RSS feed for entries


Democracy Doesn’t Work

[Fair warning: this is on the longish side. Links below lead to headings]

Free Speech | Informed, independent voters | Voting | Minority protections | Representation | Conclusions

Democracy is comatose when governments can go to war against the will of large majorities. Choosing to die for a cause is the most fundamental decision people can make, and in a democracy that decision is supposed to be in their hands. Yet governments made war in Iraq against every indication of what their citizens wanted, including in the US (where the majority at that point was small). It’s time to admit that our democracies may still be lying there, but they’re not doing much.

There are many other examples of governments ignoring majorities. In the US, poll after poll shows that most people want a medical payment system in line with the rest of the developed world, yet a little bit of advertising at election time seems to be enough to prevent us from getting it. Poll after poll shows that people the world over want sustainable environmental policies, but those policies aren’t implemented even when they cost less. The simple reason is always that some vested interest prefers otherwise. This is understandable, but it is not democratic.

Diffuse sources can be as anti-democratic as the more easily identifiable point sources. A striking recent example was the election in Iraq. There was talk of how the Sunni minority, at about 37% percent of the population according to the CIA, would deal with being outvoted by the more numerous Shia (approx. 40%), and what role the Kurds would play, who are about 20%. The exact figures don’t matter. Give or take ten percent, if you like. Meanwhile, there was almost no talk about the likely effect of–or on–a huge 60% majority. After years of war and dictatorship, women are between 55% and 65% of the adult population. Two thirds of the citizens were written off, not just by the Iraqis but also by the supposedly democratic West, which ignored this problem. When two thirds of the people are invisible, that is not democracy. It is a charade.

Recognizing how far we are from democracy leads to a series of questions. First, is it worth worrying about? Maybe some other form of government would serve us better, and poor old democracy should join all the other failed experiments on the ash heap of history. Then, if it is worth saving, how can that be done? What led to the coma to begin with, and what needs to be done differently?

The first question is easy because a few thousand years of recorded history give a clear answer. As Churchill famously said, democracy is the worst system of government, except for all the other ones that have been tried. It’d be nice, however, if we didn’t simply have to make the best of a bad job, i. e. government. There has to be a better way to find the right direction than to stumble through every wrong one first.

Science, which yanked us out of the center of the universe and into the company of apes, is helpful here as well. There is some actual evidence that many heads are better than one. In a simple experiment, people guessed the number of beans in a jar. Very rarely did any single person get it right, but the average of all the guesses was surprisingly close to the truth. “Market” experiments, where people invest (bet, really) on the outcome of anything from wars to elections to the price of turnips also tend, in the aggregate, to predict what actually happens. (James Surowiecki wrote a book, The Wisdom of Crowds (2004), on the evidence for the idea and its implications.)

That outcome feels counterintuitive, because the behavior of mobs has given the intelligence of the many-headed a bad name. The crucial difference is that certain conditions must be met for the wisdom of crowds to manifest itself. The people participating in the experiment must act independently of each other, and they must all have access to the same, reasonably complete, information. These two simple conditions are enough to explain why current “democracies” really aren’t.

Independence is the first problem. Every person on Earth will say they ignore ads, that they can make up their own minds about the news, and that it’s absurdly easy to distinguish between reality and TV. And yet, when all the shouting has died down, the politicians with the biggest advertising budgets are almost always the ones elected. Whatever we may be capable of individually, when we’re all together, you can tell us anything.

So the first requirement of a democracy is that you can’t tell the people anything. You must tell them the truth.

(I know, I know. Wait a minute. Hear me out.)

Free Speech

Requiring truth raises a whole rat’s nest of issues, starting with the deeply philosophical one of what is truth. However, absolute truth is not essential. What I’m talking about is facts, and so, within reasonably broad limits, the situation is not hopeless. It is feasible to tabulate how closely statements match actual facts. It’s also possible to find out how well-informed the consumers of given news outlets are. It’s been done. The 2003 PIPA/Knowledge Networks Poll(pdf file) showed, for instance, that 67% of the people who watched Fox News believed that Saddam Hussein and Al-Qaeda were in cahoots. This would be like a cabal between the Mafia and the people who blow up abortion clinics. It’s not impossible. Few things are impossible. But the standard of proof for the highly unlikely is higher, not lower. There needs to be some hard evidence before wild ideas are broadcast. If a news program broadcasts wild ideas based on entertainment value rather than evidence, that should be enough to lose one’s license to use the public airwaves (or internet backbone or space program).

News organizations aren’t the only culprits. Politicians have made names for themselves as liars, and what the appropriate penalty is for them, I’m not sure. Making ten retractions to the same audiences for every lie? Duct-taping their mouths shut for a while? Maybe they need a three strikes policy: three definite falsehoods, and they’re barred from holding office. Ignorance is no excuse. After all, if they, of all people, don’t know the facts, how are they supposed to run the country?

Talk of lost licenses and offices leads directly to free speech issues. There’s irony in the way unregulated free speech ends in gluts of BS from the highest bidder that drowns out free speech. The same is true of unregulated anything. Stock markets may be the core of capitalism, but they are among the most tightly regulated activities in the world in order to prevent the demise of free markets. Free speech won’t stay free for long without some regulation, which is recognized in the obvious exception that it does not cover yelling false alarms of “Fire!” in a crowded theater. However, technology has swamped the puny regulations that worked in the past. A tidal wave of political advertising drowns out any discussion. Gluts of spam have made mail, phone, and the web barely usable. We’re told we have to put up with this for the sake of free speech.

What nonsense.

Free speech itself is worth dying for, and it’s even worth some inconvenience. Agreed. But this has nothing to do with free speech. The purpose of free speech is to protect freedom of thought, not the freedom to make a buck by any means possible. As a matter of fact, the latter comes awfully close to the fire-in-a-crowded-theater misuse, which is why there are such things as truth in advertising laws and punishments for fraudulent claims. There is no such thing as commercial free speech, nor should there be. Speech whose purpose is to get something out of you, be it money or a vote, needs to be subject to consumer protections. The restrictions on speech designed to get your money could be looser than those on talk designed to get your vote–what’s money, after all, compared to the life of the republic?–but both of them have to avoid obvious lies. We need to catch up to the technology here.

Of course, nobody whose profit or job depends on pretending to the truth willingly admits to selling lies. However, when their noses are pushed into their messes, the usual plea is that they’re only doing what people want. This may or may not be true, but it is still no excuse. There are always people to buy all the heroin you can sell, but that doesn’t make it a good thing to push. Pedophiles will pay for kids to abuse, but that doesn’t mean they should be provided. If something does no harm, by all means, sell it. But lies in the body politic are killing democracy, so it doesn’t matter whether people like lies or not. Lies cause at least as much harm as yelling fire in a crowded theater. They simply cannot be allowed.

That said, let me rush to stress that I am not, NOT, saying that the interpretation of facts should be controlled in any way. It’s up to Fox News if they want to say, “There is no evidence of a connection, and we think that means we have double the worry, since there are two independent groups instead of one.” Anyone can have any opinions they want. They just can’t make up any facts they want.

One advantage of regulations is that everyone has to abide by them. Fox couldn’t provide rollicking good yarns that pull profit away from, say, CNN which then has to provide something even more rollicking. The same would go for website A and website B. Anyone who writes or talks about politics should be constrained to make sure of their facts before disseminating them, just as the same is true of commercial products or stock prices.

There is legitimate fear of free speech regulations because they could be abused to limit speech we care about. Since regulations are a can of worms, it is customary to say at this point that it is up to the voter to take responsibility, to be analytical, and to become informed. That would be great. It would also be nice if charitable contributions could stretch far enough to prevent poverty, provide good educations, and fund medical care. It would be good if corporations took social responsibility and didn’t pollute. Many people’s lives would be simpler if married people were faithful. Lots of things would be good. Basing social policy on hopeful noises about human nature, however, is a recipe for disaster. Communism tried to rely on altruism, and look what happened.

It is better to recognize the danger of abuse and to build prevention of abuse into the regulations, than it is to lose the very free speech we care about in the din of greed.

I may, of course, be laying myself open to the objection that I’m making the same mistake. I’ve taken account of the human tendency to collar the conversation, regardless of having anything to say, but what about boredom? People are supposed to be bored silly by facts. Actually, I doubt that. We all care deeply about things that affect our lives. People are bored by misdirection, obfuscation, manipulation, and everything else that requires excess energy to cut through crap. But all the important things in politics are easy to understand–or can be. It is easy to present the crucial facts about health insurance, war, pollution, and even computer-based markets versus floor-trading. However, there are very powerful interests who would rather not, so we have idiotic “he said, she said” journalism and head-bobbing pundits who say these things are beyond ordinary folk. I don’t think so.

Informed Voters, Independent Voters

The architects of democracy operated in the 1700s and 1800s, at a time when the threats to the vote were mainly at the back end. The dangers were ballot-stuffing, votes from dead people, and dishonest counts. The architects didn’t worry too much about manipulation of voters rather than of the vote because they didn’t need to. Times change. Propaganda became a high art very recently.

The fact is, if voters are to have reasonably complete information–not too little, and not so much that they’re swamped–some large changes in standard operating procedure are needed. (At least, large changes to how things work in the USA.)

The first point is that political advertising is not a form of information, at least not as currently practiced. Its purpose is to drown independent judgment in repetition. I know that this has no effect on you, or me, or anyone we know. And yet its effectiveness at manipulating the vote is so good that the first priority of politicians is to throw all the money they have at it.

Vote manipulation of any kind, whether front- or back-loaded, is a sure way to prevent the wisdom of crowds from manifesting itself. It has to be reduced to a level that is not incompatible with thought, which is not impossible. Sociologists and psychologists know how to measure comprehension and bias. My guess would bethat political campaigns should run no longer than six weeks, and ads within that time should be limited to some small number of minutes per 24 hours. Enough already, is the phrase that comes to mind.

Reducing advertising might solve an array of problems in addition to voter bamboozlement. Campaign finance should be less of an issue. If fundraising ceases to be the point of politics, corruption of the legislative process is less likely. Political advertising is a keystone in the arch that props up quite a few of the problems in modern democracies.

Politicians might object that such a straitjacket would leave them no way to spread the word about themselves, but I doubt it. People would pay attention to real debates. Debates have everything: tension, possible drama, and information to boot. There’s also face-to-face campaigning. Voter information booklets are highly useful, and could be expanded to include electronic formats, podcasts, and, for all I know, direct transfer to surgically attached phones. It doesn’t matter what form the information comes in, so long as it is information and not manipulation.

An informed electorate depends on one last point, and it is the biggest of all. The voters have to be capable of being informed. Common sense is the really essential ingredient, but it does help the process of getting the word out if voters are literate and numerate. Just imagine the whole world, including the female half of it, able to read and write and knowing where to get real information about politicians. It would be a different world, wouldn’t it? That’s how far away we are from full-fledged democracies.


Once the informed and independent voter reaches the polls, the voting process itself is not as simple as it seems. Counting the votes is the most tediously straightforward part of democracy, and yet it has sprouted enough issues to bring the whole thing down. The issues include: plurality versus majority rule, and how majorities are tallied; direct versus electoral college voting; how to register voters, where polling should take place, and how to guard against voter fraud; how to draw voting districts; and who should vote.

The actual business of filling in secret ballots, secure ballot boxes, and accurate counts are parts of the process that we know how to do at this point, although recently there have been some curious irregularities with electronic voting machines in the US. (The simplest solution to the latter problem is a proper paper trail. The solution is so obvious, forgetting to implement it seems downright peculiar.)

The more difficult problem is tallying the votes. I don’t mean the physical process of counting them, but how they are added together. It seems straightforward that the candidate with the most votes wins, but it’s not. This was brought home to Americans vividly in the 2000 election. If the majority who opposed Bush had been voting in unison, it would have been called a landslide victory. Instead, a minority candidate won because what we really have is plurality rule, not majority rule. (If we had majority rule, there would never have been an iffy situation for the Supreme Court to decide.)

If you do the math, plurality voting is the worst way of reflecting the will of the voters. There is some argument about what’s best, but none that plurality is the worst. It stands to reason, of course. If there are more than two choices–and there always are since None of the Above is implicitly present even with only two candidates–plurality voting facilitates the election of a minority candidate.

Instant runoff and approval voting are the other two main methods of determining a winner. Instant runoff allows voters to rank candidates. Rounds of elimination of candidates with the fewest first place votes finally leave two contenders, and one emerges the winner. The fundamental problem here is related to the one in plurality voting. In the latter, voter preferences for everyone except their first choice are lost. Instant runoff, however, loses sight of preferences for eliminated candidates. This is less of a problem than losing sight of all but one, but it can still be a problem. (See this example for details, or go to Science News, Nov 2, 2002, Are we using the worst voting procedure? by E. Klarreich, for a more in depth discussion.)

Approval voting gets around the instant runoff paradox described in the example, but at the price of favoring “consensus” candidates, which could also be called “least common denominator” candidates. In the approval method, voters rank candidates according to how much they approve of them, so to speak. Each rank carries its own weight, and the “heaviest” candidate wins. The result is that someone who is everyone’s second choice will win if votes for first choices are more or less evenly split.

Having seen approval voting at work in an academic department, and having seen just how mediocre a candidate can be elected that way, I’m not a big fan of the method, even though it sounds, in theory, like a rational approach. Instant runoff voting is less likely to elect a mediocre candidate (since he or she would not get many first choices and would be eliminated early on). It also has the advantage of being easy to understand, and hence less subject to voter error. Possibly what’s needed is a hybrid system in close cases, which are the ones that cause all the problems.

The important point is that there is nothing sacrosanct about any particular method. Nor is there anything sacrosanct about breaking ties, which is the focus of some theories of voting. A democracy is not about breaking ties. It’s about reflecting the will of the voters. The best method to that end is what we should be using.

Direct election versus electoral college-type procedures can lead to very different results based on the same votes, just as the voting methods themselves do. Initially, that was the whole point. The aforementioned architects of democracy were sure that the undiluted voice of the people would be an inarticulate grunt, and that a layer of more intelligent people should buffer the ignorance of the masses. Most democracies have gone to direct voting, and the US is one of the last holdouts of the antique system. The reason for this isn’t really because anyone here believes voters are too stupid to elect a president. It’s the same reason for all the other anti-democratic admixtures in democracies: powerful vested interests prefer it this way. And if they didn’t have it this way, they wouldn’t stay powerful for very long.

How to vote is not a hard question, but it keeps being muddied by people who would prefer to reduce the number of voters. The answer, of course, is voting should be as easy as possible. The current system of marching off to a physical booth and standing in line is based on 18th century technology (and the landed gentry’s ability to take half the day to vote). At this point, assuming there was genuine interest in universal suffrage, voter registration could take place at post offices, driver’s license bureaus, by mail, phone, or web. Voting itself could take place by mail, or, assuming all the security and fraud issues are worked out, electronically. These methods have been tried in various places and work perfectly well, with no more, and possibly with less, error and fraud than the traditional system. Without a secure postal service, there may be a need for a booth-and-box system, but otherwise our inconvenient, time-consuming, and boneheaded process seems designed mainly to exclude voters.

When the politicians select the voters, rather than the other way around, gerrymandering is at work. In that form of vote manipulation, a voting district is carefully carved out to contain a majority of the kind of voters that the carving politician or party needs. This is no less a manipulation of the vote than any other kind of cheating. The original idea was to allow people who form a natural unit to vote conveniently. For instance, residents of town A vote there even if they live in a valley which is closer, as the crow flies, to town B. There’s nothing wrong with the original idea, and with modern statistics, mathematics, and mapping methods, voter convenience could be accommodated in an otherwise geometrically compact area.

There is, finally, argument about who should vote. This is no longer about which groups are qualified. That question was decided decades ago, after it was discovered that citizens included not only blacks, Jews, and poor people–they included women too (!). The big question now is whether the vote should be limited to the willing, or should include everyone. Is voting a civic duty that citizens can be called upon to perform, or is it something best left to personal inclination? I tend toward the civic duty side of the argument, but then I tend to vote at every opportunity, so I’m biased. If the largest number of informed, independent voters gives the best answer, the pool of voters should be as large as possible. But if the pool then includes totally bored slackers, maybe this is not so good. It’s a difficult question.

Minority Protections

The problem of minorities is a structural weakness in democracies. Majority rule is the whole point of democracies, so they’ll necessarily be less than ideal for minorities. The solution isn’t to say, “Get used to it,” because minorities have legitimate needs that cannot be ignored. A dictatorship of the majority is still a dictatorship

Part of the purpose of bicameral legislatures is to protect minority interests. In the US, the way the Senate is elected artificially increases the power of the least populous states. At the turn of the 1800s when the system took shape, the important minority was farmers. Farmers remain a minority and they remain important, but they are far from the only minority group whose rights need protection. Ignored minorities are endangering democracies, or preventing their birth, all over the globe. Sectarian, tribal, ethnic, and racial strife are endemic at this point. By facilitating gradual change, democracies are supposed to prevent the need for violent overthrow, but their inability to accommodate minorities is leading to predictable results. Democracies must get better at respecting minorities or they really will become failed experiments on the ash heap of history.

The first line of protection for less-powerful citizens is judicial. Solid–and enforced–bills of rights and freedoms, and both punishment and prevention of all discrimination go a long way toward avoiding problems to begin with. Autonomous regions might be part of the solution in some cases. However, the point here is what can the democratic process itself do to help the situation, and that is much less obvious.

The most promising idea I have heard is Lani Guinier’s one of “proportionate interest representation.” Hendrik Hertzberg explains it on the site: “It’s really a modified at-large system. In a citywide election for five council seats, say, each voter would have five votes, which she could distribute among the five candidates any way she likes. If a fifth of the voters opted to “cumulate,” or plump, all their votes for one candidate, they would be able to elect one of the five. Blacks could do this if they chose to, but so could any cohesive group of sufficient size. This system is emphatically not racially based: it allows voters to organize themselves on whatever basis they wish. It has actually been tried in a few jurisdictions — including the proverbially American city of Peoria, Illinois — and has had notable success in all of them.”

Cumulative voting differs from approval voting in that voters express how much of a given rank they want to give to a specific choice. They don’t just vote for, say, three out of the five candidates and then have their choices weighted for them by the counting method. They can give A three votes, and B and C one each, or they can give A and B each two votes, and C one.

Cumulative voting together with instant runoff choices would make for an interesting ballot. Too interesting, some might say. However, since the will of the voters is the heart of democracy, this is not the place to cut corners. Voter information and optimal visual design of ballots can solve this problem, which is, really, no harder than tracking sports scores. The process shouldn’t be any harder than it has to be, but neither should it be so simple that democracy itself disappears.


The evidence suggests to me that the whole idea of political representatives is another structural weakness of democracies. I haven’t seen this idea mentioned before, so maybe my thoughts on the subject are loopy (or even loopier than usual), but consider the facts.

Voting, in a democracy, is not supposed to be the only expression of the will of the people. The basic idea is that they elect representatives, who then represent their interests while carrying out the day to day chores of keeping society running smoothly. Politicians do worry about what their constituents think, but mainly with a view to being re-elected, which is not the same thing. That is most clear when re-election and representation conflict, in which case most politicians would rather keep their jobs than do them. In other words, they’re just like you and me, doing what they need to do to get by, and representing themselves.

It’s silly on the face of it to think that most people in government would ignore their own needs, when most people outside government certainly don’t. Instead of bemoaning this fact, or wasting energy talking up better behavior, government needs to be designed on the premise that the people in it are human beings. The solution is to harness self-interest, not to fight it.

However, although it’s clear that calling “representatives” on the carpet every few years is not working, better alternatives are less clear. The opposite extreme to representation is government by direct vote, where everything is decided by plebiscites. That would break down in days. Information is essential for crowds (or individuals) to have any wisdom at all, and there is no way a quorum of citizens could take the time to know enough to make sensible decisions. Government is a full time job.

The core problem is interest. Government is a basically boring, administrative job, more like housework, if it is done right, than it is like ruling the world. It consists of trying to prevent messes and cleaning up other people’s when prevention failed. People who aren’t in government, and many who are, have better things to do than worry about other people’s problems. Once those problems grow bad, they just want them to go away. That is not the same as solving them.

So, two criteria emerge. One is that citizens can’t do the job of government themselves, and that therefore some form of oversight is necessary. Two is that the oversight has to take as little time and energy as possible.

Energy, in the sense of the amount of attention people are willing to give, is a big limiting factor in democracies. Citizen oversight is essential, but it won’t happen unless it’s easy to do. In this connection, it’s important that it is easier to focus on what’s wrong than what’s right. Pain draws attention. People are quicker to vote against things rather than for them, and to listen to negative ads rather than nice ones. So, on the principle of harnessing motivations rather than fighting them, the attention-getting potential of annoyance could be used to stop emerging problems. Instead of elections for candidates or initiatives, we should have votes of no confidence against failures.

In other words, citiziens wouldn’t elect leaders. They would unelect them. I don’t know if a system of unelections has already been tried somewhere, and if there are good ways to implement it. One possibility is to have no-confidence votes as a constant cudgel in the background, the way parliaments do. I could see a rolling system, which would dovetail with a convenient from-home voting process. An accumulation of legitimate complaints against either politicians or laws would lead to the calling of an unelection against them. No doubt, there would be enough of these things to warrant a regularly scheduled ballot, and the current crop of culprits with sufficient complaints against them would appear on the next one.

A recall could get rid of unsatisfactory politicians, but it says nothing about how to find officeholders in the first place. How to find competent people is the big question much wiser heads than mine have been unable to answer. Democracies answer the riddle by saying the question is unnecessary. Common sense is all we need, so the current method of selecting officeholders basically asks for volunteers on the assumption that anyone can run a government. The current state of the world suggests that the assumption is flawed.

The volunteers, once they’ve stepped forward, are then expected to campaign. That is an overlooked problem in itself. Campaigning is all right for Miss or Mr. Congeniality, but the requirements for a good campaigner and the requirements for a good administrator are just about mutually exclusive. What we need in government is the sort of person willing to worry about the effect of accounting rules on inventory. Boring? Yes, until depleted vaccine inventories result in an uncontrollable measles epidemic and brain-damage for your child. What we need are boringly competent administrators. What we want–and what we vote for– are people who feel like friends. Is it any wonder that government isn’t working too well?

So, somehow, the selection process has to be compatible with tedious competence. However, a dignified procedure to appoint the best qualified would result in nothing but the hegemony of the best-connected. Maybe the solution is to continue to rely on volunteers, but to narrow the pool to those who have more than a mosquito’s chance in a hurricane of actually handling the job. The proven way to see what people can handle is their track record. So maybe the volunteers should be limited to people with a couple of decades as excellent administrators(as determined by their subordinates and their clients), whether what they administered was MegaCorp Inc. or the church bazaar.

Having a potential pool of officeholders still doesn’t tell us how to select the one to do the job. There is no easy answer that avoids overburdening voters and cronyism, both at once. I’d like to suggest a wild idea that might be worth trying only because it hasn’t already been tried and failed many times over. How about using a simple lottery to pick one person out of the pool?

Don’t laugh. It might work. There aren’t many systems with randomly chosen leaders, but one example is Tibetan Buddhism. The supreme leader is chosen by a priest based on visions of reincarnation. From the standpoint of a non-believer, they choose someone at random and then educate him to the task from a tender age. Out of fourteen Dalai Lamas, two, the Fifth and the current one, have been outstanding. Neither democracies nor monarchies can boast a 14% success rate, and dictatorships aren’t even on the map. Maybe we should give randomness a try. The officeholders could always be unelected if they didn’t work out. Just a thought.


There are many things working against current democracies. Some of them are corruptions, such as attempts to limit who can vote or the use of political office to protect the officeholder. But some of them are structural problems. They are an intrinsic part of democracy and need to be explicitly compensated or the whole system can succumb. The big structural weaknesses are that voter attention is finite, that minority interests are ignored, and that representatives represent mainly themselves.

Some of the cures are easy. Sensible voting and tallying systems could be implemented tomorrow. They’d throw many current politicians out of office, so they won’t be, but the point is, they could be. Achieving a literate and numerate electorate is also not difficult in any real sense of the word. It’s just expensive. On the other hand, restructuring the selection process for officeholders, and working out the best combination of rewards and restraints, would require much research and experimentation to arrive at an effective solution.

So, are we depressingly far away from democracy? Yes. But, as the saying goes, “scientific studies show” that democracy would work well if we actually implemented it. The closer we’ve been to the ideal in the past, the more satisfied people have been with their governments, so the track record supports the theory.

The good news is that nobody can say democracy has failed. It can’t have done, because we haven’t tried it yet.

legal:  Site covered by a creative commons license with the following limits: content may only be used  without modification, for non-commercial purposes, and with attribution to Mia Molvray and her home page as noted here.