Oral evidence - Trade Union Political Funds and Political Party Funding

Members present

Lord Burns (Chairman)

Lord Callanan

Lord De Mauley

Baroness Dean of Thornton-le-Fylde

Baroness Drake

Earl of Kinnoull

Lord Richard

Lord Robathan

Lord Sherbourne of Didsbury

Lord Tyler

Lord Whitty

Lord Wrigglesworth

________________

Examination of Witnesses

Dr David Halpern, Chief Executive, Behavioural Insights Team, and Lord O’Donnell

Q38 The Chairman: Good morning. Welcome, Lord O’Donnell and Dr Halpern. Earlier I explained to the Committee your backgrounds and your interest in this whole subject. Do you have anything to say by way of opening remarks?

Dr David Halpern: I am in your hands. I am happy to talk about defaults in general, to get you going, if you like.

Lord O’Donnell: We might declare interests, as well. To be absolutely clear, I am the chair of the advisory panel on the Behavioural Insights Team and the chairman of Frontier Economics, which does a lot of applications of behavioural economics and other disciplines to both public and private sector issues.

Dr David Halpern: I am David Halpern. I head the Government’s Behavioural Insights Team, which is a social purpose enterprise co-owned by the Cabinet Office, NESTA and so on. We advise in many policy areas. One area of great interest in recent years has been the use of defaults and how extraordinarily powerful they are in their impacts.

Lord Robathan: I am sorry—the use of what?

Dr David Halpern: Defaults—where you set it one way or the other.

The Chairman: Opt-in or opt-out.

Dr David Halpern: Yes. That would be the simple example. The most famous example in policy in recent years has been changing the arrangements around pensions, which we have seen not only in the UK but in a number of other countries—in the US, with 401(k), and in the UK, from 2012, stemming from Adair Turner’s report. In round numbers—this may give you a clue about what we see elsewhere—less than 10% of people opt out when you change it so that people have to opt out, if they wish, rather than opt into a pensions arrangement. There are almost identical numbers in the US, for 401(k). It is happening on a phased basis, so we get a clue across the piece, and the numbers hold up pretty consistently.

If you want to know why that attracts so much attention in the policy world, you can compare it with the effectiveness of the conventional policy lever that we have used traditionally—a tax subsidy to encourage people to save more, on which we spend, potentially, literally tens of billions. Our best estimate is that 1p in the pound of extra saving comes through, if you really factor it in. A tax subsidy is not very effective in driving extra saving in relation to pensions, as opposed to simply changing the default from an opt-in to an opt-out, which has been a game changer.

From 2012, we saw it for large firms. More than 90% of people are staying in, if they are eligible. Some people are not eligible. For medium-sized firms, we see an even larger effect, because the baseline was even lower. At the moment, it is gradually reaching out to small firms. We have seen an incredibly powerful effect across the world.

The Chairman: How would that compare with the numbers under the previous opt-in system?

Dr David Halpern: For large firms, it was about 60% or less. For medium-sized firms, it was more likely to be around 40%. Now, with the change in the system, it is reaching numbers that are 20 or 30 percentage points raised from that level. As it moves from the large firms, whose arrangements are quite well organised, into medium-sized and smaller firms, it looks as if the effect is even larger.

The key issue with all defaults is what people want to do: what is the underlying preference of the population? That is perhaps one of the great unknowables in the area you are wrestling with. The general belief is that most people feel that it is a good idea to save. If you make it easy for them, they know that they should do it. That drives extra savings disproportionately among low-income groups and younger people, in particular, who were slow to sign up previously. Essentially, on the back of an envelope, if you want someone to do something, you make it easy. Those frictional costs turn out to be extraordinarily important, particularly around something that people kind of want to do anyway. In relation to other kinds of donations, one big question is: what is the underlying preference? You may have a better idea on that than I do.

Lord O’Donnell: I would add only one point. Some people say, “Ah, that is just because people are ignorant. They did not understand it”. As the auto-enrolment programme for pensions moved down to small and medium-sized enterprises, it hit my own, Frontier Economics. Most of our staff have two degrees in economics, so you would think that they knew whether or not to save for pensions. Actually, they are exactly like the rest of the population. That just shows you. If you think that is a question of education, it is not—it is a question of inertia. Inertia is incredibly powerful. That is shown up in most of the behavioural examples, not just on pensions but on things like organ donation. There is a whole range of different areas where the behavioural literature is full of examples.

Dr David Halpern: In round numbers, 5.2 million extra savers had come through the auto-enrolment route by March 2015. You can see what the graph looks like as it comes into force and moves through the population. It is incredibly powerful in its effect.

Q39 The Chairman: In other walks of life, of course, we have been pressing people to move the other way. In financial services, we have been saying to banks, “You should move in the direction of making sure that people opt in, rather than using their inertia to help the company”. Presumably, you are arguing that the virtue of opting out really applies where there is a strong public interest in the policy, as opposed to financial services, where we are saying that there is quite a strong public interest in people not being imprisoned by inertia.

Lord O’Donnell: There is strong public interest, but I would use organ donation as a good example. We know from surveys that if you ask people, “Do you want to donate your organs?” at least 75% will say yes. How many people have actually signed up to organ donation? The figure is 25%. We know that there is a mismatch. People want to do it, but they have not got around to it for some reason. The reason they have not got around to it is in the title of a book by Sunstein—keep it simple. If you want people to do the right things that work, just make it simple for them. Hence, there have been attempts to think about how we can make it simple for people to sign up for organ donation, to fulfil their own desires—not anybody else’s desires, but their own.

Lord Sherbourne of Didsbury: I want to pick up the point about ignorance, which I think I understand. Are you saying that, if you have a system of opting out, people tend not to do it, either because they cannot be bothered, through inertia, or because they are unaware of their right to opt out?

Dr David Halpern: As Lord O’Donnell said, the inertia effects are incredibly powerful. We go with the friction. Recently, a powerful experimental result by George Loewenstein looking at defaults, as we call them, showed that to be the case, even when you reveal to people that a default has been set arbitrarily. He ran an experiment particularly relevant to the US healthcare system, where people have a choice about what kind of insurance premium they pay. It is a significant choice. In crude terms, do you want to pay a lower premium, but when you get knocked over they switch off the machines sooner? How you set the default is quite consequential. They set up the experiment and made it clear to people: “It is for you to choose. We have just set it one way or the other. You should check that that is how you want it”. It was a randomised control trial, so people could have it one way or the other. People showed a very strong tendency to stick with whatever the default had been set at, even when it was drawn to their attention, in quite a consequential area. Again, that gives you a sense of how powerful defaults are. It is not just a question of transparency.

Lord O’Donnell: It is not just about whether you are in or not; it is often about amounts. If we go back to the pensions example, there is some evidence from US firms. A firm said, “The default is that 3% of your earnings go into the pension scheme but, by the way, we are going to match pound for pound everything up to 6%”. The default was 3% and, when they had that, over 25% of people stuck with the default. For new hires, the same firm decided to put the default at 6%, and hardly anybody went for 3%. The defaults matter not just on participation but on amounts.

Q40 Earl of Kinnoull: Your points about process engineering affecting what happens are very clear. Do you have evidence about whether the availability of e-techniques in the process is particularly helpful in engineering a true result, as it were?

Dr David Halpern: There is a lot of interest around that. There are lots of other things going on, but there is some evidence that people’s behaviour in e-defaults is different from that in paper defaults. Partly because it is so easy to click on or off, people seem more willing to change a default in the electronic domain. It may also be that people have learnt to be cannier. The everyday example is Ryanair, where there are lots of other charges. People are starting to learn a heuristic—a shortcut—which is always to unclick the box at the bottom that says, “Do you want your details to be shared with everyone else?” Because it is so easy to do that and people have become familiar with it, it looks as though they have become readier to make the changes.

I emphasise and echo Gus’s point that what categories are inside it looks like an incredibly important issue as well. The interesting thing about the issue I know you are wrestling with is that we know that a change in default is likely to have a big impact—although it depends a lot on the underlying intention of the population—but it is also an opportunity, potentially, to introduce other choices that would enable people to differentiate. That could be very consequential. In other words, you might say, “Do you want to give this amount, this amount or this amount?” That is a very powerful variable.

I will illustrate it with an everyday example. Most people choose the middle option on most things. The common one is coffee sizes. Most coffee chains offer three sizes. It turns out that, across countries, the absolute size of those varies greatly. If you ask people, “Which size do you take?” generally speaking 74% of people choose the middle size, across countries, but the absolute size is different. If you ask them, “Why don’t you change to the large?” they say, “It would be much too big”. There is a kind of Goldilocks effect. If you change it experimentally, so that the medium becomes the small, the majority of people shift to what was the large.

Lord O’Donnell: It translates to financial issues. When you see the little boxes on the JustGiving page—“Do you want to give £25, £50 or £100?”—I would hope that they have not been chosen randomly, but that the organisation has worked out precisely what impact the number of boxes and the amounts in them have on behaviour.

Lord Callanan: It is what Milton Friedman called “the tyranny of the status quo”. It is just fear of change—people stick with what they know. Like many people here, I study consumer behaviour and the politics of it. I know that I should change my bank account, because the details of the current one are crap, but the trouble is that I cannot be bothered to get round to it. I have seen all the figures, I have read the interest rates, et cetera, and I know I should do it, but I have been with the current one for 30 years and I just cannot be bothered. You can see it in your own personal behaviour sometimes.

Lord O’Donnell: It varies by area. There is quite a lot of evidence that people get very nervous about financial things. When nervous, they tend to default to doing nothing. They kind of assume that someone set up the default in a good way, so in public policy areas having the defaults in the right place is very important.

Dr David Halpern: Yes, hugely so. It has been said before that you are much more likely to get divorced than to change your bank. There are other everyday examples. How you set these things is a huge question for policy in general, beyond this issue, because they are so consequential. In energy markets, why is it that only one person in 10 has switched when there are so many choices? You might think that energy companies are quite skilful at adding little, subtle degrees of friction in order to discourage you from switching, so the policy response becomes, “Make it easy”. One thing that came in at the end of last year is that on your bill you should now have a QR code, which summarises your data. Why is that consequential? It means that you can put your phone against it and in a few seconds it will tell you, “This is the best tariff for you. Click here and you can switch”. It becomes a game changer to lower the friction. That is why the e-thing is very important. It is easy. How many seconds would it take you to make the choice?

Lord O’Donnell: Two clicks versus one click has been huge.

Q41 The Chairman: Does the amount make a great deal of difference, too? We are dealing with a contribution of £5 a year, on average, to the political fund. It probably varies between £15 and £1, but the average amount is about £5 a year, from what I can see from the numbers that we have had. Does that lead you to think that the effect of inertia would be greater or smaller?

Dr David Halpern: The wonderful thing about price psychology is that, like most psychology, there are two things in it. First, there is a discontinuity between zero and any small amount. The everyday policy example is adding a 5p charge for a plastic bag, which has been seen to be absolutely spectacular in its effect. People do not think twice about 5p, but when confronted with that versus zero, suddenly there is a 75% or 80% reduction in the number of plastic bags. People are crazily sensitive to even a tiny charge.

When you get beyond that, the main thing is that people often struggle to know what the absolute value is and what a reasonable amount is. The golden rule is that people can work out relative values—how good is this piece of modern art or this bottle of wine? They find it quite hard to work out in absolute terms how much you would pay for it, but they can be pretty good at saying, “This one is better than that one. I will make a contrast effect”. That is why the framing of relative choice categories is so important, not just the absolute amount. If the question is not just, “Would you like to give £5?” but, “Would you like to give £5, £20 or £50?”, what will people’s choice be? They may also be very unclear about what the previous amount was. They will probably be extremely sensitive to the way the choice is presented.

Lord O’Donnell: Way back—about 30 years ago—traditional economics looked only at price and supply and demand; only price and amounts mattered. We have moved an enormous way beyond that. We now have evidence that it depends not just on pricing and incentives but on who the messenger is, what the social norms around it are, what the defaults are, what kind of priming you have and a whole bunch of other issues—how salient it is, the effect, the commitment and the impact on your ego. What I have just gone through is a mnemonic called MINDSPACE, which goes into all those things. They are all important. Particularly in David’s example of the subsidies that we give savings, prices turn out to be, first, not very important and, secondly, incredibly inefficient at influencing behaviour.

Lord Whitty: Given that the default position is the preferred position for most of the population, does it vary at all depending on the number of times and how frequently people are asked? Presumably, in the pension scheme you have not had a long enough run to see it, but there are other equivalents. If you are given a choice more frequently, are you likely to default more frequently or to pick it up more frequently?

Dr David Halpern: It depends a bit on what your underlying preference is. A big issue that hangs over us is giving people the right to choose not to choose. There are so many possible choices in the world. Do you want to be asked every single time you go in? Sometimes people would rather not be bothered, so you solve it in some other way. You collectively make a decision and reach forward.

Going the other way a bit, an example of multiple prompts is where we are now on organ donation. The defaults on that have been much discussed, although sometimes the issue is slightly overstated. Some countries have a default of presumed content, but some have it the other way around; you have to choose actively. In the UK, we have moved to a situation of prompted choice. We are introducing multiple places where we keep prompting and reminding you. That is steadily dragging up the number of people on the register, and people seem accepting of it. Will you eventually reach the point when you are sick of being asked again every time you do your tax disc, because the system is too stupid to know whether you said yes or no last time? Perhaps, but we do not seem to be at the frontier at the moment.

Lord O’Donnell: Look at the way companies have tried to make the most of human behaviour and inertia through the strength of direct debits. How many of us have signed up to a free trial of Netflix or whatever? After 40 days, it is quite complicated to get out of it, so inertia keeps you in. You sign up to various memberships through direct debit, so you are not asked again. The reason they do not want to ask you again is that it can go only one way. If you are in, it is great for the company, so it really wants to keep it that way.

Lord Whitty: I do direct debit to a number of charities. If they keep reminding me, I do not cancel it, but if they ask me to increase it, generally speaking I do not. That may be meanness on my part. There is an underlying commitment, but constant reminders are increasingly irritating, so you do not go any further.

Baroness Drake: You have two defaults—you stay in, and you settle at the lowest level.

Lord Whitty: Exactly.

Dr David Halpern: When it is used in pensions and savings, you have a so-called smart default that has some escalation built into it. That is what many charities would love to get you to do, too, so that when you make the initial commitment you say, “It will also go up by inflation”, or something. That is essentially what we have in the pensions environment at the moment. One of the unknown questions, as the escalator moves in—because people’s initial contribution is quite small—is: will it reach a point where more people opt back out? It is generally not known. The evidence to date has shown the overwhelming force of the default.

Lord O’Donnell: There is an example where you can test that, although I do not think anybody has quite done it yet. You are asked every year about, for example, car insurance. Let us say that for some reason you have set up car insurance and house insurance so that they are on annual. The company would love you just to reinsure with it. When it sends you the renewal notice, you would look very hard to try to find what the increase on the previous year is, although in fact you almost certainly will not. However, if you go online and use a search engine, you will find that the same company would give you the same deal much more cheaply. People are starting to learn that the default—staying with the same company—is a very bad deal.

Take the example of annuities. In the old days, before the freedom, people could either take their annuity from the company they saved with or shop around. You definitely gained by shopping around. Did people all shop around? No. Quite a lot made a very bad financial choice, which was just to stick with the company. That is why I think that people will need help to make the right choices on annuities, although I am very pleased that we have the freedom. We need a bit of libertarianism, but we need a bit of paternalism as well—hence liberal paternalism.

Q42 Lord Wrigglesworth: This morning, you have given some very powerful evidence on people’s behaviour in various respects. Have you any reason to believe that people would behave any differently in relation to political funds in trade unions?

Lord O’Donnell: I am not aware of any direct evidence on political funding. The only thing I would say is that, in all the behavioural stuff, we emphasise our humility in trying to understand human behaviour. That is why David mentioned randomised control trials. You may have quite a strong prior belief; I would have a very strong prior belief that defaults would be very important in this area, but I would want to test precisely how important. I am not aware that any tests have been done, although they may have been, or of any other country that has done them. The only thing is the history, which gives you some evidence, and the Committee has that already. The feeling of both of us about the behavioural stuff is that you will not get precise estimates. We all thought that the auto-enrolment effect would be smaller than it turned out to be. We were surprised by how important the default was there. My gut feeling is that defaults will turn out to be very important.

Dr David Halpern: History is a pretty good guide. It is interesting that the numbers from the history are pretty close to what you see in other domains. You will probably see a 30-odd percentage point movement—possibly more—with more friction.

Lord Callanan: You also have the fact of Northern Ireland, of course.

Dr David Halpern: Exactly. It seems to me that the slight imponderable that bears on this issue is that it depends on how you feel about the source. If you look at it over a historical period, the world has moved in other ways in relation to mass party organisations and memberships, not only in the UK but elsewhere. If the underlying attitude towards parties has changed, maybe the effect will be even larger than history tells you. The interesting countervailing detail is that, if you end up with some kind of prompted choice that moves the default in that kind of way, there is an opportunity really to look at the detail of how it is constructed. You might well find that some people were prepared to give much more substantial amounts if they were offered the choice in a prompted form. In fact, an expression of underlying preference would come through more clearly; some people would not get around to doing it, but other people would probably give more.

Lord Wrigglesworth: You have also made it clear that the background for this and the method by which it is being done have a powerful impact on what happens.

Dr David Halpern: Absolutely.

Lord O’Donnell: Absolutely.

Lord Wrigglesworth: In the case we are looking at, electronic communication, as against writing, on paper, could have a substantial impact.

Dr David Halpern: It absolutely could. One thing that this whole area has taught us is that all these details, which generally do not interest the Cabinet, turn out to be hugely consequential. Literally, the colour of an envelope, the way something is worded or the alternative contrast all make impacts.

Lord O’Donnell: I can give you a stronger result for getting people to switch from using paper statements to using online statements by making sure that the button flashes on their screen. There are little things like that. That is what you do when you are doing a proper behavioural test: you try out all those sorts of things in randomised control trials. The difference between the worst and the best is usually pretty dramatic.

Q43 Earl of Kinnoull: Could you tell us about the propensity of people of working age at the moment to use the internet for financial transactions or decisions? What percentage of people do that? Could you also give us a view on the direction of travel there, so that we can understand it?

Dr David Halpern: I am not sure that I can give you a confident statistic. Obviously, there are some very big generational effects running through at the moment. To go back to the simple example of energy, we ran trials to see whether we could get more people to switch online or use their phones, but some of the most at-risk groups are not really using the internet or phones to do that. Including an envelope, so that they can put in their details and send them off, is much more effective for those groups. It will vary quite a lot. There continues to be a lot of unease in large sections of the population about the use of the internet for banking, although I know that banks are busy trying to make people more comfortable.

Lord O’Donnell: They are. Virtually every bank is now developing both online and mobile services quite dramatically. There is a lot more on mobile now internationally. That is the way the world is going. It has gone quite a long way. Interestingly, I was talking about issues like BBC licences, to give a random example. It turns out that quite a lot of old people are interacting online for those sorts of things, but there is very much an age issue. As you go up the age profile, the online—

Earl of Kinnoull: Can you give us a feel for that? Is half the population now comfortable with taking decisions and doing financial transactions online? Is it less or more?

Lord O’Donnell: There are two issues. It depends what you mean by a financial transaction. The reason why banks are spending so much time moving people online and to mobile is that they have observed that people who said that they were not interested in those sorts of things are transacting online when it comes to buying products. I will not advertise for them, but lots of companies are now selling stuff online. There has been a dramatic increase. When people start to become happy buying clothes, books and any particular product—you name it—online, it does not take much to go into banking online. The penetration has gone up dramatically, but it varies area by area. The reason why we are not giving you a single number is that it is hugely different in different areas. Vast numbers of people are prepared to use Amazon, let us say, online. Slightly fewer would be prepared to bank online.

Dr David Halpern: The majority of the population now have smartphones, so clearly they could do it routinely. On the other hand, a sizeable proportion of the population are not ready to do all their banking in that way. It should feel congruent to how you normally interact with the channel. Again, I use the example of electricity bills. If you are used to doing it in paper form and that is how you have always transacted with your electricity company, you feel more comfortable responding in the same way. If you are 25, you are used to doing it all online, so you think, “Why do you even send me this bit of paper any more? You know I do it all online”. There are some cohort effects moving through.

Q44 Baroness Drake: I looked up your MINDSPACE list of behavioural influences. Let us stay with the default—the opt-out—and ignore for the moment whether an opt-out default decision-making process is desirable in a given situation, because that is a different debate. If participation in something is done by opt-out and then there is a change to do it by opt-in, does the evidence support the expectation that participation rates will fall as a result of that change in the decision-making process?

Dr David Halpern: You have it in the history of this area already. If you look over the century, the cycle gives you a strong sense that, broadly, you get a kind of symmetry. People stick, and you get inertia effects one way or the other.

Sometimes there are examples where people create an event or provoke a choice, when they want it to be more active. It is not very clear, but one area where it has been rehearsed widely in the literature is around generic versus other kinds of drugs, particularly in the US context. A lot of campaigns seemed to struggle to get people to move to generic drugs, but when the patient was required to make a choice, the vast majority—more than 90%—made a choice for generics. It was not the case that there was some underlying preference. If you can create a precipitating moment when you have to choose, sometimes you get much larger numbers. That would be one interesting question about this.

The use of deadlines is an example. It may seem paradoxical, but often the use of a deadline will greatly increase the number of people who conform. You always procrastinate, but you also get a social norm effect—everybody else is making a decision at the same time. A recent one would be voter registration. There were quite effective campaigns, using deadlines, to drag large numbers of people across the line at a given time and moment.

Baroness Drake: In the absence of other factors that force more active decision-making, is the proposition reasonable that, if you switched from opt-out to opt-in, you would expect a lower participation rate?

Dr David Halpern: Yes, substantially so—20 or 30 percentage points lower.

Baroness Drake: There would be a substantial reduction in the participation rate.

Dr David Halpern: Yes.

Baroness Drake: I listened to what you said about companies being increasingly aware of how they engage with individuals electronically. If you are operating on an opt-in basis of decision-making and people’s opt-in options are to do it by post or in writing, rather than online, is the opt-in rate likely to be higher if it is done electronically, as a general proposition?

Dr David Halpern: It depends on how you normally transact. That would be one of the questions. Is it congruent? How would you normally have done it? If you are in a situation where you are routinely transacting in this domain online, you will get a lot more people to come across, as generally we know that online there is lower friction. There is a slight issue. If in your mind you have always transacted with your employer through bits of paper, in a certain way, and then you say, “Go over here to do it differently”, you will get a lot of loss from that movement between channels.

Baroness Drake: If a population in general, as opposed to a specific population, is reasonably engaged in transacting or engaging online and you introduce into that doing something by post, it is likely to have less impact than—

Dr David Halpern: Yes. We are saying that it works both ways.

Lord O’Donnell: It goes back to the basic, very strong principle of behavioural economics, which is, “Keep it simple”. People will do the simplest thing. If post is much harder than online, you will not get them doing as much.

Baroness Drake: The inertia barriers might be reduced as well, because you do not use the form, you do not have to walk down to the postbox in the rain and stuff like that. You take a lot of behavioural barriers out of the way.

Lord O’Donnell: Yes.

The Chairman: Thank you very much. That was both fascinating and helpful to us in thinking through the issue. It was very good of you.

Lord O’Donnell: I hope you will all buy David’s book. It is all there.