Why the Bank of England should target wages as well as unemployment

After months of relative inaction, MPC meetings are about to get very interesting again. While rate increases should be some way off yet, improvements in a range of economic indicators mean that ever more attention is set to shift towards the question of how and when monetary policy tightens. While calls for action are likely to build, the avoidance of premature movement will be vital. With the upturn in its infancy, household budgets still squeezed and millions of families exposed to repayment difficulties associated with debts built-up in the pre-crisis years, increases in rates that predate sustained improvements in household incomes risk choking off the recovery.

Forward guidance was designed to pre-empt such discussion and make clear that rates would remain low until the recovery was firmly entrenched. Yet, with each passing piece of economic good news, so the predictions that the MPC will have to act sooner rather than later mount. More specifically, there is a growing suggestion that the unemployment rate might fall below the 7 per cent threshold significantly earlier than the Bank initially projected.

To some extent this shouldn’t matter, because members of the MPC have been at pains to reiterate that the threshold would serve not as a trigger for monetary tightening, but merely as a staging post at which they will reconvene and discuss policy options. But there is a danger that the message may not be getting through. And that does matter: if sufficient numbers of investors factor in rate increases on the assumption that the Bank will move as soon as the unemployment threshold is reached, then the economy could face an effective tightening in monetary conditions even in the absence any change in the base rate.

With this in mind, Charlie Bean’s recent suggestion that the MPC will issue further guidance as unemployment approaches 7 per cent is understandable. Some will question the credibility of changing the framework so soon after its introduction, but forward guidance has always been an exercise in communication: as the circumstances change, so should the message.

So what might forward guidance 2.0 look like? Many of the arguments for using unemployment laid out in the initial discussion document remain germane. Of course no single measure is perfect, and a focus on unemployment could prove misleading if significant numbers of workless individuals move into economic inactivity instead, but it seems unlikely that the MPC would turn its back on this measure. As well as being simple and easy to capture, changes in unemployment internalise uncertainty about what happens to productivity as the economy recovers.

If productivity grows, then firms will make better use of existing workers (through switching them into different roles and moving them from part-time to full-time for instance) and the decline in unemployment will slow. If on the other hand potential supply in the economy has been reduced as a result of the financial crisis – due to impairment of the banking sector for instance – then we might be closer to full capacity than we think, meaning that productivity will remain weak and economic growth will continue to feed through to reductions in unemployment. Either way, reaching a particular unemployment rate can serve as a proxy for the removal of a significant amount of spare capacity. The difficulty rests in identifying just where the threshold should lie.

Further guidance could therefore simply involve lowering the existing threshold – as Bean mooted. But, if what we really care about is the sequencing of action – ensuring that incomes rise before rates do – there may be something to be said for taking the opportunity to broaden the focus. The unemployment measure could be supplemented to make explicit the fuller range of factors the MPC will have regard for in making any future interest rate decision. There’s a trade-off here, between simplicity and completeness. And the more indicators the Bank includes, the more its guidance starts to resemble a genuine trigger point: if all of the lights turn green, how can the MPC possibly refuse to go? But given that this risk seems live already the danger of market mis-interpretation is unlikely to be particularly heightened by bringing in additional measures.

And it’s worth remembering that Mark Carney has always been clear that forward guidance goes beyond a simple focus on unemployment, arguing that it is designed to provide certainty that interest rates won’t go up until “jobs, incomes and spending are recovering at a sustainable pace”. The implication is that the unemployment threshold, as well as being a direct indicator of spare capacity, also helps gauge pay and, by association, income and consumption growth. But in the current context, movements in unemployment may not feed through to earnings in a predictable way. Wages have been falling in real-terms in recent years, despite improvements in the unemployment rate. And, while it’s true that we’d expect a tightening jobs market to feed through to pay packets, there remains a lot of uncertainty about the precise relationship between unemployment and wages.

Indeed, research carried out on behalf of the Resolution Foundation by leading labour market academics Paul Gregg and Steve Machin has suggested that real wages have become more sensitive to unemployment over the last decade or so, meaning that unemployment would have to fall further than in previous decades to achieve the same magnitude of real wage growth. This is just one study and the figures are likely to be indicative rather than definitive, but the changing relationship it points to adds weight to the case for supplementing the unemployment measure – by adding an explicit focus on pay in particular.

One approach would be to establish a threshold for wage growth. Ideally the focus would be on the median, as a means of capturing variations in growth across the earnings distribution. This matters because those in the bottom half are particularly exposed to weak wage projections, wider austerity cuts and the debt overhang. If the aim is to avoid doing undue damage to the recovery via increasing rates too soon, then policymakers should have especial regard for the fortunes of low to middle earners. In truth, however, such distributional data simply isn’t available on a timely basis.

Imperfect though it would be, the Bank could instead monitor the average weekly earnings measure, while remaining sensitive to potential differences across the distribution. As with unemployment, the average earnings measure has the advantage of being regularly collected and easily understood. It also helps the MPC side-step the question of just how the relationship between unemployment and pay has altered over time. Of course, choosing the right level of average wage growth would be just as fraught as the decision on the appropriate unemployment threshold, but the Bank will need to take account of such factors when considering future policy in any event. By bringing more of these difficult trade-offs into the open, the MPC would at least reduce the risk of an over-reaction to falling unemployment that doesn’t feed through into pay and incomes.

This article originally appeared on the Independent’s Econoblog