I – II – III – IV – V – VI – VII – VIII – IX – X – XI – XII – XIII – XIV – XV – XVI – XVII – XVIII – XIX – XX – XXI – XXII – XXIII – XXIV – XXV – XXVI – XXVII
[Single-page view]

Again, none of this is to say that the whole reason democracy is preferable to autocracy is that voters are all perfectly enlightened decision-makers themselves. In fact, if there’s one thing critics of democracy are right about, it’s that voters don’t typically tend to be very well-informed about policy at all. As Brennan writes:

Sixty-five years ago, researchers began studying what voters know and how they think. The results are depressing. The median voter knows who the president is, but not much else. Voters don’t know which party controls Congress, who their representatives are, what new laws were passed, what the unemployment rate is or what’s happening to the economy. In the 2000 U.S. presidential election, while slightly more than half of voters knew that Al Gore was more liberal than George W. Bush, they did not seem to know what the word “liberal” means. Significantly less than half knew that Gore was more supportive of abortion rights, was more supportive of welfare-state programs, favored a higher degree of aid to blacks or was more supportive of environmental regulation.

Of course, I suspect people nowadays would have a much clearer understanding of the difference between liberalism and conservatism, so this specific argument might not be as strong today as it was 25 years ago. But still, the general point remains: When it comes to the finer points of policy, most voters aren’t very knowledgeable. Sure, they may have some policy preferences, but those preferences will tend to be very general broad-stroke kinds of desires without much theoretical grounding – or if they do have strong opinions on specific policies, their attention will typically be limited to just a few pet issues and not much else. And really, this shouldn’t surprise us; after all, most people are working full-time jobs and raising families and so on, so we shouldn’t expect them to have the time, the resources, or the inclination to learn all the nuances of every political issue and become elite policy experts on top of everything else. They simply don’t have the means to develop the kind of fine-grained expertise that they’d need to be able to make informed decisions on every single detail of every government policy themselves – which is why, whenever a particular jurisdiction does attempt to govern itself by putting everything directly in its citizens’ hands and having every issue decided by voter referendums (the so-called “direct democracy” approach), it tends to produce the kind of political dysfunction that has become all too familiar in places like, for instance, California. As Zakaria explains:

In many ways California is the poster child for direct democracy, having experimented most comprehensively with referendums on scores of issues, big and small. [But] if California truly is the wave of tomorrow, then we have seen the future, and it does not work.

[…]

No one disputes the facts. In the 1950s and early 1960s California had an enviable reputation as one of the best-run states in the union. “No. 1 State,” said a Newsweek cover from 1962, “Booming, Beautiful California.” Time agreed; its title read “California: A State of Excitement.” There was much to be excited about. The state’s economy was booming, and with moderate tax rates it had built extraordinary and expanding public resources, from advanced highways and irrigation systems to well-run police forces to breathtaking parks and zoos. The state’s crowning achievement was its world-class public-education system, which started in kindergarten and ended with the prestigious University of California campuses. Californians seemed blissfully content, annoying intellectuals in the cold, damp Northeast to no end. (“All those dumb, happy people,” said Woody Allen.) But for the rest of the world, sunny, prosperous, well-managed California symbolized the dazzling promise of the United States. California was the American dream.

California today is another story. In the spring of 2001 California was plunged into blackouts and electricity shortages that reminded me of India. (In fact they were worse than anything I experienced growing up.) Sure, California is home to Silicon Valley and Hollywood, two of the greatest centers of American industry and creativity. But that is its private sector. Its public sector—indeed its public life—is an unmitigated mess. The state and local governments struggle each year to avoid fiscal crises. Highways that were once models for the world are literally falling apart, and traffic has become both a nightmare and an expensive drag on productivity. In the 1950s California spent 22 percent of its budget on infrastructure; today it spends barely 5 percent. Public parks now survive only by charging hefty admission fees. The state’s education system has collapsed; its schools now rank toward the bottom in the nation when measured by spending or test scores or student skills.

The University of California system has not built a new campus in three decades, despite the fact that the state’s population has doubled. And yet, as the veteran journalist Peter Schrag points out in his penetrating book Paradise Lost, it has had to build twenty new prisons in the last two decades. In 1993 the Economist concluded that the state’s “whole system of government was a shambles.” Three years later the bipartisan Business-Higher Education Forum, which includes corporate executives and education leaders, reported that without major change “the quality of life will continue to decline in California with increasing transportation problems, rising crime and social unrest, and continued out-migration of business.” And this was written at a time when the U.S. economy was at a thirty-year high. The best evidence of California’s dismal condition is that on this one issue, both right and left agree. Echoing Schrag, who is a liberal, conservative commentator Fred Barnes explained in a cover story for the Weekly Standard that the state’s government had stopped working: “California has lost its lofty position as the state universally envied for its effective government, top-notch schools, and auto-friendly transportation system.”

Not all California’s problems can be traced to its experiment with referendums and initiatives. But much of the state’s mess is a result of its extreme form of open, non-hierarchical, non-party based, initiative-friendly democracy. California has produced a political system that is as close to anarchy as any civilized society has seen. Consider the effects of the recent spate of initiatives. After Proposition 13 passed [in 1978, limiting the state’s ability to raise taxes], the state passed dozens of other initiatives, among them Proposition 4 (which limited the growth of state spending to a certain percentage), Proposition 62 (which requires supermajorities in order to raise taxes), Proposition 98 (which requires that 40 percent of the state budget be spent on education), and Proposition 218 (which applied to local fees and taxes the restrictions of Proposition 13). Yet the state legislature has no power over funds either, since it is mandated to spend them as referendums and federal law require. Today 85 percent of the California state budget is outside of the legislature’s or the governor’s control—a situation unique in the United States and probably the world. The vast majority of the state’s budget is “pre-assigned.” The legislature squabbles over the remaining 15 percent. In California today real power resides nowhere. It has dissipated into the atmosphere, since most government is made via abstract laws and formulas. The hope appears to be that government can be run, in Schrag’s words, like “a Newtonian machine immune to any significant control or judgment by elected representatives. This not only turns democracy into some fun-house contortion of the ideal but makes it nearly impossible to govern at all.

Even with referendums dictating what they do, politicians are still required to translate these airy mandates into reality. The initiatives have simply made this process dysfunctional by giving politicians responsibility but no power. Experiences with referendums in places far from California confirm that this is not a problem unique to the golden state. The Connecticut Conference of Municipalities (CCM) determined that 72 of the state’s 169 municipalities held referendums to approve the budgets proposed by their city executive. Of those 72, 52 had to hold additional referendums, often more than once, because the proposed budgets were rejected. Most of these referendums required local officials to cut taxes and yet improve services. “You have to be a magician to accomplish those dual purposes,” complained James J. Finley, the CCM’s legislative services director.

The flurry of ever-increasing commandments from the people has created a jumble of laws, often contradictory, without any of the debate, deliberation, and compromise that characterize legislation. The stark “up or down” nature of initiatives does not allow for much nuance or accommodation with reality. If in a given year it would be more sensible to spend 36 percent of the California budget on schools instead of the mandated 40 percent, too bad.

As another unintended consequence, the initiative movement has also broken the logic of accountability that once existed between politicians and public policy. By creating a Byzantine assortment of restrictions over the process of taxing and spending, the voters of California have obscured their own ability to judge the performance of their politicians. When funds run out for a particular program, was it that the legislature allocated too little money, or that local communities spent too much, or that statewide initiatives tied their hands? You can imagine the orgy of buck-passing that must ensue, given California’s 58 counties, 447 cities, and over 5,000 special districts. Lack of power and responsibility inevitably produces a lack of respect. California’s state government and its legislature have among the lowest public-approval ratings among American states. Having thoroughly emasculated their elected leaders, Californians are shocked that they do so little about the state’s problems.

The moral of the California story, in short, is that simply handing the reins of government directly to the voters and saying “Good luck, you all figure it out” is not a recipe for success. Ordinary citizens just aren’t equipped to be able to fully understand all the planning and budgeting implications of every policy proposal and how all the various moving parts of government fit together; the best they can do is express their general preferences for which direction they want government policies to go in (or at least, that’s the most that can reasonably be asked of them).

So does this mean that democracy itself is a fatally flawed idea? Not at all. True, direct democracy may not be workable on a large scale – but direct democracy isn’t the only kind of democracy that exists. The fact that it has so many difficulties is exactly why we also have the concept of representative democracy, in which voters don’t vote directly on every individual issue themselves, but instead vote for people who can in fact make it their full-time job to gain a reasonable understanding of all the issues and then act accordingly on behalf of the people they represent. By delegating political decision-making responsibility to elected representatives in this way, voters can get some of the same benefits that having benevolent autocrats in charge would supposedly bring – i.e. having leaders who know all the ins and outs of government, have teams of budget experts and policy wonks who can help them make more informed decisions, and so on – but with the key difference being that these elected leaders will still ultimately be accountable to the voters themselves instead of just having the power to do whatever they want. Their role can simply be to implement whatever policies they think will produce the kind of outcomes their constituents will approve of (or at least won’t disapprove of) – and then, depending on whether or not they’re successful in that task, voters can either re-elect them or replace them with someone else. As for the voters themselves, they won’t be expected to understand all the fine-grained details of every single political issue – but the benefit of this system is that they won’t have to understand all those details; their most important job will simply be to notice whether or not they’re satisfied with the way things are going (relative to the alternatives), and then register those sentiments by voting. And that’s one area where they really will have more fine-grained knowledge than their political leaders will; they might not be able to understand all the minutiae of the political process better than their representatives do, but they will have a more detailed understanding of how satisfied they are personally with the current state of things. As Brennan writes (citing Thomas Christiano):

Christiano [describes representative democracy as] a sort of halfway point between [direct] democracy and [rule by enlightened elites]. He begins by noting that it’s unrealistic to expect voters to have sufficient social scientific knowledge to make good choices at the polls:

It is hard to see how citizens can satisfy any even moderate standards for beliefs about how best to achieve their political aims. Knowledge of means requires an immense amount of social science and knowledge of particular facts. For citizens to have this kind of knowledge generally would require that we abandon the division of labor in society.

Christiano believes the typical citizen is competent to deliberate about and choose the appropriate aims of government. For citizens to know the best means for achieving those aims, however, they would have to become experts in sociology, economics, and political science. They are not competent to make such determinations. Christiano’s proposed solution is to create a division of political labor: “Citizens are charged with the task of defining the aims the society is to pursue while legislators are charged with the tasks of implementing and devising the means to those aims through the making of legislation.”

Christiano argues, and I agree, that this regime qualifies as a type of democracy. Fundamental political power is still spread evenly among citizens. Under Christiano’s proposal, the legislators have only instrumental authority. They are administrators more than leaders.

As an analogy, consider the relationship of a yacht owner to the yacht’s captain. The owner tells the captain where to go, but the captain does the actual sailing. While the captain knows how to steer the boat and the owner does not, the owner is in charge. The owner can fire the captain, and as such the captain serves the owner. Christiano might contend that in the same way, under his proposal the legislators serve the democratic electorate. While the legislators set laws that the democratic body must follow, the democratic body told the legislator what direction these laws must go in.

And this kind of division of political labor makes sense, for the same reason that it makes sense for, say, a corporate CEO to delegate certain tasks to their subordinates. A CEO obviously can’t do every task that their company is responsible for all by themselves – developing new products, crunching payroll numbers, responding to customer service calls, etc. – so the most sensible thing for them to do is to delegate those responsibilities to people who can specialize in those particular tasks. Then the CEO can observe the results, and if they’re unsatisfactory, can replace those specialists with more effective ones. It’s an efficient system precisely because it doesn’t put every little responsibility in the CEO’s hands. And the same is true for democratic governance. If we tried to make voters responsible for directly handling every single aspect of governance themselves, they wouldn’t have time for anything else – jobs, families, etc. – that society requires in order to function. By delegating matters of governance to designated specialists (i.e. elected representatives), voters can free themselves up to specialize in different tasks of their own in the private sector – and together, everyone can achieve much better results than they ever could have without this division of labor. Astra Taylor puts it this way:

Most of what government does is a mystery to the average person and no one, not even the most astute legal experts, comprehends all the innumerable and intricate laws that bind us. What’s more, the principles of free choice and citizen consent could quickly become unwieldy if taken to an extreme. It’s unclear, for example, what percentage of decisions affecting our communities we should be expected to participate in. (I, for one, don’t mind letting people more knowledgeable than I determine how best to provide electricity or decide which potholes to fill.)

This is another way of articulating the difference between direct and representative democracy: by choosing representatives, we are, in a sense, choosing not to choose (which is something we do increasingly these days, outsourcing more and more of our daily decision making to, for example, recommendation engines or GPS).

Just as a quick side note, another noteworthy implication of this principle is that in some cases, a particular task or project might be so complicated or technical that even the representatives whose whole job is to work on policy might not fully understand all the relevant details – so in such cases, it may be justifiable for those representatives to themselves appoint people who can specialize even more narrowly in some specific area (i.e. bureaucrats, judges, etc.), and delegate those esoteric tasks to them. These appointees, while not directly accountable to voters, will be accountable to elected officials in much the same way that those officials are themselves accountable to voters; the elected officials will provide them with a particular set of aims to be pursued, then leave it to them to figure out how exactly to implement those aims – and depending on whether they’re able to adequately do so, they may then be allowed to either retain their position, or else be replaced by someone more capable. They’ll still be accountable to voters in the ultimate sense, but in practical terms this accountability will be one step removed; instead of having their performance evaluated and judged directly by voters, they’ll be evaluated and judged by elected representatives, who will then themselves be judged by voters based on their performance. The bureaucrats, in other words, will be acting as sort of meta-representatives – and if they hire staffers of their own to handle the finer details of their tasks, then those staffers will be sort of meta-meta-representatives, and so on. By delegating and sub-delegating tasks based on how technical and/or esoteric they are, the whole system will ultimately function more effectively – again, much like how any large company or organization works better when there are different levels of employees handling different tasks. But won’t this kind of arrangement mean that voters will have a harder time evaluating the actions of the bureaucrats furthest down the chain and holding them accountable? Yes, it will – but as Garett Jones points out, that’s not necessarily a bad thing, because the further down the chain the bureaucrats are, the more likely it is that the government officials overseeing them will be in a better position to hold them accountable than ordinary voters will be:

One of the folk arguments for electing government officials is “accountability.” Citizens, the story goes, need to be able to hold elected officials accountable, and one way to hold them accountable is to retain the right to fire and replace them. But in the case of [for instance] city treasurers, it’s easy to measure (much of) the job the treasurer is doing: just look at the interest rate on the city debt. But even when such a key quality index is so easy to measure, voting citizens do an awful job of keeping the city treasurer accountable. The better option is to let other city officials—the elected mayor, the city council, or maybe the appointed city manager—pick a treasurer and then keep an eye on the job she’s doing. Those city officials will surely notice if the treasurer is saving the city over $200,000 a year, even if voters are too preoccupied watching cat videos to do the job.

Again, this is one of those cases where it can make more sense for voters to delegate a particular task to their representatives than to try and handle it themselves. Overseeing bureaucrats like city treasurers isn’t something that most voters are especially well-equipped for, so they can simply elect someone who is better equipped to do so on their behalf. So then how can we tell which tasks require less of this kind of direct accountability to voters, and which require more? Jones answers this by citing Eric Maskin and Jean Tirole:

Nonaccountability [or rather, accountability that’s a step or two removed from the electorate] is most desirable when (a) the electorate is poorly informed about the optimal action, (b) acquiring decision-relevant information is costly, and (c) feedback about the quality of decisions is slow. Therefore, technical decisions, in particular, may be best allocated to judges or appointed bureaucrats.

–Eric Maskin and Jean Tirole

[In other words,] when it’s crucial to get the technical details right and when the policy debate is less about values and more about facts and competent execution, that’s likely a good opportunity to delegate power to unelected bureaucrats.

The distinction he makes here between policy debates based on values and policy debates based on technical details brings us back to one of our key points, which is that when we talk about voters knowing enough to be able to judge whether their representatives are adequately representing them, that’s not necessarily the same thing as saying that every voter will always know exactly what they want on every issue. After all, as Brennan pointed out earlier, a lot of the time voters won’t even have certain issues on their radar at all, so they’ll be functionally indifferent to competing candidates’ stances on them. If there are two candidates running for office, for instance, and the main difference between them is just their position on some minor regulatory question or obscure tax issue, most voters probably won’t put all that much effort into developing an informed opinion and picking a side. That being said, the flip side of this is that if there’s a situation in which one of the candidates wants to, say, triple everyone’s taxes, or criminalize homosexuality, or ban all guns (or, for that matter, forcibly convert the entire country to their new religion, or destroy everyone’s savings so they can print new bank notes that are all multiples of nine), then that’s the kind of thing that voters absolutely will take notice of and respond to. Voters might not always know exactly which set of policies they’d like best, but they don’t have to know a whole lot to know what they don’t like and won’t put up with – and fundamentally, that’s what representative democracy is really all about. As David Frum explains in a podcast interview, it’s not so much a matter of trying to ascertain voters’ exact preferences on every issue and then do exactly what they want (since that will be effectively impossible in many cases); it’s more about offering them different sets of competing policies, seeing which ones they object to, and then sticking with the ones they seem to be okay with:

DAVID FRUM: I don’t think democracy is about finding out what the people want, because there are too many of them; it’s meaningless. But what you can find out is what the people will consent to. The parties are competing for consent – not for direction, but for consent.

[…]

EZRA KLEIN: Say a word here on the distinction between what they want and consent. So consent is what they will refrain from vetoing?

DAVID FRUM: Yes. Yes. One of the things that populists say is, “I’m going to go out on the hustings, and I’m going to say to the people, ‘Do you want this? Do you want that?’ The people want this.” I think that’s just an illusion. What happens is, politically active groups compete to offer solutions – which they sometimes believe in for objective reasons, and sometimes… I don’t think anybody ever reasoned themself into supporting ethanol. But you have your ideas about what should happen, and then you go out and you try to mobilize people to give you permission to do it. But that’s all you get; you’re not actually executing some will of the masses, because you can’t aggregate the preferences of tens of millions of people in a meaningful way. What you can do is aggregate their permission – that they give you a mandate, go ahead and try it; if it seems to work, you’ll get another four years; if it doesn’t seem to work, you don’t get the four years.

In other words, voters might not always be willing or able to formulate entire political platforms all by themselves – but they don’t have to; that’s not their role. In a representative democracy, their job is just to judge the results of their elected representatives’ policies, and if they don’t like them, to weed them out in a kind of Darwinian selection process. Critics of this process might argue that relying on voters’ judgments in this way can never work because their ignorance will largely prevent them from being able to know whether they’re really getting good results from their political leaders or not – but if anything, this is more of a problem with non-democratic systems than with democratic ones. After all, under more autocratic forms of government, there’s typically little to no reason for ordinary citizens to pay much attention to politics in the first place (since they can’t have any effect on it regardless), so autocratic leaders can more easily take advantage of their citizens’ relative ignorance of public affairs and act in ways that harm their interests without them ever realizing it. By contrast, the handy thing about having a democratic system, in which candidates actually have to compete with each other to win people’s votes, is that this competition itself serves as a mechanism for bringing to light unpopular policies that voters might otherwise be unaware of. If a candidate running for office happens to have some kind of misalignment with voters’ preferences – i.e. if they have some secret fault (ideological or otherwise) that voters are largely unaware of, but which the opposing party knows voters would hate if they were more aware of it – then it’s all but guaranteed that that opposing party will do everything in its power to bring it to voters’ attention, making it a central campaign issue and hammering on it until voters who might otherwise have been ignorant of it have had it firmly ingrained in their consciousness.

Of course, none of this guarantees that no one with bad ideas will ever be able to win political power, or that the will of the people will never be violated; that’s one thing democracy has in common with autocracy and every other form of government. But what democracy has that sets it apart from more autocratic systems is a clear-cut, well-defined process for ensuring that if the will of the people ever is violated, the offending parties can actually be removed from office and swiftly replaced with others who can implement better policies. In other words, democracy offers a way for populations to correct mistakes of governance, via immediate feedback, rather than just having to live with them indefinitely until the autocrat responsible for them dies of old age. And while defenders of autocracy may argue that dissatisfied citizens under an autocratic system can easily do the same thing if their own all-powerful leader turns out to be incompetent or malevolent, by simply overthrowing the bad leader and installing a good one in their place, this argument relies on a pretty loose interpretation of words like “easily” and “simply.” Judging from the way things have actually worked in autocracies throughout history, this kind of method for replacing flawed leaders is, to say the least, far more unstable – and usually far bloodier – than the democratic alternative. And not only that, it’s not even designed to actually select for good leaders whose policies are aligned with the population’s preferences; all it’s really designed to select for are those who are most powerful and best at overthrowing governments. Maybe the whole process could theoretically go more smoothly if it could somehow be streamlined into a kind of universally-recognized formal mechanism through which the population could simply inform bad rulers that they were no longer wanted and then those leaders could peacefully step aside to be replaced – but as we discussed earlier, that’s exactly what elections are; whenever an elected leader is failing to adequately represent their constituents’ interests, those constituents can decree in a systematic, coordinated manner that they no longer wish to have that leader representing them, and can oust them in favor of someone who better reflects their preferences. Elections are essentially just a way of overthrowing political leaders in a way that’s orderly, efficient, and nonviolent.

Earlier I quoted the first half of a comment from fox-mcleod talking about how autocrats’ incentives are structured in such a way that they’re almost always forced to favor the interests of their powerful cronies over those of the broader populace. Here’s the second half of that comment:

If you’re going to be a “benevolent dictator” […] any programs that benefit the common person above the socially powerful will always come last in your priorities or your powerful supporters will overthrow you and replace you with someone who puts them first. So it turns out as dictator, you don’t have much choice.

But what if we expect our rulers to get overthrown and instead write it into the rules of the government that every 4-8 years it happens automatically and the everyday people are the ones who peacefully overthrow the rulers?

Well, that’s called democracy. It’s totally unnecessary for the people to make the best choice. What’s necessary is that in general, the power to decide who stays in power be diffused over a large number of people. Why? Because it totally rewrites the order of priorities.

Now you have a ruler who prioritizes education, building roads that everyday people use, keeping people productive and happy.

Furthermore, nations who prioritize those things tend to be richer and stronger in the long term. Why? Because it turns out education is good and science is important and culture is powerful. It turns out what’s good for the population is better for the country as a whole even though it’s bad for a dictator.

We can demonstrate through studies just how clearly democracies retard corruption.

For more on the basic principles behind why democracies are so much more successful than other forms of governance, see CGP Grey’s rules for rulers.

Again, democracy doesn’t always guarantee perfect outcomes; no system of government can. What it does provide, though, is a systematic mechanism through which we can try out different solutions for our various societal problems, explore different approaches, see which ones work, and then select for those while rejecting the ones that don’t. In this sense, as Michael Shermer explains, democracy is essentially government-by-scientific-method – and that’s its greatest strength:

Democratic elections […] are scientific experiments: every couple of years you carefully alter the variables with an election and observe the results. Many of the founding fathers of the United States, in fact, were scientists who deliberately adapted the method of data gathering, hypothesis testing, and theory formation to their nation building. Their understanding of the provisional nature of findings led them to form a social system wherein empiricism was the centerpiece of a functional polity. The new government was like a scientific laboratory conducting a series of experiments year by year, state by state. The point was not to promote this or that political system, but to set up a system whereby people could experiment to see what works. That is the principle of empiricism applied to the social world.

As Thomas Jefferson wrote in 1804: “No experiment can be more interesting than that we are now trying, and which we trust will end in establishing the fact, that man may be governed by reason and truth.”

Continued on next page →