The latest instalment of our series, in which we use the Centre for Cities’ data tools to crunch some of the numbers on Britain’s cities.

For the last few weeks, I’ve been trawling Centre for Cities’ data to first demonstrate, and then explain, a worrying piece of British exceptionalism. With the exception of London, our big cities are economically much weaker than their peers in France or Germany, Italy or Spain. Birmingham, Manchester, Glasgow and Leeds all have significantly lower productivity, expressed as GVA per worker, than similarly sized cities like Hamburg, Barcelona or Marseille.

Digging through the numbers, I found that they have smaller service sectors; lower numbers of businesses; don’t produce many patents; and have a lot of workers with low skills. I have another theory about what’s gone wrong in Britain – one which, if you’ve ever been to CityMetric before, you can probably guess. But before getting into that I want to address a question a couple of people have asked about the data.

In short, it’s this: is GVA per worker a misleading measure? Since the financial crisis, unemployment in Britain – in contrast to many European nations – has stayed low. GVA per worker is a measure that might flatter cities with higher unemployment, because they have fewer workers relative to their size. Perhaps GVA per head would be a fairer measure – on one which, the implication goes, British cities a more likely to punch their weight.

When someone first threw this argument at me, I was slightly torn, if I’m honest. The measure is clearly limited in its usefulness: taken to extremes, a city with one worker and 999,999 people on the dole could in theory have an enormous GVA per worker, but that doesn’t mean it’s a healthy economy.  

And yet, the thing we’re interested in here is productivity – the rate at which value is produced. To know that, you need to have at least some idea of the amount of work being done – so the number of workers seems a more useful measure than raw population.

That, though, is angels-on-the-head-of-a-pin stuff. We can go beyond that, to actually compare how to two measures work out in practice.

Here’s a scatter graph plotting GVA against population size. You’d expect these to be roughly correlated – more people generate more wealth. But that correlation is obviously not perfect, as some places are richer than others; and generally speaking cities that are further left than you’d expect are underperforming, while those to the right are over-performing:

Click to expand.

You can see that both Milan and Naples are outliers: northern Italy is surprisingly rich, southern Italy surprisingly poor. You can also see that German cities (in black) are generally also more productive than the norm.

But from a British perspective, it’s the position of Birmingham and Manchester that’s most concerning: two big cities that look way less affluent than you’d expect. Indeed, both Hamburg and Munich are significantly smaller – yet also significantly wealthier.

The key question, though, is what does switching from GVA per head do to the rankings? Here are our 19 cities, ordered by GVA per worker:

Click to expand.

And here they are ordered by GVA per head:

Click to expand.

Switching measure does change the rankings, a little. On the per head measure, Seville looks significantly better, going from worst to best performing of the three Spanish cities: that suggests a lower employment rate than in Valencia or Barcelona. And Naples goes from a little way behind the pack to way, way behind which I think suggests a much higher one, although my head is starting to hurt.

At any rate, that suggests that judging cities on GVA per head, rather than GVA per worker, can make a difference.

But it doesn’t change the fact that the British cities are clustered near the bottom. Okay, Glasgow and especially Leeds do rather better on the per head measure than they do on the per worker one. But Birmingham and Manchester still look terrible. Britain’s secondary cities are still, as a group, performing worse than those of the other four countries.

To put it bluntly: whichever measure you use, something has clearly gone wrong in Britain’s big cities. The only question is what.

Next time, in the final instalment of this mini-series, I’m going to put forward a theory as to what that might be. Until then, why not have a play with the Centre for Cities “Competing with the Continent” database?

Jonn Elledge is the editor of CityMetric. He is on Twitter as @jonnelledge and also has a Facebook page now for some reason. 

Want more of this stuff? Follow CityMetric on Twitter or Facebook.