In the wake of the Brexit vote, Peter, Nick and Fraser discuss what referendums can tell us, and what technology means for democracy.

For more information on Aleph Insights visit our website https://alephinsights.com or to get in touch about our podcast email [email protected]

Hello and welcome to the Cognitive Engineering podcast produced by Tell Me Studios for Aleph Insights. In this series of podcasts we take a look at interesting topics and discuss what we think they tell us about analysis and decision making. I'm Fraser McGruer and I'm here with Nick Hare and Peter Coghill of Aleph Insights and this week we're discussing referendums. So the background to this is the recent referendum we had here in the UK on the country's membership of the European Union. The result was 52% leave versus 48% remain in the UK. So Nick let's start off with this question, do referendums work? I think the answer here would

depend on what you're trying to achieve of course. I mean the referendum we've just had worked in the sense that we had a very high turnout, historically you know higher than any election for about the last 30 years and so in that sense you know it was a very good way of polling what people actually wanted to do. But there's things that obviously referendums can't do and one of them particularly the way that it's set up is to get people's degree of attachment to the outcome. So for example what we can't, what we're not going to get out of a referendum which is simply yes or no, is a potential situation where perhaps you know 52% of people are kind of slightly attached to leaving the EU whereas the other 48% are really attached to remaining.

Hold on, sorry can I stop you there for a second? Just let's just go slightly broader for a moment because just a question, I'm not a political scientist, I don't think many of us are, but something that intrigues me is why would you have a referendum and not just go through your existing legislature? Why wouldn't you do that? So why not just go through in our case parliamentary democracy? What do people say about that? What's the sort of, what do the

political theorists say about that? Well I don't think there's a, there isn't, it's not as though there's a really clear set of guidelines for what things you should have a referendum about. I mean in this case I think a fairly objective summary of the situation would be that the Conservative Party in the United Kingdom has been racked by division over the the issue of whether the UK should remain in the in the EU for quite a long time and I think this was a bit of a gamble really on David Cameron's part to try and settle that issue once and for all, in that a referendum gives you, if you like, a sort of final view of what the population as a last course of appeal isn't it? To say well let's ask everyone in the country. At the moment the system we have where you delegate an MP who makes decisions on your behalf, you know essentially for around about five years, that system means that you're always voting on a quite a large basket of issues and of course you're also not in a position really to change your viewpoint halfway through the term. You're often electing them on the basis of a set of principles that they have and not on their attitude to specific bits of legislation which might come up during that term. So it's very imperfect, the parliamentary system, in terms of actually finding out what people want and acting on it. But of course it does have its uses, it's quite efficient. Unlike in Switzerland where they have referendums about almost all major bits of legislation, here the parliamentary system doesn't require that much engagement by people. You're getting a benefit in terms of reduced admin but the cost is that you get slightly less resolution about what people actually want.

Okay so before I interrupt you, I'm going to come on to you in a moment Peter, but you were just talking about degrees of attachment and you were talking about there's a potential downside to referendums. So I interrupted you, so can you delve into that a bit more, continue with what you were

saying there? Well we know that there's quite a few potential voting systems and every country has slightly different systems by and large. So in the UK we have a first-past-the-post system which means that effectively all the people who voted for anyone who didn't win in a particular constituency, their vote completely disappears. It has no effect. Contrast that with a system like AV where your second preferences might make a difference if your first preference guy doesn't get through or proportional representation where you effectively add up all the votes in the entire country. They're all doing different things and they all have different real-world advantages and disadvantages. A referendum is no different, it's not perfect. As I said, the thing it won't give you is a sense of how much people care about the thing they're voting for and it might be possible to design a kind of referendum where people say, well I'm a yes but I'm only a one out of ten yes or I'm a no but I'm a nine out of ten no. The problem being, and it's a general problem with these kinds of issues, is how do you encourage people to tell the truth? So you know someone who is only a one out of ten but would still rather see that thing happen than not happen would have

an incentive to say they're a ten out of ten. Okay I'm just thinking you could build in some or maybe there's some natural obstacles in the way to people voting that you can judge them to be a one to a ten and maybe we should build assault courses or obstacle courses around polling stations. Although I guess that starts to favour the more physically fit. Mind you in this referendum that would mean that less old people have been voting. I think there are these you're

actually not, it sounds ludicrous but what you're suggesting is that we force people to indicate their strength of commitment to something by effectively undertaking a kind of cost. And actually that's something we might talk about in a minute but there are systems, there are there are you know these revealed preference systems which effectively do do that. Okay great

so I knew I'd hit on something genius there. Peter anything I've got some specific questions I want to ask but anything you want to say at this point? Well I think the referendum is you

can you could argue it's quite a blunt instrument for political decisions. The way we do them in the UK they give you a binary decision either yes or no on a particular topic and they take a long time to organise, take a lot of money to organise. I think the estimates for this referendum suggest 142 million pounds were spent just on organisation and running the poll stations and the counts and that's not including donations to the to the leave campaign and the 50 the 10 million or so pounds spent by the government of taxpayers money for the for the remain campaign.

So there's a lot of money spent. There's probably some opportunity costs in there as well when politicians could have been doing other stuff other than campaigning. Indeed yes yeah yeah

art to to have the same effect. Okay so but that makes me think though it's easy to hijack but it's easy part of the reason it's high easy to hijack is in is in the recent case it's because you're dealing with big issues and people the politicians say this is so important we must go to the public with this and so immediately by its very nature by definition becomes an emotive subject and so therefore vulnerable to being hijacked as you say. So what about you know I don't know how it works in Switzerland but as you're saying they're famous for you know doing a referendum everything I believe there's a famous case of a half-built bridge which halfway through they decided they had another referendum following the previous one saying actually we're going to stop building this bridge I don't know if that's apocryphal or not but what about if we I mean the thing about having your parliamentary system or your federal system is that it's a nice efficient there's efficiency gains right but what about if you could harness technology now and pretty much most houses are hooked up to the internet and there are other things that you could do to support those who are not what about if you had a referendum all the time every week on issues sort of local issues that people care about whether it's schools traffic calming measures etc etc and just make that in a simple sort of binary way in which referendums already do is that really stupid or is that possible Nick tell me well first of all we need to

answer a slightly more fundamental question which what is the point of public policy why do we have governments making laws and stuff and I know it sort of seems it's so obviously a part of life that we rarely stop to ask that question and the I think generally accepted answer is that you need these things where there are where there are what are referred to as externalities right where your behavior affects other people in a way that there's no market for you know where you have a where you have a market we you know we've got very little legislation about you know how many bananas should be in Tesco's because because there's a market for bananas you know people who want more bananas pay more you know if there aren't enough bananas then then Tesco will put the will put the price up and supply more bananas when it when it comes to issues like education policy because of the way that we have public education or things like policing and national security there isn't a market it's hard to imagine how there could be a market there so we need to set a policy for everyone at once we've all got to decide how we're going to allocate resources for those things and that is the general problem that policy tries to solve so now that your question of how do we could we could we come up with an efficient way of doing these things without having these these delegated approaches like having a having a parliamentary system there are actually some interesting fundamental limits on what you can achieve through getting people to reveal their preference so there's a famous theorem called arrows theorem which which is sometimes referred to as a paradox because it's you know when you when you have a large number of issues when you have a bunch of voters who are voting on those issues there are certain things you you sort of think you'd like to achieve so you'd like to you'd like it to be the case for example that if everyone prefers one option to another that that option would be adopted and the group's preferences are going to be expressed in the same way regardless of whether you add something irrelevant to the mix of potential options and all these fairly straightforward things that you'd like it to do and arrows theorem demonstrates you can't have a system that will guarantee doing all of those things so so there's there's limits on what you can get people to reveal through through these kinds of systems there are other systems where you you are effectively trying to get trying to get people to decide how much they would pay for something to happen so just to bring this back to the referendum what if we effectively all bid on whether we stay in the EU and you know an optimal sort of outcome might be well let's let's say as actually appears to be the case that wealthier people would prefer to be in the EU and less wealthy people would prefer to be out of the EU in theory if you could get it right you you could imagine a system where the wealthy people effectively compensate poor people for staying in the EU like a sort of subscription charge now that that might be a more efficient approach but again we come up against another fundamental limit here which is that there is no method of eliciting how much people are willing to pay that will guarantee to be balanced budget so so the only and so this is fairly fairly convoluted but the idea being that let's say I want to say how much I care about being in the EU and I say well I'm willing to pay you know 200 pounds a year to be to be in the EU how do we make sure that I'm telling the truth well there are systems which make sure that you're telling the truth but it might be that when you do that you discover that the whole that the all the preferences as a whole add up to an enormous amount of money and so you can't adequately compensate everyone so that so there are these theoretical limits on what you can achieve with with even with very sophisticated systems okay so so you know in that sense we everything will always be a fudge of some

sort okay that's a good well we need to wrap up everything will be a fudge of some sort I like that okay before we wrap up I don't have we have you said everything you want to say on this Peter

anything final thing to say I agree with you that maybe there is an opportunity with using technology to devolve more decision making responsibility to to people but I think the the the the way that you would have it have it balanced by some degree of rationality and how rational are people able to be about these decisions because we pay we pay we pay politicians and they're supporting civil servants to make decisions on our behalf and we try to empower them with information so they can make rational sound decisions that are in our best interest now I was asked to vote last week about the membership of the EU but I'm not I don't have nearly as much access to information that would inform my decision as other people so I don't believe that I was in a sound position to make a rational decision about whether or not it was it was best for best for the country but I think yeah the point I just wanted to raise about the leaves campaign was that they they they argued that to leave the EU would be to regain sovereignty to regain power and arguing with the assumption that sovereignty is a good thing in the first place yeah and so that's that's a big assumption now I think there are other models of government government where more power is devolved and I think there are there are there are concepts like this subscription-based governance where as Nick describes you can vote for your preference in various things and your your preferences revealed either through just how you behave or how you actually what you specify so there's no the idea of the sovereignty being geographically tied is less relevant in the modern day so we're

moving into different issues they're moving in question of sub sovereignty I mean one thing I don't think we can cover it here but I'm interested if these modes of decision making that you're talking about a subscription for example if they're not being used in government anywhere if they are being used in other forms I don't know but we can come on to that another time but the other thing is the thing is do we want to be asked these questions all the time I mean that's one of the reasons why we have politicians is we can't be bothered we want to get on with our lives that's how I see it we don't want to be asked questions maybe the Swiss love their referenda I don't know Nick just one last thing yeah just I think this is perhaps an alternative

vision for how technology can can do these things and it's looking perhaps a bit further out instead of using technology to for essentially voting on things which is kind of replicating a it's effectively just a digital replication of the analog system we have already is to solve the fundamental problem about about externalities and there's a very famous theorem called the coast theorem in economics which says that if you can allocate ownership of everything then you no longer effectively need government policy and so the equivalent would be close to what Peter was sketching out there you know you you effectively sign up to policies on an individual basis and you pay what they cost so you know in future why couldn't we have a situation where I was in the EU and you weren't you know we happen to live in the same country but I'm in the single market and I can move about but you can't and you know at the moment it's you know it's really only the physical practicalities of doing that that stop us but in future maybe that won't be the case for