Government data on spending is woefully inadequate, according to Francis Maude. At a recent round table, civil servants discussed how to get better data – and what to do in the meantime. Suzannah Brecknell reports.

 Last week Cabinet Office minister Francis Maude announced that a new body will carry out centralised buying on behalf of the government as a whole. This did not come as a surprise to anyone who has been following the progress of the coalition’s procurement reforms, but it did underscore the priority which is being given to re-shaping the way in which departments buy goods and services.

Information – or a lack thereof – has played an important part in that agenda. Earlier this year, Maude told the public administration select committee that the quality of government data when he began to renegotiate contracts with key suppliers last year was “woefully inadequate”. 

With good information on what they are buying, procurement teams can work to consolidate and reduce spend; track trends in spending and compare prices to find better deals; identify fraud and payment errors more easily; spot when departments are paying twice for similar services; and help departments improve their resource allocation and accountability by giving a clearer picture of the cost of given services and goods.

Yet if information is power, civil servants seem to have been rather powerless in this area. A survey carried out by Civil Service World and recovery audit firm PRGX earlier this year found that over 80 per cent of senior civil servants said they find it difficult to access data on public spending in particular categories or with specific suppliers.

At a recent CSW round table, held in partnership with PRGX, attendees began by discussing why the quality of government spending data is so poor. It is not, suggested PRGX client services director Jon Francis, for want of investment. The survey revealed a multiplicity of different systems in use to monitor spending across government, many of which are top-end systems, he said. But the plethora of systems can also cause difficulty in collating data across departments or using historical data within departments to track changes in spending over time.

Iris Montecalvo, head of finance for the NOMIS project within the Ministry of Justice (MoJ), pointed out that machinery of government changes also make things complicated: as her team has moved from the Home Office to the MoJ and, most recently, to the Prison Service, data has been lost and systems changed.

Does this matter? The fact that government runs a variety of IT systems doesn’t need to be a limiting problem, said Dave Harrower, of the performance and management information team in the Cabinet Office’s Efficiency and Reform Group. What really matters, he said, is agreeing to use standardised definitions and categories within these systems so that “you can interpret the data in a consistent matter”.

Building a common set of categories also allows one to collect data in what Harrower described as “the only way you can get decent information” – that is, “collect it at the lowest level of detail and build it up”. If information is collected at a higher level, it can lack the detail which would allow meaningful analysis of spend, and Montecalvo noted with approval that information is often held at a much lower level in local government.

A pragmatic solution

The approach taken by the ERG as it has sought to gather data on which to base procurement strategies for common spending categories has been to invest in software that can connect the data gathered from existing systems, and apply a common taxonomy of categories, rather than introducing within departments an entirely new system to collect all transactional data.
 
Francis said he is encouraged to hear that ERG is taking this realistic approach “rather than some grand system which is going to cost billions and fall over: they’re aiming for something that will work,” he said. “I think they’re going the right way.”

Although Harrower advocated a pragmatic approach to how data is collected, he emphasised that it’s still important to make data as accurate as possible if it is to inform effective procurement strategies and reduce costs. “Unless you actually base [the procurement strategy] on the real numbers then you could very well finish up with the wrong strategy: it’s fundamental,” he said.

Lee Ferguson, head of performance improvement in procurement at the Department for International Development, compared it to prospecting for oil: you might think there’s £10m of savings to be found in one area, but six months of mining for data might reveal only £100,000. Without good data, “you end up drilling a lot of holes,” he said.

There is a balance to be struck, however. Liz Hirst, director of commercial services at the Valuation Office Agency, said you can’t “wait for the data to reach perfection” before improving procurement policies. There is much that can be done to save money and improve the value of government spending, she suggested, even without detailed data.

Harrower agreed that rather than waiting for perfect data before changing strategies, “you need to get on and do both.” So even without a final agreement on common definitions and standards, he continued, it is important to improve the collection of data: “You might not be able to make a lot of sense of it [at the centre], but it’s important that people get the processes in place so that they actually can collect this data.”

Spotting errors

As with developing better procurement strategies, perfect data isn’t a pre-requisite for reducing fraud and error costs. Almost all of the organisations for which PRGX works struggle to provide straightforward, clean sources of data, said Francis – including private sector companies – but savings can still be found.

The anomalies which point to fraud or error are “all sitting there inside your data,” said Adam Simon, global managing director of PRGX; finding a way to mine that data using internal or external expertise has “fantastic potential” to find savings.

Payment errors are particularly likely to be high, Francis suggested, if an organisation has changed its finance system in the last six years; has a high turnover of staff; or doesn’t strictly impose three-way check systems that match purchase orders to invoices and receipts. When delegates discussed the reasons for payment errors within departments, a common theme was either a lack of adequate data to make the payments correctly; or users’ failure to comply with systems for maintaining data and checking payments.

Ferguson added that payment teams may be focusing on meeting the five-day payment target set out by the last government. This, he said, “drives the wrong behaviour” in terms of payment accuracy. Taking all these factors into account, it seems that payment errors could be reduced by focusing on both keeping data up to date, and holding users accountable for the speed and accuracy of payments.

Lesley Godwin, a procurement manager from the Home Office, gave another example of how improving accountability can help to tackle fraud. UKBA pays hotels to accommodate asylum seekers, she said, but in the past it had no way of checking whether the asylum seekers had in fact stayed at particular hotels. The shared service centre which processes payments could only check whether an invoice matched up with a receipt on the system. Nobody was asking whether the invoices themselves were actually correct, she explained, until budgets were devolved – then a change in the way purchase

“Until the introduction of [the new system] people weren't made responsible for budgets and it was very much: ‘We'll just spend what we like. If we need any money, it comes our way’,” she said – but the new system has given individual managers responsibility for their budgets, forcing them to be ensure that what they are paying out is accurate.

Now copies of all accommodation invoices are sent to an internal UKBA team in the UKBA, which checks the invoice information – so they can spot when a hotel is charging fraudulently, and discourage suppliers from attempting to overcharge in the first place.

Users and local accountability

This example, suggested Ferguson, demonstrates that improving data is often not a systems issue – “Systems can do whatever you want them to do,” he said – but an issue of accountability “and how it’s applied at a local level”. Other delegates also emphasised the importance of ensuring that individuals enter data accurately and follow new rules on categorising spend. Harrower said that – although data can be improved retrospectively with business intelligence tools that can allocate items to an agreed taxonomy of categories – it’s “much more effective to get people to apply their procurements into the taxonomy up front if at all possible”.

For example, he said, “if a purchase order is spread across several categories then you need to put proportions [into the details of the order] because trying to do that retrospectively is a pretty hopeless task”. This sort of detail must be supplied by the originator of the purchase order request – so understanding of the importance of good data, and accountability to maintain good data, must be pushed right down to the levels where requests originate, not retained in finance teams.

For Ferguson, the most important factor in improving the quality of procurement data is ensuring that compliance with systems is “mandated at the highest level”. Ministerial endorsement, he said, will be vital to ensuring that all employees understand the importance of getting data correct.

The experience of introducing the coalition’s spending control regimes from the centre has shown that, with strong political support, mandation can have an effect, said Harrower. When spending was frozen in various categories – such as consultancy and advertising – “everybody said: ‘It won’t happen; they're not going to bother following up’. But in reality it has happened and it has been followed up.” What has made the controls a success, he said, is the “political will and a burning platform: these things have definitely come together and have helped to change the mindset to a very great extent.”

Public accountability

As the round table continued, Simons noted that he could see two main discussions emerging around the value of collecting spending data in a department. To realise cost savings, he said, “you can afford to be far more pragmatic, work around systems, find ways of getting data and doing your consolidation.” But improving accountability to the public on how money is being spent is a “much more fundamental type of social requirement” which requires a greater level of investment and accuracy.

The issue of public accountability was highlighted as delegates discussed the different ways in which data is held in departments. Different teams collect financial information in different ways and for different purposes, said Emma Eckered, commercial accountant in ICT finance at the MoJ: procurement teams work in cash, for example, while financial teams are using the data to create annual accounts.

Eckered pointed out that the differing ways in which procurement and financial teams treat spending information can mean that they offer different answers to the same question. If you ask the finance team how much business travel has cost your department in a given month, for example, they supply the cost of services used in that month, even if invoices haven’t been sent in the same time period. The procurement team, though, supply the total of invoices received in that month – and the two figures may not align.

This sort of discrepancy can present “a reputational risk” suggested Dave Pankhurst, a principal analyst in resource modeling at the Department for Work and Pensions. If a department is “providing advice to ministers and public and through to the parliamentary questions,” he said, “we need a single source of data that represents our view of the answer”.

While these problems may occur because of the different ways in which information is structured, rather than real differences in data, it’s of course important to ensure procurement or resource decisions are based on a shared view of the data. For Diana West from the Ministry of Defence, the key issue is whether information – regardless of source – is being presented to decision-makers clearly so that they can understand it and use it to make better choices.

She returned to the importance of enforcing clear accountability at all stages of data collection and use, whether that be for entering data correctly, estimating costs accurately where actual spend isn’t yet known, or making decisions on how that data will be used in the business. She feels the transparency agenda has already begun to help make people more accountable for the quality of their data and estimates, she said.

Hirst, too, was positive about the transparency agenda. “It strikes me that [this] is already proving to be a helpful move from this administration,” said Hirst. “The data may not be perfect, but letting it see the light of day means that people can ask questions about it.”

West said there have been examples where the MoD has been able to use data released under the transparency agenda to compare its purchases with similar ones made in the Home Office, offering a wider pool of comparison to ensure that the department is paying sensible prices. “The transparency agenda is opening the doors to allow everybody to think in that way,” she said, so that even outside the common procurement categories departments can learn from each other’s work.

The fact that transparency data is allowing departments to make improvements – even though it is far from perfect – reflects Harrower’s view that civil servants can’t wait for data to be flawless before taking steps to make better use of it; and nor should they. Pankhurst summed up the general feeling when he asked: “Do we need better data? Actually, we’ve got absolutely reams of data in our departments; they're stuffed with it. It’s all about making better use of what we’ve got. The data hasn't got to be perfect. [We should] use the skills that are out there to analyse it, and use that analysis properly in the business.

Written by Suzannah Brecknell