British Growth and the Graduate Premium

Tom Forth, .

In early 2023, Anna Stansbury at MIT Sloan and Dan Turner and Ed Balls at Harvard published one of the most important summaries of Britain’s regional economic growth problems to date.

“Tackling the UK’s regional economic inequality: Binding constraints and avenues for policy intervention” echoed, reproduced, and improved on much of what I’ve been writing about for over a decade.

I agreed with it all. I’d written a lot of it online as far back as 2013 and pretty constantly since. And while in some places I wished the data was even better, it was mostly better than mine, and Anna and others have since worked to make it even better still.

But most interesting to me was where Anna looked at something I had never looked at, and hadn’t heard other economists talk about. She calculated and plotted the graduate premium for each region and nation of the UK (Northern Ireland only got chopped out in the FT visualisation) separately and showed that as the graduate share had increased, the graduate premium had fallen. This was especially true for those with non-STEM degrees and the pattern was consistent in every region, except London where graduate premiums had held up.

While this matched what I felt myself and had heard at every business roundtable I’ve attended in Leeds, Manchester, and Birmingham over the past fifteen years — poor infrastructure and lack of government support in the region always tops availability of skills and planning permissions as a constraint on growth — this has long been heavily questioned by most British development economists. To my shame, and I think even more to the shame of those economists, Anna seemed to be the first person to actually get the data and do the calculations.

You can see the graphs in Anna’s paper. They’re complicated and not that beautiful. So you’ll probably prefer to look at them via the excellent simplification in the FT published later in 2023.

The graduate premium has fallen in every region of Great Britain except London as graduate share has rise. In the USA, it has kept rising.

The expansion of the analysis to the USA, where a similar increase in graduate share had occurred at the same time as an increase in the graduate premium, widened the interest enormously.

In the time when John Burn-Murdoch was sourcing comparable US data on graduate share and graduate premiums, British economics had been busy.

I was at the time an advisor to the Resolution Foundations’ Economy 2030 Inquiry which involved lots of trips on incredibly delayed and frequently cancelled TransPennine Express trains to Manchester from Leeds. Before one such meeting at the beautiful Greater Manchester Chamber of Commerce I arrived to find a room of policy and academic economists all having read Anna’s paper. They were a fantastic mixture of embarrassed they hadn’t done it first, frustrated they hadn’t done it first, and desperate to check if it was right.

There were some moans about the methodology, and then quickly an agreement that they’d try to reproduce the result with both the published methodology and another that they thought was better. Ahead of that, the biggest suspicion in the room was that while the graduate premium had likely fallen in the West Midlands and the North West of England, it had probably grown in the largest cities of those regions. It was agreed to look at the graduate premium in Manchester, Greater Manchester, Birmingham, and Greater Birmingham (the WMCA area). Some other cities were added later on.

The result was a surprise.

Many British economists expected that graduate premiums would have kept up in other cities in London, even if the regional premiums fell. They were wrong.

Not only did the original result looking at graduate premium reproduce by both methods, but it also reproduced when looking just at the cities, by whichever definition was chosen. And still, London bucked the trend, maintaining its graduate premium when everywhere else it had fallen.

As I have told many people, I was hugely impressed by the rigour, urgency, and efforts of the people who did this work. I had disagreed with many of them for over a decade on the importance of skills to productivity in big British cities. They desperately wanted Anna’s paper to be wrong. But they accepted that it was not wrong with speed and grace.

The Economy 2030 Inquiry was not the place for a complete change in direction in British thinking about skills and economic growth. The report, which I read in a Wetherspoons in Gateshead before catching up with a friend from school and later fellow skills-sceptical economic geographer Nick Gray of Teesside University, continued to argue that increasing graduate share in Greater Manchester and Greater Birmingham would be key to achieving the catch up productivity growth that Britain so desperately needs those cities to lead on. I got round to finishing the most complete summary of my disagreement on that narrow point in a blog post on City Size and Productivity in August 2025.

In recent months interest in graduate premiums has broadened in Britain. Graduates with much higher student loans thanks to England’s large increase in university tuition fees are starting to see, after a decade in work, just how much is taken from their pay packets and how small their repayments are compared to the size of the loans and the interest they are accumulating. Continued underperformance of the British economy despite the recent huge increase in graduate share of the population has pushed more and more people to question what was until recently a near-consensus in British economic thinking that a more highly skilled population would create a more productive economy.

In his most recent piece, ‘Is university still worth it?’ is the wrong question, John Burn-Murdoch shows that Britain is unusual in having seen its graduate premium erode as graduate share has increased in recent decades. British consensus thinking on this was reasonable, even if it turned out wrong.

The UK is an outlier in having seen its graduate premium fall as the graduate share has risen.

There is much good discussion to be had of the impact of high and rising minimum wages and Britain’s pronounced stagnation in productivity at a level below our neighbours and rivals. But “if Britain can haul its productivity growth and skilled job creation back into line with its peers, graduate earnings will be stronger” as the piece ends is largely uncontested.

Of course I agree too. But I also think that it’s only a slight oversimplification and exaggeration to say that a consistent theme of our national plan for growth, backed by a near consensus among Britain’s national economics institutions in South East England (economists in the rest of the country, especially in Scotland, were always more sceptical), applied by three parties of government in thirty years, was hoped to achieve exactly this. It really was believed that skills were the main constraint to British growth and that by increasing the skill levels of the population, as measured by the proxy of graduate share, especially in areas with weaker economies, we would unleash productivity growth.

Blair’s three priorities in government were “education, education, education,… because nothing matters more to the future prosperity of the country”. In a fantastic late 2021 episode of The Centre for Cities podcast Andrew Carter, talking with East German academics on how East Germany had caught up and overtaken similar regions of England summed it up very well. “We hear this a lot in Britain that one of the ways we’re going to level up is large investment in physical capital whether it’s land or buildings and transport and actually as we all know it’s investment in human capital that makes a material difference over time” he said. This really was, and still is quite close to, the consensus in Britain’s national economics institutions.

I wrote a whole blog post on how, as Andrew’s interviewees say, human capital [skills] is not what made East Germany’s cities grow at twice the rate of North England’s since they abandoned communism.

East German cities have grown their economic strength about twice as fast since reunification as North English cities.

I don’t think skills, or graduate share, should be our focus for growth now. Though I have quite a lot of time for arguments to return polytechnics to local control and hugely increase their funding via local taxes. If skills matter, and I think their importance as a measurable property is overstated, graduate share is not a good way of measuring them.

So what should we do?

I don’t think that Britain should try to increase its graduate share any further. In my region of Yorkshire, home to two of Britain’s largest cities, Greater Leeds (2.2m) and Greater Sheffield (1.6m) the graduate premium is already so low that according to my very rough maths, adding more graduates will actually damage the economy.

Iain Mansfield has done more serious work on this and he leans towards reducing university participation by reducing public subsidies, especially on courses with low expected economic returns. I’m sceptical that central government will make as good decisions as young people and their families on that, but that’s a disagreement on the detail. I think we’d both strongly support using the vast data we have on graduate outcomes to better inform those considering going to university. Some, or many, would choose not to.

Up to a point my opinion on this would remain the same even if the graduate premium started to increase in Britain. I am convinced by arguments that the economic value to the graduate of university attendance is substantially, somewhere between a third and two thirds, about signalling. Graduates will, for lots of reasons good and bad, often get the best jobs if we can create more of those jobs. And if there are too few graduates to fill them, I expect that non-graduates will take their share of them with only a small reduction in productivity, on average.

But if not by further increasing graduate share, how should Britain “haul its productivity growth and skilled job creation back into line with its peers”?

My suggestions remain astonishingly unchanged from 15 years ago despite an enormous change in my circumstances and experiences. Just as in my piece on how East Germany overtook North England, and in pieces I have published every few months for a decade, I think that we need to stop focusing on human capital as a path to prosperity. We should,

  1. Start investing heavily (and ideally more efficiently) in better transport, especially in our big cities, so that we can achieve the agglomeration benefits that we currently miss out on completely.
  2. Move our national public R&D funding away from places with low private R&D investment like London and small places that refuse to grow like Cambridge and Oxford and fund much more scientific research in the big cities with cheap land, homes, and offices like Birmingham, Liverpool, Leeds, Manchester, Sheffield, Newcastle, Belfast, and Glasgow that built our country into an intellectual industrial powerhouse in the first place.
  3. Reverse the centralisation of the British state that has kept North England poor and removed its agency to succeed by itself when the UK government will not back it.

We should be very clear that we have not tried any of this in the last three decades. We have done the opposite and centralised ever more power and investment in South East England.

I get a lot of pushback, sadly often nasty, from people that my three points are worthless without a well-developed theory of change. They suggest I move to London, where they almost all live and work, to develop one. Change things from the inside, they say, while quietly, I suspect and sometimes see proof of, thinking that I’m not good enough to succeed outside of the provinces.

I agree with them that with all of my suggestions reducing the power and short-term prosperity of our current capital city, and thus of people close to power in Britain, there’s not much in it for those with the power to change things. And while I think that in the long run London will be a much more prosperous city, with a much higher graduate premium, if it doesn’t have to spend 8% of its GDP on a national welfare state the rest of the country can’t afford, the long run is a long way off. So maybe I should move to the capital and try to change things from the inside.

But I’ve seen many people like myself, some of them better and smarter people than me, try that path. And all the data shows pretty clearly that it hasn’t worked.

So I’m going to stay on my own path. Build a business and accumulate wealth which I can spend to buy that theory of change from other more talented people, when I have enough money to succeed.

blog comments powered by Disqus