives a majority of first-place votes, the last-place candidate is eliminated, giving his/her votes to the others according to voter preferences. The process is repeated until a majority winner emerges.

San Francisco voters adopted IRV in 2002 and began implementing it in 2005 in city council elections. Similarly, Oakland approved IRV in 2005 and first implemented it in 2010. See FairVote.org (an advocacy group that promotes more inclusive voting methods) for more information, especially on instant runoff voting.

Yes, today Congress is dumb, in cognitive diversity one of the dumbest in the world perhaps. But, it can smarten up rather easily on some simple electoral reforms.

[I do not advocate any particular method of vote counting as the solution to all political problems. In fact, I will argue later that voting in-itself tends to introduce systemic bias. What I am saying is that IRV, cumulative voting, and similar methods would be a significant first step in improving our representative bodies. —AT]