How to fix democracy: empower the best informed voters

What if 80% of the public is well informed, but they vote 40% right-wing and 40% left-wing? Then it is the most ignorant 20%, voting somewhat randomly, that will determine every election.

What if 80% of the public follows the political issues and is well informed? What if 40% have rational reasons to vote for the right-wing, and 40% have rational reasons to vote for the left-wing? Then it is the most ignorant, badly informed 20%, voting somewhat randomly, that will determine every election. And therefore elections become random events. So if we want better elections, we should want to limit the influence of that 20%.

Achen and Bartels describe the current situation:

Democracy for Realists, 2016

Why Elections Do Not Produce Responsive Government

By Christopher H. Achen & Larry M. Bartels

Page 312

The result is that, from the viewpoint of governmental representativeness and accountability, election outcomes are essentially random choices among the available parties – musical chairs. Elections that “throw the bums out” typically do not produce genuine policy mandates, not even when they are landslides. They simply put a different elite coalition in charge. This bloodless change of government is a great deal better than bloody revolution, but it is not deliberate policy change. The parties have policy views and they carry them out when in office, but most voters are not listening, or are simply thinking what their party tells them they should be thinking. This is what an honest view of electoral democracy looks like. It is a blunder to expect elections to deliver more.

We all understand that elections are supposed to show “the will of the voters” and thus deliver to the ruling group “the mandate of the masses.” But this is a romantic view. In the early 20th Century economists did their best to argue that elections really did work correctly, and manifested the public’s desires:

Page 23-25

The most systematic and sophisticated instantiation of the populist ideal is the “spatial model” of voting and elections. Although the model has been a mainstay of political science for the past half century, it was originally formulated primarily by economists – perhaps because the intellectual framework of economics meshed naturally with “the liberal view” that “the aim of democracy is to aggregate individual preferences into a collective choice in as fair and efficient a way as possible,” as David Miller (1992, 55) put it. Miller acknowledged in a footnote that some readers might object to the limited focus on “one strand of liberalism – the importance it attaches to individual preferences and their expression”; however, he argued that that strand “prevails in contemporary liberal societies, where democracy is predominantly understood as involving the aggregation of independently formed preferences.” Thus, in effect the goal of the spatial model was to give mathematical form to the folk theory of democracy.

But then in 1953, the great economist Kenneth Arrow largely demolished this view with his Impossibility Theorem, proving that elections could not work they way we want them to work (especially if the elections elect just one person). The end result of rank voting is a result that is simply too coarse-grained to represent the fine-grained will of the people. Fixing this problem will take a few changes, which we will explore in a few essays, but for now, let’s focus on how we can limit the damage done by the 20% of poorly informed voters.

There is a simple way to fix democracy

Achen & Bartels spend most of 400 pages proving that the “folk theory” of democracy does not work, but they don’t spend any time explaining what parts of the theory do work. The statistics they quote seem to suggest that something like 10% to 20% of the population does engage in most of the behaviors suggested by the folk theory, which is to say, some portion of the public follows the news, knows the issues, and remembers what each politician does over the course of many years. And there is probably a much larger group that follows politics somewhat, though not perfectly. But it is also clear, there is a good 20% to 30% of the public that is deeply ignorant of politics, and when they vote, they turn elections into random events, in which the stupidest people decide the election.

Then there is the evil of pandering. Many voters only pay attention to politics during the months leading up to elections, which allows a politician to spend years undermining the needs of his/her supporters — so long as that politician is seen doing good things for his/her supporters in the months before the election, that politician can still win re-election, and then they can go back to undermining the needs of his/her supporters. A party that is devoted to austerity can promise lavish spending a month before an election, then win, then go back to austerity as soon as they’ve won. This is how the Tories have managed to stay in power in Britain in recent years. Pure dishonest pandering.

And there is the great problem that many voters are unable to understand what policy responses are the best, when facing a crisis:

Page 205

A crucial feature of this brief litany of electoral responses to the Depression is that the ideological interpretation customarily provided for voters’ reactions in the United States does not turn out to travel well. Where conservatives were in power when the Depression hit they were often replaced with liberals or socialists, as in the United States and Sweden. But where relatively leftist governments were in power during significant downturns they were often replaced with more conservative alternatives, as in Britain and Australia. Where the existing party system was oriented around noneconomic issues, as in Ireland, voters rejected the “ins” and replaced them with “outs” whose policy positions cannot even be sensibly classified in left-right terms. Where the timing of elections forced more than one major party to stand for reelection during the worst years of the Depression, as in Canada and Sweden, voters seem to have been perfectly willing to reject both in turn. Where complex coalition politics diffused responsibility, as in France and Germany, discontented citizens turned to unstable coalitions, to fringe parties, or to the streets. Simply put, there is no consistent ideological logic evident in voters’ responses to the Depression when we look beyond the American case. When voters got a chicken in every pot at election time, they usually liked the incumbent party’s ideology just fine, whatever it happened to be. But when incomes eroded and unemployment escalated, they became ripe for defection to anyone who promised to bring home the poultry.

And yet, this suggests an easy fix for some of the problems that Achen & Bartels lay out. While they focused on the weakness of the 51%, they accidentally make clear the wisdom of the 10% or maybe the 20%. Achen & Bartels focus on the fact that large majorities often fail quizzes about basic political and economic facts, yet in every one of their stories there is a solid 10% or 20% that gets everything correct.

Therefore we simply need to elect these people to another level of representation, some group in-between the public and the legislature. I’m going to call these people the Notables.

Let’s assume we want to elect about 5% or 6% of the population to serve as Notables. Let’s assume we divide the country into districts of 100,000. That’s 75,000 adults. 5% would be 3,250.

Let’s assume a system where:

1. elections occur every month (some interesting properties of monthly elections are described here)

2. 100 people can be nominated each month, perhaps drawn from a queue that anyone can sign up for

3. every voter gets 10 votes

4. the top 20 vote getting candidates are elected

5. they serve a term of 15 years

So that gives 3,600 Notables, about 5.5% of the adults in the district. You can play around with the numbers a bit, I don’t think the exact numbers matter. The goal is to get a broadly representative group that can speak for the community, but within that community, we want to find the people who have some real interest in politics. Presumably, the kinds of people who would sign up to be candidates to be Notables would be people of above average motivation. And at this level, motivation matters more than raw intellect. We simply want to reduce the influence of the truly ignorant. We ultimately want elections to be less random.

Think of your personal Dunbar Number. Of the 160 adults that you know best, can you think of 8 who you think would do a good job of electing the legislature? Would the world be a better place if those 8 were empowered to chose the legislature, rather than have the entire 160 do so?

For all of the hassle of introducing a new layer of representation, there would be some advantages. Right now, political activism is informal and without recognition. In theory, anyone can become a political activist, in reality, only a small handful of citizens ever become politically active. Electing Notables would offer formal recognition for political activism. So far, we’ve never had a moment when as much as 5% of adults participated in political activism, so electing 5% of adults should mean that every adult who is interested in politics could become a Notable. This applies to left-wing and right-wing activists, there is no bias here.

Electing 5% of the adult population means that 1 out of every 20 adults will be Notables.

That ratio is low enough that every citizen would know multiple Notables, personally. It is possible that such fine-grained representation will increase the legitimacy of the system.

Of course, the Notables themselves are relatively powerless. In the USA, with a population of 330,000,000 people, and an adult population of 240,000,000 the Notables would consist of 12 million people. That’s a large, diffuse group, too large to engage in anything like legislative action. All that we can say is, this is the group that shows interest in public affairs, so people in Congress would have to take them much more seriously than they might take some random citizen. More so, this would be the group that elects Congress, so Congresspeople would suddenly be facing an informed electorate, rather than the apolitical and often ignorant public that makes up the typical 51%.

Crucially, politicians would no longer be able to pander.

Politicians would no longer be able to undermine the interests of their supporters, year after year, but then do something generous a few months before the election, and thus win re-election. They would now be elected by the kind of people who follow the news and who remember bad things the politician did 3 years ago. Therefore the politicians would always have to be on their best behavior.

There are some crucial issues to be addressed about electing such a group of Notables, and whether such a system would be truly representative, or would there be ugly interactions with existing advantages held by the wealthy, or held by those of particular racial groups. Such a large subject deserves its own essay, so I’m going to postpone that discussion till later in this series.

At times, Achen & Bartels suggest that political opinions move from the party to the public, never the other way around. But who is the party? There must be some limited group that makes up the in-group, and it is not clear if this might be 5% of the population or 0.5% of the population. This needs further research, but I suspect the 5% of Notables would not be nearly as passive as the 51% that currently wins elections. A system of Notables means the 51% of adults no longer elects Congresspeople, but rather, it’s 2.51% (half the Notables) that elects Congresspeople. And this smaller, better informed, better motivated group is likely to do a much better job of holding Congresspeople to account. If we want elections to be part of our system of “checks and balances” then we need a better motivated electorate to do the work of checking and balancing. Sifting the highly motivated and giving them an official status is the obvious way to do this.

Again, the goal is to limit the damage done by those voters who are poorly informed. Since 50% of all people have below-average intelligence, and some people have far-below-average intelligence, we should worry about how they affect the system.

At 5% we are talking about very broad representation — the Notables would simply be those natural leaders that everyone admires in a community. We are asking people to elect their friends, and then trust their friends to make decisions about who should serve in the legislature. If a person does not have any friends, and does not trust anyone, then we have an especially urgent need to limit their impact on the system, as such people (anti-social and distrustful) are unlikely to support democratic norms. The average citizen does not follow politics closely, but they have several friends who do, and they should be willing to empower those friends to make decisions about the legislature.

One nice feature of such intimate elections is that it engages a non-partisan part of people’s brains. They are not being asked to vote against the enemy team, they are only being asked to vote for their friends, people in the community who they know and trust. This is a simple filter, yet it is likely to do much good. (I’ve worked for 20 years in the tech industry and I’m aware, when we try to boost a signal and rescue it from a great deal of noise, sometimes the simplest filters work the best.)

Electing an intermediate group, in-between the public and the legislature, should fix many of the problems that Achen and Bartels identify with the “folk theory” of democracy. In particular, it helps to limit the damage done by the least informed voters.