Ignorant but Rational


Don Brash accused the public of being ignorant, and then said that this was caused by “the failure of teachers to teach and politicians to explain some of the basic facts of life.” The claim of ignorance on issues of public policy is certainly correct, but the diagnosis for why this is the case is wrong.

When are people more likely to be ignorant? When the costs of being informed are high, and the benefits of being informed are low. So, being ignorant about how much you earn could end up being very costly – you may accumulate debt, you will poorly assess your preferred consumption bundle, etc. That’s why people tend to know how much they earn.

But what are the costs of ignorance in public policy? Well, the cost of voting for a nonsensical idea is pretty low, because the chance of your vote determining the outcome of the election is tiny. If there’s a one in a million chance of your vote affecting the election, and the cost of tariffs for you personally is $5, then the expected cost to you is one ten thousandth of 5 cents if you vote for a party that supports tariffs. Equally, voting for policies that are illogical may have psychic benefits – you might decide to do something because of personal bias.

But if ignorance and irrationality is driven by our current institutions, then institutional change could alter those incentives. As opposed to each citizen getting to vote, we could have a higher level of aggregation achieved by cross-sectional juries of citizens. We could massively decentralise state functions and then allow greater personal choice in jurisidiction, including the possibility of virtual jurisdictions. An easy first step would be to end get out the vote campaigns.


12 Responses to “Ignorant but Rational”

  1. Jonathan Says:

    A related point is that the distribution of costs and benefits can also explain why bad policies can prevail. Free trade is a good example. The benefits of free trade are distributed widely across the economy to all consumers and producers who use imported goods as inputs. They all benefit a small amount from the lower prices for imported goods. The costs, in the form of lost subsidies, are borne by a much smaller group (exporters). The costs of being uninformed for exporters are high but the costs for consumers are quite low.

  2. Jesse Says:

    The concept of rational ignorance also applies in other voting contexts. A small shareholder, for example, may be rationally apathetic about certain corporate decisions which are put up for a shareholder vote. It is costly to become informed about certain proposals and the cost may exceed the value implication of the decision for his or her parcel of shares.

  3. Colin Hunter Says:

    Juries of citizens has some serious issues, even putting aside equality issues, it is logically flawed. Its most important tenant that is incorrect is that it assumes that a random sample will give and equal distribution of the electorate. This is incorrect on two levels. Firstly at any one time virtually any combination of people can come up, it could be an electorate of union officials, it could be an electorate of political parter staffers who knows and while over an infinite amount of time this will even out, the smaller the sample the more likely any one sample will be biased. As anyone who understands chance will know, it isn’t how you get lucky it is when. Holding 4 aces doesn’t mater if your opponent has a royal flush right? So it is quite possible with random sampling you will get year after year where say trade unionist dominate the sample, with a simple majority (of say one vote) and then one year you get a sample of completely act party affiliates the next year making up all (all the sample the next year). What will matter more? All you are doing, by taking a random sample is determining your election by chance. In fact it is quite possible that for years the same unlikely event will happen year in, year out. Sure over the entirety of time it will even out, but within say a century or 33 elections, who knows what would happen. Hell people win the lotter twice, the odds on that are so astronomically low as to be virtually inconceiveable, but it happens.

  4. Colin Hunter Says:

    I might add the smaller the sample and the less often it is taken the more impact this will have. So making your sample smaller will inherently be rather counter productive, is that one of the those unforseen consequence I keep hearing about? :p

  5. stephenwhittington Says:

    Having a royal flush when someone else has four aces is impossible.

    Regardless, your case seems to take public opinion as a given – hence your fear of unionists or ACT Party affiliates dominating my group. The whole point is that you need to worry less about that, because you have in-built incentives for truth seeking.

    Jesse, that is correct, but the rights are tradeable. I may not know what Telecom should do because I am rationally ignorant, but prices will see me move the property to the highest value user – and that may be a private equity firm, a larger stakeholder, etc.

  6. Colin Hunter Says:

    You are correct Stephen, how embarassing, a straight flush then, it still beats 4 aces. 🙂

    The thing is that, a random selection has fraught, if you understand statistics at all or randomization you will see why that it is. The less people the deciding the more standard deviation will matter, the more likely you will achieve . Even if you have built in incentives for truth seeking you won’t sway act party members or communits, you are dealing with people who have ideological differences. Sure you could get middle of the road voters who could be swayed, but the problem is you have no way of achieving that. In the article you site for example the author acknowledges that a broad cross section of the community is important (equal number of men and woment etc…), but the problem is a random sample, that is small (12 people) won’t achieve that, sample have to be relatively large in order to work and be frequent.

  7. Colin Hunter Says:

    Sorry that last post was rather incomprehensible…. oops, I didn’t mean to hit submit, I hope I still make some sense. Also while things may even out, when variation happens is still very important. Statistical swings of the random samples will likely prove decisive in such elections.

  8. Colin Hunter Says:

    One last thing, is that in Texas Holdem, which I now realize I was thinking of, it is indeed possible to have 4 Aces and a straight flush. Well, at least I’m saying in retrospect I was thinking of that. 8)

  9. stephenwhittington Says:

    Colin, the scheme could be easily modified:

    For example, we could elect next year in groups of, say, 1000, who would vote on our behalf three elections hence.

  10. Colin Hunter Says:

    Sure, but it is a scale, anytime you have a randomly chosen voting group, you will add luck into the mix, which will in turn make the election more arbitrary. What you need to know is that if the incentives to make smart policy, will out weight the negatives of arbitrary elections. That would require significant case studies and evidence, which I’d be keen to look at, but I suspect no political party would be silly enough to implement a policy (for various reasons) and as a result I’d be surprised if there was much scientific evidence either way. Also even increasing the numbers to 1000 say, would still leave possibilities of odd results. Think about opinion polls and the deviation in them, you would probably have similar deviation in a 1000 person election (which is about the same size as many polls), this is of course more stable, but it can easily see spikes. Also the more you increase the number of people, the more you decrease the benefit of small number of people. At a 1000 people, you still have a significant collective action problem. Sure not as bad 1 million, but enough for most people a 1 in 1000 chance at making a decisions probably pretty close to the apathy of a normal election.

    Luck isn’t always a bad thing, but it certainly makes decision making a lot tougher, for example political parties and campaigns would still be quite important, but much of this would have to do with luck mitigation and various trade offs around that, for example, if you are desperate and a minor party, you may be better off preaching to your base in order to try and get a lucky draw, rather than try and contest on a broad front. Major parties with more to loose would probably try to be more conservative, this may also effect policy. There is already a great deal of uncertainty, but increased uncertainty would, I imagine, alter the political game.

    One of the issues I think you would run into a bit, is that while perhaps a more motivated jury might make better policy, at what point are these people not just randomly chosen politicians? Aren’t professionals the best at such things anyway? (This may be contentious here, because I understand a lot of you think professionals are not as smart as market forces). Wouldn’t an economic incentive, allowing the people to quit their jobs and focus on policy decisions be even better? What about people with relevant skills, would they be better at choosing policy? Would the arbitrary nature of the electorate actually turn people off the democratic process?

    I think though Stephen, you have hit on one important point, how do you involve people in a meaningful way into democracy in a modern society? I’m not sure I know the answer, but it is interesting to think about.

  11. Jesse Says:

    I was reminded of this post when I read the coverage of Iceland’s referendum on whether to repay money owed to Britain and the Netherlands in relation to the bailout of Icesave. The Icelandic parliament approved repayment but was vetoed by the President, whose personal approval ratings soared.

    In the subsequent referendum 93% of Icelandic voters voted against repaying the money. I was struck by the following two passages in the Times’ coverage:

    “Icelanders shot fireworks into the night sky over Reykjavik harbour to celebrate the referendum result.”

    “The credit-ratings agencies have already threatened to downgrade the Icelandic krona to junk status in the event of a no vote. Iceland is dependent on receiving the next tranche of an IMF loan and an Icesave settlement is crucial in establishing Reykjavik’s credibility.”

    This seems to be a case where the costs of irrationality are high indeed.

  12. Colin Hunter Says:

    That is why I think we can all agree Referendums are a load of BS Jesse. Experts have to make the tough decisions and then be politically accountable. Hell look at all the trouble California has.

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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: