You have spent your entire academic career at the LSE and earned your PhD at the University of California, Berkeley as a Fulbright Scholar. Looking back, was there a particular moment or experience early in your career that set you on the path toward studying the economics of the welfare state?
When I went to LSE as an undergraduate, I was blown away by my first-year economics teacher, Richard Lipsey. So I promptly switched into the economic theory track for the rest of my degree, which then led on to a master’s degree at LSE and, as you said, a PhD at the University of California, Berkeley, including a year at MIT.
Why the welfare state? I was looking for a thesis topic and was out in the park in Berkeley talking to a friend who pointed out that I had always been interested in the welfare state and suggested that I write a thesis about some aspect of social policy. There was a dataset about US welfare recipients which I used for my PhD thesis and, having started on that track, then continued on it.
Your work spans theory, books, and direct policy advice to governments across several continents. How do you see these different roles – academic, author, adviser – reinforcing each other, and has the balance between them shifted over the course of your career?
They all contribute to each other. There’s a feedback loop. I started out as a junior academic working on the theory, but I’ve always been interested in policy. When I was first involved in a policy question, I started from the theory and translated that into policy. That policy work fed back to theory. So it’s been a constant feedback loop between theory and policy.
I see my role as a policy advisor in two parts. Wearing my hat as an academic, I want to draw a circle. Outside the circle are policy designs that economic theory suggests are demonstrably crazy – like over-reliance on private medical insurance as the main way of financing healthcare. Inside the circle are those policies which, if put together sensibly and implemented properly, have the makings of something that is likely to work. In my role as policy designer, the second part is to try to make sure that the political compromises are those that do the least damage to the underlying strategy.
You have long advocated for income-contingent loan systems as the most equitable way to finance higher education. For a student in continental Europe, where tuition is often low or free but public funding is under pressure – what is the strongest economic argument for rethinking that model?
My argument is that a combination of taxpayer support and income-contingent loans is not only the most equitable way but also the most efficient way to finance higher education. There are at least three arguments for a cost share. One is that higher education creates social benefits – it’s right the taxpayer should pay for those – but also creates private benefits. People with degrees on average earn more and – often underestimated – typically have work they enjoy more. So job satisfaction is very much a return to the degree. Where something has external benefits, the cost should be shared, with the taxpayer paying for the social benefits and the individual beneficiary for their private benefits.
The second argument for a cost share is affordability. Time was when higher education did not really matter; only a few people went to university so higher education could be financed from taxation. Since then technical advance means that countries need more skills, a greater diversity of skills, and because skills go out of date more quickly, need more repeated retraining. So today we need a mass high-quality system of tertiary education which is simply unaffordable from taxation alone. As a matter of brute politics, higher education will always lose out to healthcare, pensions and other priorities.
England has made something of a mess of its higher education reforms, but there was one time we got it right – the 2006 reforms, with which I was involved. The 2006 reform had three legs: financing universities through a mix of taxpayer support and tuition fees; financing students mainly through income-contingent loans; and policies earlier in the system, from nursery education onwards, to widen participation. There is something I call pub economics – something that’s obviously right, and everyone knows it’s right … but it’s wrong. Pub economics argues that fees harm access. What the evidence shows is that what stops people from poor backgrounds going to university is that they don’t get good high school graduation grades, because of problems earlier in the system. Access to higher education is more a 0-18 problem than an 18+ problem.
With the rapid rise of artificial intelligence, online learning, and freely accessible knowledge, where do you see the unique value proposition of a traditional university education in five to ten years – and does the enormous price differential between traditional institutions and AI-enabled online self-learning tools still justify itself?
There is a bit in portfolio theory that argues that if a new financial instrument is devised, everything shuffles up – you end up with a different mix. Cinema didn’t kill off live theatre. When DVDs first became available, people were saying the whole world would be dominated by Microsoft University. It didn’t happen. Zoom has not killed off in-person meetings. After Covid I am back in the classroom giving lectures as a live performance because in-person is different and more powerful. Innovation changed the mix, but did not kill off the old technology.
AI might be different, but if I have to guess, I would say that AI might hollow out mid-level skills, leaving tertiary education providers to do what they are best at. One of them is building analytical capacity. I always tell my students at the London School of Economics that I am not there to give them job training, I’m there to give them intellectual training. If they can develop an independent analytical capacity they can apply that to whatever circumstances they find themselves in – broad, flexible skills allied to imagination and creativity. Face-to-face teaching will still be an important part of the picture.
Critics argue that income-contingent loans ultimately discourage students from lower-income backgrounds, despite being designed to do the opposite. From your experience across different countries, what is the single most important design feature that determines whether a loan system genuinely promotes access?
Mortgage-type loans work well for home loans because the house acts as collateral. But if you borrow to get a degree there is no collateral, and that makes it risky. Conventional loans would undoubtedly discourage students from lower-income backgrounds – that’s exactly the case for income-contingent repayments which reduce the risk faced by the borrower because the income-contingent formula protects low earnings. If you take out a loan and then have low earnings, you make low or no repayments. If you take out a loan, have a job, but then become unemployed or leave the labour force for childbearing, your earnings fall to zero and your loan repayments instantly and automatically fall to zero. Thus loans have built-in insurance – nobody goes bankrupt.
But well-designed loans are not the main driver of participation. The determinants of access lie much earlier in the system. I remember the best education minister I have ever worked with losing his temper at a student debate. ‚If I were a real socialist,‘ he said, ‚I wouldn’t spend a penny on higher education, I’d spend it all on nursery education.‘ He was exaggerating, of course, but the underlying point is right.
The 2006 reforms showed the way. Between 2006 and 2012, the number of applicants from the most disadvantaged backgrounds rose by 53 percent. That staggering increase was not because of the loan and fees regime – it was because of policies earlier in the system. So what is needed are loans with income-contingent repayments to finance university studies together with policies much earlier in the system to make participation more equitable.
A 25-year-old entering the workforce today will retire around 2065. Given what you know about demographic trends and the current design of pension systems, should that person be optimistic or worried?
My bottom line is that it depends on what happens to worker productivity. A response to higher life expectancy includes longer working life. In addition, fertility has declined in many countries which means that the future workforce will be smaller than it would otherwise have been. A rational policy is to make each member of that future smaller workforce more productive by investing in their human capital and in the physical capital they use.
The question is how realistic is an increase in productivity; the answer is that there is a huge amount of uncertainty. On the one hand, technological progress and artificial intelligence might lead to significant improvement in productivity. On the other, there are wars, pandemics, and other major disruptions. If I were young I would be cautiously optimistic but only cautiously optimistic. Bottom line: the policy mix needs to include longer working life and greater investment in human and physical capital.
You have advised on pension reform in countries as different as China, Chile, Sweden, and the UK. Is there a common mistake that governments keep making, and if so, what drives it?
The worst problem is short-termism. Governments know the problem is coming but kick the can down the road. For how long have we known about the 1940s baby boom? Since some time in the 1950s, i.e. we’ve known about the problem for 70 years. If governments had linked pension age to life expectancy in the 1950s the problem would be much less acute today. The pension systems in Italy and Greece illustrate how things can go wrong: as a result of their well-known unsustainability, the systems broke down under the stress test of the 2008 economic crisis, resulting in very sharp decreases in pension benefits.
Sweden got it right. In the 1990s they realised that their pension system was unsustainable and set up a committee, the Pensions Group, made up of all the then major parties in parliament, to design the reform. Once the legislation had passed, the government turned the Pensions Group into a standing body – the guardians of the reform. As a result, the pension system was adjusted rarely and carefully with a long-term view. That was possible because Sweden can get long-run cross-party support for things like pensions that need a long-term view. It made my heart ache, because with Britain’s adversarial politics, long-term cross-party political agreement simply doesn’t happen.
Canada, in an interestingly different way, also got it right. The law that frames the Canada Pension Plan has protection almost as strong as the Canadian constitution – it is not possible to change the pensions legislation through a simple majority in the federal parliament. Change requires a majority in the federal parliament and in the parliaments of two-thirds of the provinces representing at least two-thirds of the Canadian population.
The big problem with pensions is governments taking a short-term view of something that needs a long-term view.
The concept of intergenerational fairness is rarely defined precisely. As an economist, how would you define it – and are today’s welfare states failing that test?
In some countries, at least, they are. Inequality has risen substantially in countries like the United States and United Kingdom, and my personal view is that it has risen too much.
There is also a conceptual point: complete intergenerational fairness is possible only in a world with static demography and static technology. If the age structure is constant and output is constant the system can simply roll on from generation to generation. But the moment some cohorts are larger than others, or growth is different in different periods because of wars or oil shocks or pandemics, there is no single definition of fairness. Suppose that my generation is large and your generation small – either your generation has to pay higher contributions so that I get the pension I was promised, in which case costs fall on the younger generation; or you don’t pay higher contributions, in which case the older generation does not get the pension they were promised. The costs arise because the cohorts are of different sizes, and any solution will impose costs on one or the other group, or both. The idea that you could have complete fairness across cohorts is a myth. What is needed is a new social contract, as Minouche Shafik discusses in her book What We Owe Each Other: A New Social Contract (Princeton University Press 2021).
Your academic work on market failures in the welfare state has directly influenced legislation in multiple countries. When you sit across the table from a finance minister, how do you translate an economic model into something politically actionable?
The first thing is to let them know that I understand what it looks like from their point of view – you let them know that you know where they’re coming from. As I mentioned earlier, I then draw a metaphorical circle: outside it are policies that are crazy, inside those policies which, if put together sensibly and implemented well, can make sense. And the task is then to choose the political compromises that minimise damage to the strategy.
In a British context, I remember being invited to an event which the new education minister was going to attend. I figured that I would have maybe 90 seconds of his undivided attention and worked out the three key points I wanted to make. He listened to me with eyes half closed; when I had finished, there was a half-second pause; he then opened his eyes and said, ‘Very clear, thank you.‘ As a teacher, you can tell when something has landed. All one can hope for as a policy adviser is that the minister understands your argument and why you are making it. What they then do with it is a matter for the political process.
My view is that effective reform needs three sets of skills: strategic policy design, which is what academics like me do (in many ways the easy bit); practical political skills; and technical implementation skills. I have always been clear that politics involves skills that are utterly essential and which I don’t have. I love being on the fringes of the political process, but am not a natural politician.
Is there a reform you proposed that was adopted but implemented in a way that disappointed you – and what did that teach you about the gap between design and political execution?
In 1990 the government introduced mortgage-type loans. My colleague and I fought against that proposal, saying that student loans were the right policy but should have income-contingent repayments. We lost that time round and spent the next few years talking to journalists and policy designers. Thus by the time of an inquiry in 1997 support for income-contingent loans was unanimous. Income-contingent loans were introduced in the UK in 1998.
In 2006, the government implemented our strategy in full. On the day the White Paper was published, I got a call from the Secretary of State’s office saying: ‚You will notice that we have not abolished interest subsidies. The minister wanted you to know that he’s heard your arguments, he agrees with you, you’re right, but he feels it would be politically a step too far to get rid of interest subsidies at the same time as introducing variable tuition fees.‘ When it came to the crucial parliamentary vote – this was the Tony Blair government at the height of its powers with a parliamentary majority of 160 – the higher education bill passed with a majority of five. That was the minister making the right political judgment.
In contrast, reform in 2012 under the coalition government wrecked the system for short-term political gain. The lesson is: if you want effective reform you need a minister with three characteristics. They have to have the right motivation, they need to be bright enough to understand the idea of a strategy, and they need to be sufficiently a political big beast to prevent the strategy from being cherry-picked. Without a strong minister governments will cherry-pick – they will do the bits of the reform that are politically popular and not do the bits that are politically difficult. That is not the way to get good outcomes, but it’s politicians who decide, and I have to accept that.
You have described the heart of your work as exploring how market failures can both explain and justify the existence of welfare states. After decades of research, has your conviction in that thesis grown stronger or have you found cases where it reaches its limits?
When I first started, the main market failures on which I was working were information problems. But since then, there has been a growing literature on behavioural economics, search theory and incomplete contracts. Thus my conviction in the soundness of the economic arguments has grown stronger. But I’ve also learned that that is only part of the story. Politics also matters.
Something I have come to only recently is the importance of emotions. There was a scandal in Britain where our post office used a computer system that was flawed and led to postmasters and postmistresses being prosecuted as criminals. That the system was flawed had been known about for several years, documented on television and explained in the serious press. Then a television programme put on a three-part series called Mr. Bates vs. the Post Office – about the individual lives of a small number of postmasters. That made it personal. That got people emotionally engaged, and there was such a row that the government acted very fast.
Another example was the Brexit referendum. I posted a blog (‘Letter to friends’) explaining why I was going to vote to stay in the EU, and it went viral and crashed the LSE server. The best comment I got was from a colleague: ‚Nice work, Nick. Let’s hope it matters.‘ What he meant was: you’re being rational, but that’s not where the debate is. The Vote Leave campaign was playing to people’s emotions. I am now thinking more and more about how, if you want something to get into play and stay in play, you may need an emotional dimension to the way the proposal is presented.
If you could redesign one element of any welfare state from scratch – without political constraints – what would it be and why?
Give me the education system. Let me redesign it from nursery education onwards so that it runs as a whole – ideally from minus nine months onwards. All the evidence shows that the returns to investment in education and socialisation decline with age. So the greatest investment should be in the early years, which is why we should be spending a lot more on nursery education, and should finance higher education as a cost share. If my fairy god-person allows me one wish, it is to give me the education system.
Many of our readers are students or young professionals deciding between academia, consulting, public policy, or the private sector. Based on your own experience, what would you tell someone who wants their work to have a real impact on society?
Academia, consulting, and public policy are all necessary and important. Ideally, it would be good for people to have experience in more than one of those sectors. The UK is very bad at cross-fertilisation – academics tend to stay in universities, civil servants stay in the civil service, business people stay in business. The US is much better at having former politicians teach at universities, or having business people in government for periods.
My view is that the key thing is a good analytical training. Good policy work, whether in the public or the private sector, benefits from highly trained minds. In terms of jobs, people should start in the sector they think they will find most fulfilling, but should then take opportunities to work in other sectors.
And people totally underestimate the huge importance of serendipity – pure luck. Working at home one day I got a phone call from a colleague asking whether I would be able to free two weeks to go to Poland with the World Bank. It had never occurred to me to work for the World Bank. This was 1989 – the communist economic system in Central and Eastern Europe was on the point of collapse. Poland needed an unemployment benefit law, and I was invited to help. My generic message: get a good analytical training; go into the sector you think you’re most likely to enjoy; take opportunities to work in other sectors; and if luck occurs, grab it with both hands.
