Mutant algorithims, dodgy hair salons and bad debts

Top of the Charts

Afternoon all,

Schools’ back. Well actually, ours isn’t – which I am totally, 100 per cent, relaxed about… But everyone other one seems to be, which is good news. I’ve always been in the “we should probably educate the kids” camp. It’s a rather basic civic duty. Like voting, which apparently Donald Trump is now so keen on he thinks people should do it twice. That’s more crime, than civic duty, but it’s the thought that counts.

Our civic duty is obviously to provide some weekend reads, and this week we’ve got everything from colluding algorithms to singing central banks. Controversially, unlike school attendance, TOTCs isn’t compulsory. Yet. While we await the legislation to sort this anomaly, feel free to encourage a friend/acquaintance/stranger to sign up.

Have a great weekend,

Chief Executive
Resolution Foundation

Mutant algorithms. Since we’re all algorithm focused at present, take a look at interesting research from Bank of Canada economists examining the impact of humans handing over price setting in German petrol stations to AI. The good news for algorithm fans: gas stations using them made profit margins around 9 per cent higher than those still letting humans set prices. The less good news: this seems to be mainly because algorithms reduced price competition (algorithms made no difference to profit levels for stations that were already local monopolies). Crucially the authors say this is evidence that algorithms can learn to collude with each other to set higher prices – in local areas with two petrol stations that both introduced algorithms profits rose a massive 28 per cent. Human decisions lay behind our exam fiasco – but we do need to be on the look out for the impact of real mutant algorithms.

Monopolising workers. Last week the prestigious Alfred Marshall Lecture at the European Economics Association annual conference was given by the UK’s own Alan Manning of LSE. It’s worth watching the whole thing. He argues that while policy makers spend lots of time worrying about the damage monopolies can do to consumers (fewer options generally = higher prices/less choice), we don’t do enough worrying about ‘monopsony’ (where lack of competition between employers/options for employees to change jobs is the problem). Alan’s best UK illustration is a High Court judgement which accepted a hairdressing salon barring their hairdressers from working for other salons within half a mile of their premises. Given that half of hairdressers’ commute for less than 10 minutes, a half mile exclusion zone might mean not a lot of other salons to work at – which isn’t a great position to bargain for higher wages from. For a shorter version of his argument, check out Alan’s blog.

IOUs. Lots of us owe the Chancellor a debt of gratitude: we scoffed 100 million subsidised meals over August. Shockingly helping out by eating out with a 50% discount is an easy sell on the patriotic duty front. Much less sweat involved than “Dig for Victory”. Luckily the Chancellor isn’t asking for that cash back (at least not yet). But more traditional debts do come in the form of the £50bn worth of loans to businesses that the government has guaranteed during this crisis. How much of this gets repaid matters for the public finances (the OBR estimate up to £33bn might be written off), but the strength of the economic recovery also depends on how much firms struggle to repay debts. ECB research examines 88 post-1990 banking crises to see how big a problem loans close to default became. Their conclusion: around a fifth end up as non-performing loans, and this slows the recovery if not resolved.).  The good news is this doesn’t look like becoming a banking crisis, but how to unravel lots of IOUs to the Chancellor is going to be a big feature of the years ahead.

Bank research. Some good public service announcements from the Bank of England this week. First a Bank Underground blog, bringing together lots of the real time data work that has been done during this crisis. The Bank also published their “Agenda for Research” – setting out their five priority areas for work over the coming years. Now stop yawning, this matters. The Bank has the UK’s biggest concentration of economists working on public policy questions (the Treasury is tiny in comparison – something that should be rebalanced). Things to note include an explicit focus on the “level and nature of the monetary policy target”, which we very much agree should be a priority. The US Federal Reserve made big news at its Jackson Hole Conference last week announcing a shift to average inflation targeting (useful write-up). This is welcome given one of the big lessons of the last decade is you need to let your economy actually get some wage pressure building before worrying about raising interest rates (as we set out last Autumn). But there’s lots more research for the Bank to do given the reality that while our inflation target remains 2 per cent, interest rates will be spending a lot of time stuck near zero.

Bank songs. Forget the Fed or the Bank of England, the best central bank news of the past week came from the Bank of Jamaica who are taking monetary policy communications to a whole new level… Who knew “low and stable inflation” could inspire such passion.

Chart of the Week

If you’re after a nostalgic trip back to Britain’s pre-covid economy, then look no further than the current unemployment rate, which has been hovering merrily at close to an historic low of 4 per cent throughout the pandemic. Should this make us relaxed? No because, as Michael Saunders of the MPC notes in a speech today, all this really means is that the headline data is not capturing our labour market crisis, with widespread furloughing muddying the picture. As Chart of the Week shows, if you look at other leading indicators such as Google searches for redundancy, then the picture is more worrying. People aren’t typing that into a search engine for light relief. How big exactly the coming rise in unemployment will be is very uncertain. The Bank is forecasting a near doubling to 7.5 per cent by year end. Personally I think it may take longer to build to that level, but even if it does this would be by far the sharpest rise in unemployment for half a century. So, while all the political debate in Westminster this week is about whether or not to put taxes up, the real exam question for policy makers this Autumn is how to keep the rise in unemployment down.