The perils of overconfident youth, and underconfident shoppers

So. Many. By-elections. Will no-one think of the voters. Or the pundits, who can only say seismic so many times. Yesterday’s ones weren’t even close, which is particularly surprising in Mid-Bedfordshire given Keir Starmer’s main conference promise was to concrete over decent chunks of the shires house building wise. I’m starting to think a 15-20 point poll lead might be more important to national politics than outer west London not loving ULEZ.

If anyone still thinks we’re getting an early election next spring, I’m happy to take your money. But you get TOTCs for free, including reminders this week about the joys of good management and the doom of… lower survey response rates (see COTW).

Have a great weekend all.

Torsten

PS No TOTCs next week with half-term so good luck surviving Halloween and I’ll see you in November.

Creaking China. British consumers aren’t the only ones in trouble – time to talk about China’s local government financing vehicles (LGFVs). These are the routes by which sub-national governments borrowed for investments, powering a good chunk of China’s post-financial crisis growth. The only problem? Lots of the investments are rubbish, leaving LGFVs highly leveraged. An interesting/gloomy blog from Marco Polo – a China focused think tank in Chicago – gives you the lowdown, arguing this is the next epicentre of China’s debt problems (after highly publicised property developer traumas). With $12 trillion in LGFV debt, the question is whether a managed deleveraging can be achieved. The authors think not because 1) China doesn’t have the growth tailwinds it enjoyed in recent decades from fully entering the global trading system and 2) the national government won’t kick the can down the road/bail them out (given its attempts to move away from a debt financed growth model). Their prediction? Widespread defaults, with $5 trillion of losses hitting households and taking down some regional banks. Happy days.

Magnificent managers. We know managers affect performance (just ask Spurs), but a new paper investigates how that happens. Using data from a single large company over 10 years, ‘good’ managers are identified as those who get promoted quickly and the impact of them arriving to manage a team is assessed. It’s big and lasting: good managers trigger more reallocation of workers across the firm/team, with better matching of workers to jobs boosting performance in a lasting way – 7 years on, workers earn 30 per cent more and are assessed as higher performers. The effect’s so large that the author argues doubling the share of good managers would lead to increases in output of, an almost implausibly high, 61 per cent. So it turns out we need to choose our managers, not just our parents, carefully.

Skill suitability. So good managers help match us to jobs we’re good at. You know what else helps? Self-awareness. To see why read this research that measures not only the actual communication/numeracy skills of job seekers in South Africa, but also their beliefs about their skill levels. The first conclusion is that we’re VERY bad at assessing our own strengths (i.e. knowing in which area our performance is relatively strong compared to other job seekers). Predictably, we also overestimate our own abilities. The research shows that this matters, shaping which jobs we apply for i.e. we apply for ones that require skills we think we’re (rather than are) relatively strong on. The perky authors then show that if jobseekers are given information about their actual strengths this changes what jobs they apply for and boosts both their wages (by 25 per cent) and job satisfaction (but not employment). Now this is interesting, but note the research is focused on the youth – I’m less sure sudden self-awareness will have the same benefit for the middle-aged among us. It’s a bit late for me to find out I should have joined the actual circus rather than the economic policy one.

Damaging discrimination. Labour market discrimination happens. UK evidence shows those from an ethnic minority background have to send 60 per cent more applications to get a positive response compared to white British candidates. A new study (free version) tells us more, specifically about discrimination against Muslim women. By sending off thousands of job applications in three European countries (Germany, the Netherlands and Spain), the researchers investigate whether discrimination reflects attitudes towards Muslims in general, or specifically those signifying religious observance (by including a photo of the applicant wearing a headscarf). The small sample sizes make the results less conclusive than you’d like, but they do find that veiled Muslim women are discriminated against in Germany and the Netherlands (but not Spain), particularly when applying for customer-facing jobs, and materially more than Muslims without a headscarf. This is one of many reasons it’s good we don’t normally include photos on job applications in the UK.

Chart of the Week

This was a big week for UK economic data, with the ONS treating us to 2.5 releases. Yes, 2.5 – because while we got full-fat data on prices and the public finances, critical labour force survey data on employment, unemployment and economic activity has been held back by a week (we had to make do with semi-skimmed stats on pay, vacancies and employee jobs instead). Why are you telling me this you massive nerd I hear you cry. Because it matters – the delayed data is central to big decisions like whether to raise interest rates or panic about labour market inactivity. Behind the delay is ONS anxiety about the accuracy of their flagship Labour Force Survey (LFS). COTW highlights why this should have been picked up sooner, comparing what’s happened to jobs over the past two years according to our three biggest jobs/employment data sources. While HMRC tax data (PAYE) and a survey of businesses (WFJ) indicate employment is up around two million post-pandemic, the LFS tells a very different story – with half that jobs growth. We don’t know for sure which employment path is correct, but COTW also highlights a growing concern for the LFS – that you lot aren’t answering it. The sample size has collapsed from 90,000 to barely 50,000 in just two years, down from 150,000 in the 1990s. This is bad news people, and for more than just nerds.