Everyone agrees that the economy seems to be recovering. So what?

I ask because a recovery does not mean that everyone gets better off. Just as millions of people keep their jobs in recessions, so thousands lose theirs even in good times. My chart shows the point. It shows the quarterly job separation rate. I define this as the numbers moving from employment to either unemployment or inactivity in a three month period, expressed as a percentage of the previous quarter's stock of employment. Even during the good times of 2002-07, this rate averaged over 3% per quarter. This implies that the average worker in decent times has a greater than 3% chance of not being in work three months later. The risk is higher for younger or less educated workers.

Now, this is an imperfect measure of job insecurity. On the one hand, it overstates it because it includes people leaving jobs voluntarily, for example to retire or have children. But on the other, it understates it to the extent that it ignores people who lose their job and get a new one within the same quarter.

Another measure of job insecurity is the redundancy rate. This averaged 0.6% per quarter in 2002-07, but rose to over 1% in the recession. This, though, understates job insecurity, to the extent that workers can be eased out of jobs without being formally made redundant.

What this means is that, even though the recession is over, many workers still face a significant risk of unemployment. And this, remember, means not just a huge loss of income, but also considerable unhappiness.

Two other facts corroborate this - that between 1997 and 2008 an average of 47,000 jobs (pdf) were lost per week; and that official data show that between 2002 and 2007 an average of around 10% of businesses ceased trading each year.

I say all this as a counterweight to the statistical fetishism which obsesses with noisy quarterly GDP data but ignores the insecurity which many people face even in normal times. Economics, it is often forgotten, is about people, not numbers.

In this sense, the economic debate misses something important. The question is not just: how can we achieve macroeconomic stability? It is: are we doing enough to cushion individuals from microeconomic volatilty and idiosyncratic risk?

I'm not at all sure we are.