Public spending is set to come in at around £800 billion this year. The biggest single element, about £220 billion, is spending on transfers — pensions and other welfare benefits. Not far behind is the £180 billion or so we spend on employing more than five million public sector workers: the nurses and doctors, teachers, soldiers, police and administrators who deliver the public services we rely on.

How we pay these public servants matters. Pay them too little and we will struggle to recruit and retain the right staff and deliver high-quality services. Pay them too much and not only will we waste an awful lot of taxpayer money but we will also create difficulties for the private sector looking to employ the best and most talented people it can find.

Lacking the obvious market mechanisms that help to determine pay in the private sector, governments have struggled for decades to find the best way of setting pay for their employees. Politics meets economics meets spending control meets collective bargaining. With nurses planning a summer of action, after recent pay disputes involving teachers and doctors, we seem set on a collision course once again.

That’s because public sector pay has been held down for a long time now. Pay scales were largely frozen in 2011 and 2012 and have risen by only 1 per cent each year since then. More years of 1 per cent increases are in the pipeline.

Had pay in the private sector been growing at anything like a normal pace this level of pay restraint probably would have been impossible. But it hasn’t. Until recently, pay misery has been pretty much equally shared. Those in the private sector have, on average, done no better than those in the public sector. As soon as the recession and financial crisis hit in 2008, private sector pay took a dive. It recovered a bit before the recent Brexit-vote induced spike in inflation, but in real terms it remains below its pre-crisis level.

Wages in the public sector have ended up in a rather similar place, but they took a different route. They carried on growing through the crisis itself, hitting their highest level relative to the private sector in many years in 2011. What the subsequent pay restraint did was reel them back in towards pre-crisis relativities. On average they probably got back to their pre-crisis par within the past year or so. That is one of the reasons why, until recently at least, the government found it relatively easy to impose what looks like pretty draconian pay restraint without sparking either more widespread industrial action or a worse crisis in recruitment and retention.

Things are likely to get more difficult from now on. Pay in the private sector is not going great guns, but it is rising by more than 1 per cent a year. If the government does stick to its plans to hold public pay growth to an annual 1 per cent over the next three years, then public sector wages will start falling well behind those in the private sector, soon reaching historically low levels. At some point that is likely to result in more widespread recruitment problems — to accompany those problems we are already creating by making working here much less attractive to the foreign recruits on whom many public services depend.

At some point pay will have to rise. But government really ought to think hard about more than merely the overall pay limits. How pay is set, how it varies across the sector and what role other elements of reward, notably pensions, play in the remuneration package all need a careful look.

In terms of setting pay, drawing the sting from some of the politics and industrial relations problems is important. That’s why, in the decades before the recent central pay policy, independent pay review bodies were responsible for recommending pay levels for about half of public sector workers, including teachers, nurses, doctors and the armed forces. There is a good case for returning to that system as soon as possible.

We also need to look at different bits of the public sector. Some do better relative to their private sector peers than others. In broad brush terms, the highest-educated and highest-paid do better in the private sector, the lower-skilled are better off in the public sector. There is less inequality in most dimensions. Some of this is good: women do relatively better in the public sector. Some of it probably less so: the public sector often underpays in London and the southeast and overpays in poorer regions. It’s not merely averages that matter and we need to think about whether these differences are appropriate.

The biggest public-private difference, though, is in pension provision. Relatively generous, guaranteed, salary-related pensions, of a kind that has almost entirely disappeared from the private sector, are still ubiquitous in the public sector. There is a case for this. Government can provide such pensions while most businesses find doing so is too expensive and too risky. But we need at least to ensure that we are getting good value from these costly promises and that their cost and value are fully appreciated by those who are accruing them. Whether the scale of difference between the sectors makes sense is questionable.

We need to get this, and much else about the deal we offer to our public servants, right. A new era of higher pay increases should come soon, but with reforms that ensure three sorts of fairness: a fair deal for public sector workers; fairness between them and their private sector counterparts; and a fair deal to the taxpayers who foot the bill.

This piece was first published by The Times and is reproduced here in full with permission. Paul Johnson is director of the Institute for Fiscal Studies. Follow him on @PJTheEconomist.