Crossposted from the blog The Hazy Path to which I contribute.

If you believe that democracy is guaranteed to triumph in the end, you might be in for a nasty surprise. You see, the path of democracy, once thought to be on an inexorable upward trajectory, has recently taken a turn for the worse. In 2017, the Economic Intelligence Unit downgraded the USA from a “full democracy” to a “flawed democracy” for the first time ever. This means a country with free elections, but weighed down by weak governance, corruption and low trust in the political system. And this trend isn’t confined to the US: it’s a global wave. In that same report, 72 countries had a lower democracy score than the previous year, while only 38 improved on their ratings. This process, known as democratic backsliding, has been steadily occurring around the world since 2006. And the only way to halt this process is to re-establish the case for a strong, truly representative democracy.

However, defenders of democracy often find themselves on unstable ground, when asked the question “why is democracy so great?” After all, there are plenty of non-democratic states which seem to be doing alright. There’s China with its centralized planning by a one-party regime, Singapore with its ever-present surveillance state, and the United Arab Emirates with its absolute monarchy and limited political freedoms. The share of global GDP produced by non-democracies is steadily rising, now exceeding one third. What’s more, governments across the world are seeking to emulate the “China model” of “stable, if repressive politics and high-speed economic growth”. So how can we prove that democracy is the natural endpoint of a country’s development, rather than merely one option among a vast array of possible governance models?

Well, to prove to the world that democracy is good, we must first understand why it’s good. In other words, what goal is democracy trying to achieve?

I believe that democracy is best used as a means to an end, rather than a sacred principle with intrinsic value. I will attempt to convince you that this view of democracy, the “instrumental” view, is more useful in building better government and ultimately a better society. But first, let’s define some terms.

The conflicting views

One theory of democracy’s purpose is to elect the “right” leaders. In this view, questions such as “Who is best equipped to lead this nation?” have a correct answer, and democracy is merely the most effective way to find that answer. This is the “instrumental” theory of democracy, as it treats democracies as a means to an end. Their existence is justified by the effectiveness of their outcomes.

The other theory states that democracy’s purpose is to respect the will of the people, and give them the power which is rightfully theirs. According to this theory, the “right” leader is simply defined by what the voters choose. This is known as the “procedural” theory, as it finds democracy’s value within the democratic procedure itself. The procedural view talks in the language of rights and justice, rather than effectiveness and success.

The jury theorem

Imagine that the New York Yankees have a slight edge over the Boston Red Sox. Let’s say that they have a 51% chance of winning each head-to-head match-up. If they only compete once, the Yankees have a 51% chance of winning, while the Red Sox has a slightly lower 49% chance. So far, so obvious. But if they compete in a best-of-seven series, the Yankees’ odds of winning overall will be slightly higher. Their advantage will extend a little further each additional time the teams clash. If all you need to win overall is to be the victor in a majority of games, even the slightest advantage approaches near-certain victory when enough games are played.

This is Condorcet’s jury theorem, so named because it demonstrates the increased accuracy of jury decisions as the jury expands. Arguably, the theorem is even more relevant to democracy, because while a jury could only consist of a few dozen people at most, elections have thousands or even millions of participants. Therefore, if each voter picks the right candidate slightly more often than 50%, democratic elections will be almost guaranteed to choose the correct leader. That’s a colossal IF, but more on that later.

Why Instrumentalism?

Not all decisions are suitable for democracy.
There are surely some issues upon which we must defer to the experts: let me explain. Several awful economic ideas, routinely denounced by everyone who knows what they’re talking about, enjoy great popularity among the public. For example, consider the gold standard. The public often remember it fondly, along with the halcyon era of the Bretton Woods agreement. Columnists and politicians alike frequently toy with the idea of returning to it. However, while the voting public can be persuaded by nostalgia, a diverse panel of the world’s top economists unanimously agreed that the gold standard is a bad idea. Other examples of experts’ universal agreement on a controversial political topic include: the benefits of free trade agreements far outweigh the costs, and government policy is not responsible for high gas prices. And yet, the precise opposites of these viewpoints frequently poll well with the public. Economic concepts are often counter-intuitive, which is why some positions must be appointed rather than elected. Handing over the reins of the economy to a misinformed, nonexpert public would end in disaster. Makers of economic policy must be independent from the government, not subject to the pressures of election cycles. This is just one example of how democracy should be used as a tool where fit, and cast aside when unfit.

It incentivises us to improve our democracy.
The procedural view holds that the people’s decision is correct by definition. The winner of the election is always the right leader, as long as they were selected by a free and fair election. However, this view can quickly slide into a sense of complacency. It seems to imply that we ought not question why people made the choices they did: the people have spoken, so what business do we have interrogating their choices? We saw this kind of faulty thinking in the aftermath of the Brexit referendum. Any scrutiny of the misleading Facebook ads that presumably swayed voters was quickly shut down, by the argument that it was somehow patronising to suggest that voters had been led astray. The instrumentalist view, however, holds that democracy has the potential to be the best political system, but it isn’t guaranteed to be. Only an adequately informed population can distinguish valid arguments from fallacies and fictions. Therefore, our media ecosystem must be principled and sound: not as a nice-to-have, but to ensure that democracy itself works.

Remember Condorcet’s jury theorem from earlier? It also works in reverse. Suppose that false information spread by a nefarious media source guides people away from the best candidate. If the voters’ chance of selecting the correct candidate falls below 50%, then adding more participants simply makes the problem worse. If democracy is a tool, we must sharpen it to the best of our ability.

There are plenty of ideas on how to make the media serve democracy, including “news vouchers”. This would be a democratic process for allocating government subsidies to the media. Each person would be allocated news vouchers that they can exchange for a subscription at various news outlets. This could help to revive local news outlets, but crucially, it could raise the bar for truth. An independent overseer could introduce rules that news outlets would have to follow to be eligible for funding from news vouchers. Though very interesting, this idea isn’t the main point of this article. The main point is that these ideas seem far less urgent if we believe democracy is an inherent, rather than an instrumental, good. If we believe that democracy is good no matter what, we are lulled into a sense of complacency, with the possibility for a dysfunctional, deceptive media to steadily distort our democracy. But if we believe that democracy isn’t guaranteed to succeed, we must dedicate much more time and energy to nurturing the sources of information that help people make their decisions.

The opposing views

These are, in my reckoning, the most commonly made arguments for the procedural view of democracy. (If you believe I’ve got this wrong, or left something out, I’d be glad to hear: just leave a comment below.)

Legitimacy. Provides a principled defence of democracy not reliant on any subjective definition of “good outcomes”.

Fairness. Democracy is the fairest way to make decisions among people who see each other as free and equal.

In my view, the legitimacy point falls short. It’s not clear why we need a defence of democracy that works in all situations. Democracy is a good system of government because, in general, it gives us more of what we want (accountability, economic development, freedom of speech) and less of what we don’t want (wars, oppression, abuse of human rights). If a situation arises in which democracy is not the optimal tool, we should have no qualms about casting it aside. And for the point about subjective definitions, it’s certainly true that a good outcome for some might not be a good outcome for others. But this problem affects democratically made decisions just as much: a majority of a nation’s citizens may vote in their own self-interest, at the expense of minority groups, other nations, or future citizens. This is why democracy must not be treated as a one-size-fits-all solution.

As for the point on fairness, why is fairness inherently good? If there existed a deeply unfair system which nonetheless provided better outcomes for the population, I would switch to it in a heartbeat, and I think most people would too. Fairness is only valuable to the extent that it improves people's lives. It may be easy to conflate the two, since fairness does improve lives in the vast majority of cases. But we must recognise that in some cases, an unfair process may produce good outcomes. For example, consider a short-staffed, hastily erected emergency hospital in the aftermath of a natural disaster. The triage ward is full of injured patients, and their conditions are deteriorating by the minute. Unfortunately, the limited resources make treating all of them impossible. It certainly doesn’t seem fair to start treating people at random, or based on some arbitrary characteristic such as which side of the room they entered. But in the time it would take to calculate in detail how to save the most lives, it may already be too late for some. Therefore, an unfair process may be the best way to achieve the desired outcome: saving more lives.
To give a more on-topic example, imagine an unstable coalition government is ruling a country, when it is suddenly attacked by a rogue state. This coalition government only has overall control in Parliament by a hair’s breadth, and is itself composed of multiple parties of similar sizes. It would no doubt be unfair for the largest party within the coalition to immediately receive executive control: other parties may have different views on armed combat. However, this government’s relentless process of back-and-forth consensus building fails to meet the strenuous demands of war. 
Both of the above examples go to show that while fairness usually is synonymous with good outcomes, there are exceptions. Fairness should therefore not be held as a guiding principle: indeed, quality of outcome is the only guiding principle that matters.

If this article must be distilled down to a single coherent claim, let it be this: “Democracy is a tool, not a sacred process. It should be used and sharpened when appropriate, and cast aside when unsuitable.” And that’s a hill I’m willing to die on. Though perhaps it’s not the most inspiring quote to write on my tombstone.