At a recent Civil Service World round table, officials discussed how to gather and use reliable, comparative business information to appraise services and improve departmental efficiency. Stuart Watson reports.

 Information is the essential fuel of any decision-making machine, but during a period of swingeing spending cuts it has become more crucial than ever to the business of public administration. In order to know which services to protect and which to cut, civil servants need the best possible data on whether government activities offer value for money.

Ministers, MPs, and the National Audit Office (NAO) have all noted that there’s a lack of good quality finance and performance information across Whitehall. Last year, the NAO’s report on progress toward previous efficiency targets also found that departments have not been using business information intelligently to inform their decisions on making savings.

The Cabinet Office and Treasury are working to improve the quality of financial data across government; but even with improved data, senior civil servants will still need to ensure that the information is used effectively. And as another recent NAO report pointed out, it’s important not only to gather audited and reliable data, but also to analyse and present that data in a meaningful way, setting out past and projected trends and emerging risks and opportunities.

At a recent CSW round table, held in partnership with information experts SAS, delegates began by discussing how departments could use information to improve performance and efficiency in an ideal scenario. Ranjoo Kainth, of the chief information officer’s unit at the Ministry of Defence, delivered a concise summary with which it was difficult to disagree. Departments should, she said, give “the right information to the right people at the right time and of the right quality in order to make the right decisions.”

Informing better decision-making
Stephen Mitchell, a deputy director in the Department of Health’s finance group, outlined three main ways in which government uses information: to enable Parliament and the public to hold it to account; to support the ongoing management of services; and to contribute to making correct policy choices.

Anne-Marie Vine-Lott, head of benchmarking and performance reporting at the Treasury, said that she would like to see a cultural change within Whitehall so that policy and data management teams work together to base policies around measures of cost-effectiveness.

There are some examples of this happening already, she said, in the Environment Agency and through HM Revenue & Custom’s Performance Hub. Karina Singh, a programme director at HMRC, explained that the organisation tries hard to allocate resources where the data shows they will produce the best results. Good data allows decision-makers to understand “whether a certain tax is producing enough yield for us”, she explained, and thus whether resources should be shifted to another tax to improve value for money.

Although the ideal is that policy and data teams work together, civil servants “shouldn’t be naïve,” said Guy Ker, head of publishing at the Cabinet Office. “We are working in a political process,” he noted, and “every four and a half years” general elections tend to separate policymaking from the best data on cost-effectiveness. Vine-Lott, however, argued that while the decision-makers might change, the core information available to them should stay the same.

Mitchell felt that the use of information to guide decisions has suffered a decline since the early 2000s, when the PM’s delivery unit – then led by Sir Michael Barber – had succeeded in putting together performance data in an accessible way to make an impact on decision-making. “We have an absolute plethora of information, particularly about the health service,” he said. “What is important is to be able to pull [those indicators] together analytically so you have some way of influencing decisions.”

He hoped that the new departmental business plans will provide a new framework within which information can be effectively presented to decision-makers – particularly “the apparatus that surrounds those”: the monthly updates and quarterly scorecards comparing performance across departments.

These quarterly scorecards will soon be published by the Treasury and Cabinet Office. Their aim, according to Cabinet Office chief operating officer Ian Watmore – speaking last year at a Public Accounts Committee hearing – will be to capture key performance data from departments and allow departmental boards to compare this information easily.

For these kinds of comparisons to be meaningful, delegates agreed, you need accurate and consistent information not only within but across departments. As SAS UK’s Keith Valder said: “You need one version of the truth from one consistent source; one system that sits above the others. Otherwise, you end up sending in 10 more people to check the information.”

However, all agreed that achieving that aim in government is a difficult proposition. The difficulty, said Mitchell is not a lack of data, but making a choice between key indicators when so many are available. “That scorecard has a whole apparatus of performance information underneath it which raises lots of issues about what you’re talking about when you’re looking at the performance of something as complicated as the DH,” he said. “It mixes health outcomes data, like life expectancy, with staffing data about the DH, expenditure data about the health service and the unit costs of in-patients, procurement spend by the department, and the total cost of our office estate. It’s quite difficult to encapsulate performance.”

Mitchell said it will be important to get the balance of information right in these scorecards. They could cover so much ground that the information contained within them is indigestible for the public; but if they’re oversimplified to concentrate only on finance and leave out data about outcomes, they will be of little value to service users. “It’s fairly heroic to try to construct one of these scorecards across government departments, because they are all very different,” he said.

Ker argued that despite these difficulties, there must be some common measures across departments. Making the system too complicated could allow individual departments to duck accountability, he said; better to put in place an imperfect system that can be improved over time: “An imperfect tool that delivers a 50 per cent improvement is better than no improvement at all.”

Ker also called for measures to focus on outcomes and delivery rather than inputs. Online transactions, for example, should be measured by the number completed successfully, rather than the number of people who logged in but failed to finish the task.

Improving data analysis
SAS UK’s head of public sector, Graham Kemp, suggested that measurements of process can be helpful as well as outcomes data. In procurement, for example, departments could be measured on the administrative cost of buying goods and services: “Who has the most effective procurement process for every pound that’s procured?” Establishing a common system to analyse procurement processes would also help to show up areas of fraudulent activity or error, he said, as well as “giving you the competitive angle.”

Kemp also advocated the use of analytics to generate “very fast business improvements”. Analytics software allows organisations to uncover trends and patterns in vast amounts of data in order to support decision-making, take a more proactive approach to forecasting demand, reduce risk and improve outcomes.

He asked whether analytics might be applied to the government’s digital delivery strategy, to find out why more people are not accessing services online and therefore highlight potential remedies. Ker responded that the government already knows that 8-10m people need assistance in going online. Work is taking place to analyse that group, he said – though he suggested that “we need to be more predictive about what the outcomes will be. At the moment, it’s a little bit hopeful – that we will get everyone’s grandmother online by unleashing the kids on her. There is a bit of a gap of reality, and we need a more accurate, scientific view.”

Singh complained that there is rarely time to conduct proper analysis: “Do we ever get the time to pause and think: ‘If that’s the objective, then that’s the strategy. What information do we need?’ It doesn’t happen,” she said. And Joseph added that even when an opportunity is available for analytical thinking, too often another department meanwhile presses ahead with a policy not founded on such good data.

Sharing information across government
Cross-Whitehall working is also hampered by the lack of commonly-accepted metrics across government departments, said Suzanne Joseph, head of diversity consultancy at the Home Office. “Sometimes departments working on issues with clear overlap take such different approaches to get to the same answer it becomes difficult to use the information,” she said.

The Cabinet Office’s Chris Mitchell, who works on the government’s digital delivery strategy, said that the challenge presented by more limited resources and the need to make efficiency improvements could be used as an opportunity to set clear benchmarks across government and work together to leverage improvements. “We need to work with each other and buddy-up. I am really struck by how much departments tend to work in silos in really similar areas,” he argued.

Again, delegates noted that the Cabinet Office and Treasury are working to standardise performance information across government; but Singh didn’t think that departments are ready yet to take advantage of data gathered and distributed by the centre. Departmental silos often colour people’s views of such info, she said: “We are only slicing off various aspects of it.”

“There are huge gains to be made by looking at [data] holistically,” she added, explaining that by collating the different pieces of information that departments gather about citizens and outcomes, government as a whole could be more effective. “At the risk of sounding slightly sinister and Big Brother-ish,” she said, government has much to gain from “putting the information we have together on things like taxation, benefits and health, so we have a better understanding of the customer group we are all serving.”

Joseph wondered about the likely impact of outsourcing and the localism agenda on the way that the civil service uses information. She suggested that civil servants will need to be very clear about what kind of information is required from external providers to measure performance, while Singh suggested that data protection responsibilities may impose some limitations on sharing data with non-governmental bodies.

Transparency and public accountability
Greater public transparency has been at the heart of the coalition’s approach to improving services while delivering efficiency savings, and Ker identified himself as a believer in that ideal. Consultations have been asking the public what data they want to see, he said: “If we can get information to the public in a way that they can understand, we can make decisions informed by public opinion.”

However, the panel agreed that much government data is not yet fit for public consumption. Like other departments, the DH is publicly reporting every transaction of £25,000 or more, but Stephen Mitchell asked what meaning or use that has for those without the right tools to analyse it.

Gathering information from the public using new technology was also discussed. Kemp noted that technologies are now available to gather 10 gigabytes of data per second from web pages, allowing instant analysis of public opinion about policy decisions. “There is an opportunity there, but we have to be a bit careful,” said Chris Mitchell, “about jumping in and using that info straight away. We have to see if it’s working and giving us good information about policy.”

Cultural change
Several of the panel identified a culture of resistance within the civil service, where people are reluctant to publish statistics, take responsibility for the performance level revealed, and act upon them. “There is reluctance even within departments with good information-gathering practices, because they point to poor performance levels,” claimed Ker.

Joseph suggested that the drive to cost cuts and offer services through a broader base of providers is bound to lead to change: “The way things are going, we have to be thinking about getting value for money. Quantitative measures will have to come in and [the public] will be able to choose how they access their services much more than they did before,” she said.

Civil servants should take responsibility for developing their information management skills, argued Kainth, noting that the MoD is offering training courses on data skills for all of its civil servants.

However much information is gathered, though, there will always be tough decisions to be taken, said Stephen Mitchell. He pointed to the example of the Dilnot report on social care financing, which at the time of the discussion had not yet been published. “I’m sure that will be a deeply analytical report, given Andrew Dilnot’s credentials, and a lot of information has been collected,” he said, “but some very difficult decisions will still have to be made on the basis of that information. You still have to have the political wherewithal and the will to do something about it.” ?

Around the table
Peter Jones, deputy director, IDEAS data reporting programme, Strategic Planning, Finance and Performance Department, Welsh Government
Suzanne Joseph, head of diversity consultancy,
Home Office
Ranjoo Kainth, skills and training team, Chief Information Office, Ministry of Defence
Graham Kemp, head of public sector, SAS UK
Guy Ker, director of publishing, Cabinet Office
Robin Martin, research team, Institute for Government
Stephen Mitchell, deputy director for governance, group finance, Department of Health
Chris Mitchell, head of quality monitoring and accreditation, digital government services, Cabinet Office
Karina Singh, programme director, HM Revenue and Customs
Keith Valder, chief operating officer, SAS UK
Anne-Marie Vine-Lott, head of benchmarking and performance reporting, HM Treasury

Written by Stuart Watson, CSW