How to measure the grimness of a recession

GO FIGURE Different ways of seeing stats

Unemployment is up. But look at the numbers - so far it's not nearly as gloomy as the last recession, says Michael Blastland in the first of his regular columns.

How do you measure how grim a recession is? You could compare it with the boom times and discover that life is worse. Now there's a surprise.

Job Centres are getting busy

Or, at a time when we hear suggestions that this is the worst crisis for generations, we could try to find out if that's true, and compare this recession with the last recession: that is, compare bad with bad and see which is worst.

Mostly though, we hear comparisons of the bad with the best. So on the most recent, and genuinely-depressing, data we have about the state of the economy - this week's unemployment numbers - the widely reported comparison has been with unemployment in the past 10 years. That is, with the boom years. It is about the worst it has been in that period.

But what if we also compare what's happening now with what happened at roughly the same point in the last recession? These charts - of those on the dole, and the number of jobless - begin to do that, and start the month before the numbers begin to rise consistently, in the spring of 1990 and 2008.

What does this tell us, other than that most reports about unemployment are not telling us much? A few facts stand out:

• The starting level this time is well below the starting level last time
• Despite recent rises in unemployment and the claimant count, and at least two quarters of recession, we are still not up to the levels at the start of the 1990 recession;
• The lines are about the same gradient, so far - slightly steeper for unemployment, slightly shallower for the claimant count, indicating that people are losing jobs at roughly the same speed as last time.

The effect on the figures of the large number of people on incapacity benefit is worth bearing in mind here, and might suggest that the true unemployment level is higher than it looks.

But incapacity benefit was also there in the 90s recession. If we take inactivity as a whole, male inactivity rates in particular rose steeply in the 80s but have actually fallen back slightly since the early 90s. (Whereas female activity rates have been rising for years).

And here's another thought to bear in mind when raw unemployment numbers are quoted over the coming months. There are about 3 million more people of working age this time than last, and 4.5 million more than in the early 80s recession. If - and it's a big if - the raw numbers are no worse than last time, we'll have done rather better.

Crystal ball

So how bad is it this time? We'll see. We don't know yet and this is not a prediction. But comparing similar points in the recession suggest that unemployment is no worse - yet - than in the last recession, and in some ways better.

More of these are being filled out

In other ways, of course, things are not so good. Public finances are in worse shape now than then, for example. The Bank of England also predicts that the economy will contract more sharply than it did in 1990. We'll soon find out.

But as the recession takes its course, it's worth keeping up with the real data and seeing how it compares, not just bad with good, but bad with previous bad. We've been treated recently to a great deal of coverage of the "up 146,000, what a shock" variety. Is 146,000 bad? Yes, of course it is, these are people's livelihoods.

But how bad? About as bad as last time, but probably no worse, and starting from a lower level. About as bad, in fact, as it tends to be at this point in a recession. To know this, we need some relevant context, and that is in short supply.

---

• And finally, here are some numerical nuggets. You may never have wondered how many women in Japan with part-time jobs walk their dogs at 3am, but I bet you'd like to know.

"Yeah, we got that," says Jonathan Soma, who's taken a five-yearly survey of what the Japanese get up to and turned it into an interactive stream graph, which represents the changing volume of people undertaking a particular activity at any given time.

Walkies!

It's drawn from the Japanese census, which legitimises the gossipy question of what are you doing right now. "Japan slapped a bunch of people with notebooks and a sacred numbers mission: keep a log of what you do during the day, in 15-minute intervals. And those people did," says Jonathan.

The data comes from the Statistics Bureau of Japan's Survey on Time Use and Leisure Activities: Results on Activities by Time of the Day. So how many study in a private cram school at 5.29am? Zero. That's a relief.

And how many do vehicle maintenance at 10.29pm? 19,902 men. Single men? In some cases, perhaps imminently.

Is it accurate? Very roughly, would be my guess, and subject to a fair bit of variation day to day, the more so for less common activities. Strangely fascinating? You bet. October 18 is Statistics Day in Japan, and its slogan is "Statistical Surveys Owe You and You Owe Statistical Data".

• When the UK's own Statistics Authority ticked off the government just before Christmas for its knife-crime figures, less noticed was that the Home Office had previous.

The authority received an earlier complaint that at one of its briefings, a Home Office press officer had crept in with well-spun handouts praising the government's record on immigration. The complainant - no less than the President of the Royal Society of Statistics - worried that the Home Office was attempting to pass off partisan comment as independent Statistics Authority background. The Home Office has since apologised.