Assessing the performance of government

The previous chapters assessed what happened to each of the selected public services between 2009/10 and 2016/17. This chapter draws the data together to provide an overall assessment of performance.

We begin by reviewing spending, and assessing how well governments since 2009 have maintained control of it. But knowing how much spending on services has gone up and down only tells half the story. To assess performance, we need to assess how well government has used that money. This chapter therefore pulls together information on three other factors:

  • Demand: We need to know if there is a potential ‘spending gap’ emerging, due to spending and the demand for services changing at different rates. An increase in spending is effectively a cut if it is accompanied by an even larger growth in demand.
  • Efficiency: A potential spending gap does not have to mean a decline in scope or quality, if the Government can find ways to achieve more with each pound of the public’s money that is spent.
  • Scope and quality: If the Government fails to make the efficiency changes needed to bridge that gap, standards will decline, or people will be forced to wait for – or do without – services, often in unplanned ways.

In light of how public services have performed, the chapter ends by considering some lessons that can be learned from the implementation of both the 2010 and 2015 Spending Reviews.

Since 2009/10, the Government has largely controlled spending, but since 2016 emergency cash injections have become common.

Austerity provides the overall backdrop for all the stories we tell in this report, but that does not mean that all the services we have looked at had their funding slashed. We estimate that across all nine, the total change in spending was in fact a 3.6% real increase between 2010/11 and 2015/16.* This ranged from a 22% real-terms reduction in prisons to an increase of 15% for hospitals.

In some cases, there was also a shift from general taxpayers to specific users meeting the cost of services. This is clearest in the case of visas and immigration, where some charges for services exceed the direct cost of providing those services.** In courts, increased charges for civil and family cases, but not criminal ones, have helped to sustain spending. While formally there should be no cross-subsidy, it is likely that some of these charges have supported the common infrastructure of the courts system.

For the vast majority of the period, none of the departments overseeing these services overspent its planned day-to-day spending budgets.*** Spending levels were in line with those set at Spending Reviews. There were exceptions – MoJ received extra money from the reserve, following the cancellation of planned sentencing reforms in 2011.

However, since 2015/16 the Government’s ability to control spending in line with its initial plans has weakened significantly. In total, over £10bn of extra resources are being diverted to keep the various services going over a five-year period. Much of this has been funded by squeezing other parts of departments’ budgets. These reversals are strongly related to what has been happening to demand, efficiency and quality of services, which we now consider.

Large potential spending gaps arose in most of the services, and were particularly pronounced in policing, prisons and adult social care.

Alongside the change in spending, there were major shifts in the demand for services between 2009/10 and 2016/17. In adult social care, police and prisons, around a 20% ‘gap’ emerged between actual spending and the counterfactual level of spending that would have been required in 2016/17 if spending had risen in line with demand.

  • There was a rise in demand for adult social care, with growing numbers of over-65s (up by 18% between 2009 and 2016) and of working-age adults with long-term needs. At the same time, spending fell by 6% in real terms.
  • In prisons, demand remained relatively constant, with a very slight increase in the number of prisoners over the time period. But spending fell by over 20% in real terms.
  • Demand in policing is harder to measure, but the overall population increase, around 5% during the period, could be said to have put pressure on this ‘public good’ service. This compares to a 17% real-terms fall in spending.

There were much smaller changes in the difference between spending and demand in courts, hospitals and GPs.

  • For courts, while there has been some drop in the number of cases, spending fell by more. The biggest drops were in the more expensive county court cases, which fell by 25%, but most cases are dealt with in magistrates’ courts where the fall in numbers was less. The data need to be treated with care, as it is difficult to establish exactly what was spent on criminal trials, but overall the data suggest that a potential spending gap of up to 10% opened up over this period.
  • Demand in hospitals has been rising faster than the increases in spending: funding rose by 6% in real terms up to 2013/14, with activity in hospitals rising 11% over the same period. But the consistent missing of key waiting times targets since 2013, along with rising GP referrals and A&E admissions, suggests that relentless increases in demand have outstripped the service’s ability to deliver, despite the continued increases in spending.
  • For GPs, funding was flat in the early period, but started to rise more steeply after 2014–15. However, this is in the context of a growing population, especially among older age groups. Over the period, the data suggest that the same potential spending gap opened up for GPs as it did for hospitals. But this needs to be treated with care, as there are only survey-based estimations for the number of GP consultations.

Finally in two services, schools and UK visas and immigration, no significant potential spending gap opened up.

  • While schools faced an increase in pupil numbers, they fared relatively well compared with other services. Pupil numbers (up 6% overall) rose by around the same amount as spending (also up 6% in real terms).
  • Finally for visa and immigration services, spending kept pace with changing numbers as they are funded largely through user charges. After the EU referendum, there was a sharp rise in permanent residency applications, which were accommodated by the system.

Substantial efficiency improvements bridged the gaps in policing, in prisons up to 2013 and in visas after 2016, with smaller improvements elsewhere.

A potential spending gap opening up does not have to mean a decline in scope or quality, if government can find ways to raise the efficiency of public services – to do more with every pound spent. We can look at two aspects of this: making economies and raising productivity.

Making economies. This is essentially about buying things cheaper. It involves reducing the amount paid for the people, goods and services used to produce the service. If prices or wages are lowered, then the same service can be produced for less money.****

Most of the money public services spend goes on wages: pay bills typically account for around 70% of the cost of providing public services. In all of our services where specific wage data are available, median wages have grown at a slower rate than for the economy as a whole, and have fallen in real terms. For example, data from the Annual Survey of Hours and Earnings show that median wages for police officers grew by 7% in cash terms, and median wages for prison officers grew by only 2.5% between 2009 and 2016. This compares to an increase of 11% across all jobs in the public and private sectors. In other areas, the prices paid for goods and services have also fallen. For example, local authorities have reduced the amount paid for privately provided care places by around 6% since 2011.

Raising productivity. This is essentially about doing things better. It involves increasing the amount of a service that is produced by a given number of people or assets (say, prisons or schools). There are lots of ways to raise productivity, from simply using staff time better to developing a completely new technology to deliver the service.

We can get a sense of how much productivity has been improved by comparing the actual cost of providing a service with the cost implied if productivity improvements had not taken place. Given changes in demand, how much would services have cost to provide if they had continued to convert inputs to outputs at the same rate?† This exercise – giving us a counterfactual cost to compare with the actual cost – suggests there could have been substantial increases in productivity across services.

This is most striking in the areas that faced the biggest reductions in expenditure, with prisons holding roughly the same number of prisoners for over 20% less than the counterfactual cost, and policing achieving the same level of public satisfaction for around 10% less. Improvements in courts appear to be on roughly the same scale as those in the police. The numbers suggest that more modest improvements in productivity were achieved in hospitals and among GPs, and that increases in efficiency in schools were reversed as spending continued to increase in later years.

There was little evidence of improvements in productivity in adult social care or UK visas and immigration.††

Efficiency or poorer quality? Through both economies and productivity increases, the Government potentially managed to do more for less in many areas. But to establish whether there were actual efficiencies, we need to assess one other factor – quality. A service that is making efficiency gains will do the same (or more for less) to the same standard. Quality is, by its nature, hard to measure, but in the preceding chapters we found a range of proxy measures to capture what might be happening.

Again the patterns vary, with the most striking divergence in the criminal justice system.

  • The police faced the challenge of 17% spending reductions, which was managed without a comparable drop in service quality (although victim satisfaction is beginning to decline). This suggests that there were indeed substantial improvements in efficiency in policing across the period.
  • While the data need to be treated with caution, the same pattern was clear in the courts, where the available indicators did not show any large drop in quality.

For prisons, initially our proxies for quality held up, again suggesting substantial improvements in efficiency. But from 2014 onwards, there was a clear and rapid deterioration in the indicators relating to violence and safety. There was also a fall-off in the provision of programmes that may be related to rehabilitation, for example, there was a drop in the number of prisoners gaining formal level 1 and 2 qualifications in maths and English. It is clear that the apparent improvements in efficiency within prisons after 2014 were in fact largely a deterioration in quality.

In hospitals and schools, it appears that the modest improvements in efficiency were real.†††

In hospitals, for those actually receiving treatment, the data suggest that quality was holding up. For example, the number of patients contracting bacterial infections (e.g. MRSA and Clostridium difficile) and developing pressure ulcers decreased year on year. Similarly, satisfaction levels among service users hardly changed.

  • The same is true for schools, which managed to keep teacher-pupil ratios and academic standards consistent.

The available data make it difficult to judge quality in GP consultations, as there is no information about clinical outcomes.

So where did improvements in efficiency manage to bridge a potential spending gap? It appears that improvements in efficiency bridged the large gap in policing and the courts throughout the period, and the similarly large gap in prisons in the earlier years. In schools and UK visas and immigration, there was no sustained potential spending gap to be closed.

Finally we need to consider local neighbourhood services. The data are too weak to allow any assessment of the size of a potential spending gap, or of detailed changes in efficiency. The headline facts though – large falls in spending, the employment of a range of strategies to adapt to this, and overall satisfaction with services as broadly holding up – suggest that there have been substantial improvements in efficiency.

Signs of pressures on council reserves, however, may suggest that a turning point is being reached.

Reductions in scope and quality were used to bridge the gap in adult social care throughout, and from around 2014 onwards in hospitals and prisons.

In other areas, however, efficiencies were not enough to bridge the potential spending gaps which opened up during this period. Here it is clear that the scope and quality of services took the strain.

In adult social care, the pressures have largely manifested themselves in formal restrictions on people’s ability to access services. Between 2010/11 and 2013/14 the proportion of councils paying for services for people with low or moderate needs fell to just 13%. There was a 25% reduction in the numbers receiving support – and since then, there appear to have been further reductions. There have also been knock-on effects from these restrictions. They are increasingly leading to delayed transfers from the NHS as people end up staying in hospital longer than necessary, waiting for care packages in their home or the community. This is damaging for individuals and places further pressure on the hospital sector.

Hospitals, unlike adult social care, in most cases cannot directly restrict access to services by raising the eligibility threshold, given that the NHS is ‘free at the point of delivery’ for everyone.†††† Rather than formally restricting access to services, they are simply requiring people to wait longer for them, while running up deficits as activity outstrips spending. The A&E four-hour target for type 1 admission, transfer or discharge has not been met on a quarterly basis since December 2012; the standard for treating cancer patients within 62 days of an urgent GP referral was breached for the first time in March 2014 and has been declining since; and the proportion of patients waiting more than 18 weeks to begin treatment for non-urgent conditions has been consistently below the target since March 2016 – the worst performance since the target was introduced in April 2012.

For prisons, we have already seen that a deterioration in quality, specifically around safety and security, was clear post-2014, setting the service on an unsustainable course. The reductions in staffing in this period were not a sign of efficiencies being achieved, but a change that would eventually have to be reversed. The recent commitment to increase staff numbers has not yet had any effect.

The 2010 Spending Review was initially successful, but that success has not been repeated. The current Government is struggling to implement the spending limits set in Spending Review 2015.

The analysis laid out above tells a tale of two spending reviews: one successful (at least initially), and one less so. It shows that the tactics deployed in the initial period of spending control after 2010 began to run out of steam during the last parliament, but that neither the last government nor this one has successfully managed to change course. As pressures on the scale and quality of services have grown, so the Government has found it harder to maintain spending control. The current Government is showing signs of being caught in a reactive spending cycle, responding too late to long-evident problems.

The 2010 Spending Review: short-term success

Spending Review 2010’s hard budgets were successful in the short term. Spending was controlled, and efficiencies were achieved, for the first three to four years of the spending review period – and continue to be achieved in the police and courts. Three key factors explain this:

  • The Government set hard budgets and stuck to them. Departments were clear about their resource limits and understood that the Government’s commitment to deficit reduction would not allow for any increase.
  • The high levels of spending growth – by historical standards – in the 2000s meant that at least initially there were some easy savings to make, including reductions in administrative costs.

Relatively weak growth of earnings in the rest of the economy may have made it easier to hold down input costs.

But this did not stimulate the transformation needed in the longer term. These measures may have been sufficient if 2014 had marked the end of austerity, as was originally planned. But while they allowed services to spend less money, they did not make the services fundamentally cheaper to run.

Not enough action was taken during this period to advance those politically and organisationally challenging ‘transformational’ reform agendas – for example, reducing demand, making better use of IT, or integrating services. Changes to people’s behaviour and the way the NHS works have not succeeded in controlling hospital admissions.[1] The MoJ’s plans to introduce sentence ‘discounts’ for early guilty pleas - and save £130m in the process - were dropped in 2011.[2] Plans to integrate health and social care following the 2012 Lansley reforms have had little national impact, with delays in hospital discharges increasing dramatically.

Looking back from 2017, with further spending reductions tabled and economic uncertainty ahead, the early years of austerity look like a missed opportunity to successfully implement radical reforms.

The strains caused by services simply tightening their belts began to show in the run-up to the 2015 Spending Review, as it became clear that the economy was performing worse than expected. Government was increasingly relying on falling quality, explicit rationing and increased queueing to take the strain of managing spending constraint, in prisons, adult social care and hospitals respectively.

The 2015 Spending Review: crisis, cash, repeat?

In 2015, the new Conservative Government (but the same Chancellor, George Osborne) faced a very different challenge compared with 2010. The easy efficiency gains – holding down staff numbers and costs – had largely already been made. Some of the services which had enjoyed historically high levels of spending growth in the years preceding 2010 were now facing large pressures, which were clearly evident in the data. In a few areas, the Government changed course. So for GPs, the tight spending control post-2010 was reversed, with new commitments to spend more and to recruit more GPs.

But many 2015 Spending Review settlements were not driven by the data. In November 2015, more than £10bn was earmarked to be removed from departmental budgets in real terms by 2019/20. Yet there is little sign that the settlements handed to departments were fundamentally driven by an assessment of how services had fared after 2010.

  • The Ministry of Justice received a 15% real-terms cut, with no protection for prisons despite spiralling violence.
  • The funding settlement for local government was complex, but the net result was no real increase in funding for adult social care, despite continued growth in demand and accelerating problems in discharging people from hospitals.
  • The NHS received extra real-terms funding, front-loaded into 2016/17, but much of this was absorbed in meeting the hospital deficits being run in 2015/16.

In schools and the police, there were fewer immediate pressures evident in the data in the run-up to the 2015 Spending Review: pupil attainment was holding up, as was public confidence in the police. Yet in both these areas spending was protected, with police and schools budgets flat in real terms (i.e. broadly similar to adult social care, which faced severe pressures).

Since the Review, the pressures have only intensified. The Government has failed to develop alternative strategies – new ways to manage demand or make services more efficient – and has continued to pursue belt-tightening approaches that in many cases have already run their course.

Since 2015/16 the Government’s ability to control spending in line with its initial plans has weakened significantly.

  • In 2015/16 hospitals were running record deficits of over £2bn, though the Department of Health managed to remain within budget. In 2016/17, despite the injection of an additional £1.8bn from the Sustainability and Transformation Fund, hospitals were still running a deficit of around £800m.
  • In November 2016, the Government provided an extra £500m over three years of resources for prisons.
  • In December 2016, the Government provided around £400m of extra resources for adult social care in 2017/18, partly funded by changing the timing of council tax increases.
  • In March 2017, the Government provided an extra £2bn over three years for adult social care, on top of the December 2016 announcement.
  • In July 2017, the Government provided an extra £1.3bn over two years for schools.
  • In September 2017, the Government abandoned its policy of allowing only 1% pay increases in the public sector, announcing pay increases of 2% for police offices and 1.7% for prison officers respectively.

The Government therefore currently has two sets of issues to tackle. The first is the actual pressures which are building up – or have built up – in different services. But the second concerns flaws in the process for making spending decisions. The absence of a consistent planning and performance framework to support spending decisions has left government vulnerable to making over-optimistic assumptions and missing key warning signs.

The next chapter outlines our recommendations for dealing with both sets of issues.



*     The total amount of spending covered is £164.1bn in nominal terms in 2015/16. This is 60% of total resource Departmental Expenditure Limit (DEL), after excluding DEL transferred to the devolved governments.

**    Where charges exceed the cost of provision, this is essentially a tax levied on service users who are, by definition, non-nationals.

***   Note that the DfE has in recent years overspent on its capital budget, in relation to a privately financed school building programme.

****  Wages seldom fall in nominal terms, but services can effectively become cheaper in real terms if wage increases are kept below the rate of inflation.

† In a counterfactual the values of all other elements – specifically activity, prices and wages – are at the actual level.

†† Office for National Statistics, ‘Econonomic Output and Productivity’, retrieved 5 October 2017,

††† This is broadly in line with the ONS’s estimates. These show productivity rising in health (a wider measure than just hospitals) up to 2014. In education (again a wider measure than just schools), productivity rose up to 2012, but these gains were reversed in the period 2013–14. There are no estimates available for 2015 onwards.

†††† Robertson R (2016) ‘NHS rationing under the radar’, blog, The King's Fund, 17 August, retrieved 5 October 2017, https://www.