Who’s afraid of the big band bond vigilantes?

Top of the charts

Afternoon all,

Autumn is in the air, Parliament is back in session, and party conference season has already begun. With Ruth at Reform UK’s shindig, I have returned to your inboxes to ease you back into the season of mists and mellow fruitfulness (#culture).

On Tuesday, Andy Burnham launched his premiership in the House of Commons, once again putting living standards – sorry, “breathing space” – front and centre. But any additional government support on that front is looking increasingly difficult to deliver as gilts respond precipitously to further conflict in the Middle East – more on that in Chart of the week. For those of you feeling out of the loop after too long in the sun, we and our friends at UK in a Changing Europe will be chatting on Tuesday about all the policy problems in Mr Burnham’s in-tray.

Below, we also look at bland bots, treacherous transitions, and ask whether it pays, or hurts, to be a manly man.

Have a great weekend,

Mike
Deputy Chief Executive


Bot-speak. How you write can reveal a lot about you – your education, where you’re from, and even your mood at the time. So what do we lose when AI is used as a writing assistant? Researchers (and, er, their AI tools) have analysed over 800,000 texts to reveal that AI editing reduced the variation in writing complexity by up to half. They also used AI to “polish” texts written by humans and then tried to retroactively deduce the authors’ characteristics based on the polished texts. The findings are striking: while core meaning was preserved, it became harder to infer any personal characteristics about the original authors because distinctive linguistic cues had been watered down polished out. So, next time you run a draft past Claude, check how much of yourself it has removed.

Standing still. This blog from the Bank Underground sets our recent productivity gloom in a long-run and international context. It suggests Britain’s post-financial crisis slowdown is neither new nor uniquely British, although it is particularly large. Market- sector productivity has grown just 0.5 per cent a year since 2008, behind the US (1.7 per cent) and eurozone (1.0 per cent), and down from 2.5 per cent in the previous two decades. Much of the global pre-2008 boom came from the IT revolution and manufacturing shake-up, but this had faded by the mid-2000s. The UK has fared worse partly because manufacturing shrank faster, but also thanks to Brexit, due to knock off an additional 3.25 per cent by 2028. So far, so gloomy. One hint of good news, though, is emerging from post-2024 data, and hinges on how you count workers. According to the ONS’s Labour Force Survey, productivity fell 0.2 per cent in the year to Q2 2026; on HMRC’s (more accurate) admin data, it grew 1.1 per cent, better even than the late 2010s’ 0.7 per cent.

Sun (investment) spots. The US Inflation Reduction Act sought to boost clean energy projects in fossil-fuel-dependent, high-unemployment, areas but new research found a tale of two halves. Designated areas saw a 144 per cent rise in the probability of landing a solar project, but the associated jobs had little employment effects, and the projects didn’t shift the political mood. However, loyal readers may remember we previously highlighted a paper that found something different: a rise in the Democratic presidential vote share. So, maybe the Democratic-launched plants are a vote winner depending on the visibility of their local impact. Relatedly, and closer to home, we published research yesterday after a six-month deep dive into Aberdeen and Grimsby. In both places, there is significant potential from net zero, but the UK Government needs to take a more deliberate, hands-on approach to concentrate green industries and opportunities both to help cushion oil’s decline in Aberdeen and form an anchor for the local economy in Grimsby. Spoiler – there’s more to this than simply drilling at Jackdaw.

Mr Worldwide? A new survey of more than 100,000 men across 70 countries has tried to assess the state of masculinity around the world, and found it’s surprisingly stable across countries. Now, there’s nothing we love more than evidence-based macho stereotypes, so strap in! Men who score more on stereotypically manly metrics – like competitiveness, violence and, my favourite, (not) asking for help – are five percentage points more likely to be in work, earn nearly three per cent more, and lean into stereotypically male industries. However, they’re also more likely take risks, report higher depression and are less likely to seek help for mental health (come on, guys!!!). The authors find that exposure to war or economic recession during one’s formative years increases these feelings of masculinity. So, from the world of work to the GP surgery to the ballot box, masculinity is quietly calling the shots.

Kids these days. A decade ago, one in 100 young people (16-21-year-olds) said that they have no close friends, this is up to 1 in 20 today. 14-15-year-olds are also less happy with their friendships than ten years back. What could be underneath a decline in quantity and quality? The authors put it down to falling self-confidence: only a third of young women are happy with how they look, and life satisfaction among young men has dropped from over half to just under two-fifths since the 2010s. The share of young people who report accomplishing less due to mental health has also quadrupled over this period. This fits very closely with our recent work which shows that young people are alone up to two hours more than a decade ago, in part due to lone working. These trends are yet another reminder of the many challenges young people face.


Chart of the week

It’s been another week of fretting about the “the markets” as the benchmark cost of government borrowing reached its highest level in nearly 30 years. But sometimes it pays to step WAY back, so below we show how borrowing costs have changed since then-Chancellor Henry Pelham (I’m sure you all remember him) first consolidated outstanding government debt into one handy bond in 1751. Admittedly we’re worrying less about the War of Austrian Succession these days, but we do have our own problems. Borrowing costs reached 5.2 per cent this week, the highest since 2007. The media focus on day-to-day moves was, yet again, unwarranted (Tuesday saw a change of around 0.07 percentage points, which is not unusual), and the rise arguably reflected concerns that central banks will put up rates to counter rising oil prices, rather than anything the Prime Minister said. The chart shows that rates at current levels are actually very close to historical averages, and it’s the very low borrowing costs seen through the 2010s and early 2020s that are unusual. Some think we’ll get those ultra-low rates back; that would be nice. But it would be foolish to bank on this, so Mr Healey should think about how taxes and spending need to change in a world where we are paying debt interest of over £130 billion a year, over a £100 billion more than we did during the pandemic.