Democracy

A number of things have converged at once where I want to push back against how most people understand democracy. I’m going to try and first outline how democracy and governance works in general, how a parliamentary government works as a clear example, and then dive down into the UK tory, labour, and lib dem parties. First post democracy:

TL;DR: Government can’t know enough to reliably predict good policy in advance because Hayek. Hence the most important feature of democracy is the people being able to reverse bad policy. Doing this policy by policy is hard. It’s much easier to give politicians discretion to make policy and replace them with people who disagree if they do it wrong. Hence public discussion of policy: campaigns, debates, promises, facts, an educated electorate etc, are not vital to running a democracy. Political parties: including careerism, cronyism, ideological blinkers, partisan voting etc, are a vital part.

I’ll ignore for this post differences of interest within the country. Let’s think of goals like “not having a recession”, rather than, “giving help to people like me rather than people like you”. I will take it as granted that devising better policies is hard. If you think government is bad because they deliberately choose not to do obvious good things this post won’t interest you. I want to work out how good people who care about their fellow citizens can discover and implement good policies.

Government can’t know enough to reliably predict good policy in advance because Hayek.

Governments do a lot of things. They set tax rates, they set school policy, they start wars, they make trade deals, they set food labeling standards. And for all of these things they are at the very limit of how well we can predict the outcome of their policies.

Consider an industry where people really are motivated to do their absolute best to get a good outcome for people, have the resources needed to do research, and are focused only on a very small problem: fizzy drinks.

Consider coca-cola, a company which knows more about its tiny tiny domain of interest than many government departments know about huge areas of interest. A company most of us directly interact with more often than we do our national government. They knew everything there was to know about soft drinks in 1985.  And they launched New Coke anyway.

It wasn’t a failure because they didn’t know enough. They had done surveys and focus groups and taste tests and every other thing they could have done. They couldn’t reasonably have known more short of forcing everyone to undergo a brain scan. It turns out it was just not possible to know how people would react without doing it.

Capitalism doesn’t depend on companies getting it right first time, it would totally fall apart if it did. What it depends on is trying things, failing most of the time, getting rid of the failures, and keeping the good. This is essentially just Hayek on why planned economies fail.

Hence the most important feature of democracy is the people being able to reverse bad policy.

Governments are messing in things that are *much* harder to understand than fizzy drink. The science of taste is well understood and can be easily and cheaply improved. Macroeconomics is murky at best, and experiments are generally impossible. It would be a miracle if a government came up with the correct economic policy. Let alone the correct family policy. We must simply try things, fail, and keep the good.

Economist Robin Hanson has a wonderful idea called Futarchy. He thinks the main reason why policy is hard is that governments don’t honestly collect accurate data about what policies are going to work.

We take betting markets (the best tool to collect everyone’s knowledge on a subject), and ask them what will happen if we introduce various policies. If the markets say things will get better we do it.

I agree with Hanson that Futarchy is the best way to devise good policies. But I think that on its own such a system is not enough. It’s not enough simply to come up with the best policy we have enough information to devise. You need to change policy in light of data you only get after you try it. (Of course, you can do this inside Futarchy, have some automatic rule that we ask about repeal after implementing anything, but see below in indivisibility of policy).

Doing this policy by policy is hard. It’s much easier to give politicians discretion to make policy and replace them with people who disagree if they do it wrong.

Policies are interconnected. A trivial example is to observe everyone would favour “lower taxes”, “increase public spending”, “reduce the deficit” if presented individually, but together they are contradictory.

A more realistic example is to consider policies in many areas meshing together in a way you only really get out of something like an ideology. Consider employment: an economic system that promotes high-tech industry is going to fail unless there’s enough education in place to produce the required technicians. It’s also going to fail unless labour markets are flexible enough and people are secure enough that they can re-train. It’ll fail if your welfare system is designed around getting people a job next month rather than getting them a new career. It’ll fail if fifteen other things aren’t all set up together all running in harmony.

A vital case of this is the executive. Many a good policy in the US has been driven into the ground because the president didn’t want to implement it properly. Or even just because he wasn’t employing officers with the right frame of mind.

It’s also not always clear which policy did the thing we care about. If schools get worse is it because of not teaching phonics, a lack of sports facilities, or because some change in public transport means kids don’t have time to get breakfast? Rather than trying to guess which did it, it might be best to say that whatever did it, if you put the same group of people in charge you’ll probably get the same results. And if you fire them you’ll likely get different results.

Hence public discussion of policy: campaigns, debates, promises, facts, an educated electorate etc, are not vital to running a democracy.

There is lots of room for policy discussion. Thank god, it’s one of my favorite hobbies. But democracy works fine without it. Consider two systems for running a country:

First, the commonly accepted way: people are good at predicting in advance what policies will work. They find people who promise to implement policies that work well and vote for them. These people implement the policies, good outcomes happen.

Second, the evolutionary strategy I’ve suggested before. Various groups of people propose clusters of policies. We pick one of these groups, they implement the policy. If things turn out well, we vote for them again, and they implement the same kind of policies again. If things turn out badly we fire them and vote for other people who enact other kinds of policies.

Note that this first one depends on a very poor form of feedback. The great body of the people have to learn about policies, they have to understand economics and criminal justice and education. And then they have to elect politicians who will keep their promises.

The second one depends on a rather more direct form of feedback. The great body of the people have to identify good and bad policy after it has happened. They have to elect politicians who will act the same way as other politicians. And they have to trust that a good policy from the past will produce good outcomes a second time.

Note what I’m not saying: I’m not saying that policy discussion is impossible. We are far from totally ignorant. We know about things like free trade, infrastructure, victimless crime, public health. We know a lot of things about what good policy looks like. And it’s important that as many people possible debate and discuss policy. I will do so as part of my political party. I encourage you to do so in your own ways.

I’m also not saying that political lies, broken manifesto promises etc etc are good things. We’d prefer to do without them. My suggestion is only that they make things worse, rather than fundamentally broken. If we are, as many (perspectiveless) thinkpieces have suggested, entering a “post-fact” political environment that will not prevent democracy from functioning correctly.

Political parties: including careerism, cronyism, ideological blinkers, partisan voting etc, are a vital part.

The important thing then is giving people a choice between the current set of policies and another set. It doesn’t matter what the two sets are, or even if people know what they are. What matters is that they can chose between the two and reliably get it.

Political parties are important for giving an actual incentive in politics.

Consider a legislator. How do we actually give them a good reason to perform well? We can’t judge if they followed their manifesto commitments the same way we’d judge fulfilling a contract or even an advertising pitch. If we did we’d force people to enact some damn silly policies when events changed the facts on the ground. Consider a member of the executive. We can’t judge if they administered well the same way we can with an employee because we don’t know nearly enough about the info they have, their task is much more poorly defined etc. And in both cases they might be out of their job in the next election. So long term benefits are not relevant.

What does give people a reason to vote for good things? In general the best incentives for long term action come from bodies that will exist long term.

Companies that expect to be around for decades plan for events decades in the future. Not because anyone in the company expects they’ll be around then. But they expect that later on people will judge them based on how the company is going to do in future, because those people will be judged in turn on how well the company is doing when that future arrives.

Churches, companies, dynasties, and political parties all work this way. They give members now reasons to work hard for members in future. Where in the case of parties this means passing good laws that will remain good long into the future.

Lots and lots of politicians want to be Ted Cruz or Donald Trump. They want to shut down the government or blow up the world to get a bit of short term popularity. The reason we don’t have these people all the time is that normally political parties are powerful enough to shut them down. If the Republican Party had any carrots to offer people for good behaviour then Cruz would never have shut down the government, and Trump wouldn’t be making the world tremble.

On that subject, a final word in favour of honours. Honours are probably the best possible way to reward people for political service. They’re much better than cash bribes, “pork barrel” government programs, government offices for incompetent people and all the other ways people have found over time to reward loyal politicians. A nice medal, some meaningless letters after your name, a nice smug feeling, and that’s it. Doesn’t cost the rest of us a thing.

Advertisements

2 thoughts on “Democracy

  1. Unfortunately, human nature makes most government policies effectively irreversible. Psychological experiments found that people place high premium on the things they already possess. For example, in one experiment people in one group were offered to buy a mug while people in another were given it for free and later offered to sell it. The researchers then found that people would only agree to sell the mugs for several time the price that they would be willing to pay in order to buy them. In politics, this means that policies started by one party are usually continued by the opposition even when the latter was strongly against them initially.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s