Sure, this wouldn't be the intention of such a move. Quite the opposite. The case for such a target is that, with unemployment likely to stay high for a long time, it would signal that interest rates will stay low for a long time. This would give us greater certainty about monetary policy, which could give extra confidence to businesses considering whether to invest.
I'm not sure even this is convincing. In increasing borrowers' confidence that rates will stay low, such a policy would also increase savers' fears they will do so. This could prompt them to cut spending in anticipation of lower returns.
But let's leave this aside. A further problem is that extra certainty about interest rates comes at the price of less certainty about future inflation.
In theory, there's a simple reason for this. The conventional Phillips curve tells us that lower unemployment raises subsequent inflation. Targeting unemployment, therefore, might well entail higher inflation. This would create uncertainty about the question: when will inflation return to target?
In practice, though, the uncertainty has another dimension. The link between unemployment and subsequent inflation has been unstable. Between 1997 and 2007, it was weakly negative, as the orthodox Phillips curve predicts. Since then, however, it has been positive; both rose after 2007. This instability raises a problem. Let's say unemployment were to rise. Would this be a sign of more spare capacity in the economy and hence that inflation will fall? Or would it instead tell us we've suffered an adverse supply shock in which case we might see both rising unemployment and rising inflation?
If unemployment tells us little about the state of the economy, why target it?
In fact, unemployment is uninformative for other reasons. To see why, remember what unemployment is. The ONS says it comprises two groups: those people out of work who have found a job and are waiting to start it in the next two weeks; and those without a job who have actively sought work in the last four weeks and are available to start work in the next two weeks.
These definitions exclude lots of people who would like to work more. For example there are 1.4 million part-time workers who'd like a full-time job. There are over half a million self-employed people who work less than 15 hours a week. Sure, some of these are home-makers and the semi-retired who are happy to run micro-businesses. But others are freelancers and handymen waiting for the phone to ring, who'd like more work. And there are also 2.3 million who are economically inactive who'd like a job. The unemployment rate excludes all these.
Now, you might think the solution to this is to simply have a low target for unemployment, to capture the fact that the formal rate understates the number of under-employed workers. But this might not be sufficient. Think about that definition of unemployment again. What exactly is "actively seeking work"? A man who blindly fires off a few CVs on spec might consider himself seeking work and so counts as unemployed. One who puts out feelers amongst a wide social network might not consider himself actively seeking work and so be "inactive." But the latter could well have a higher chance of getting work than the former. The distinction between "unemployed" and "inactive" is thus a fine one. This means that changes in unemployment might not tell us much.
To see this, consider the scandalously under-reported data on flows between unemployment, employment and inactivity.
These tell us that, in Q1, 960,000 people moved into employment, with 577,000 doing so from unemployment and 383,000 from inactivity. What if employers were to recruit more people from the unemployment count but fewer from the "inactivity" count? Then unemployment would fall without the labour market strengthening.
The data also tell us that, in Q1, 941,000 moved out of employment, with 416,000 entering unemployment and 525,000 entering inactivity. What if more were to transition into inactivity - say by taking early retirement - and fewer into unemployment? Again, unemployment would fall without the economy improving.
There's a third thing. In Q1, 515,000 moved from inactivity to unemployment and 386,000 from unemployment to inactivity. What if these numbers were reversed? Again, unemployment would fall without the economy changing.
Unemployment, therefore, is as much a bureaucratic category as an economic one. It can fall without the economy improving, and rise without it worsening. Sure a big change in unemployment probably does tell us something. But precise targets for it don't make much sense.
In fact, the only reason I can think of for using them is to rely upon Murphy's corollary to Goodhart's law. Goodhart's law says: "Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes." Murphy's law says that any law, including Goodhart's, breaks down when replied upon. Murphy's corollary to Goodhart's law thus implies that a statistical irregularity might become regular if adopted as policy.
