Locks, stock and apparel

Procurement activities in all organisations share many common features, but they also face specific challenges unique to their sector or industry. Consider the case of HM Prison Service. Each day 128 prisons in England and Wales have to feed, clothe and hold securely a population of 70,000 prisoners. This is a substantial challenge in itself, but it is made more complicated by the logistical issues of delivering goods and services to secure institutions - after all, opening the gates every other minute to receive deliveries is hardly conducive to good security. Moreover, some prisons were built more than a century ago and their design does not encourage the efficient distribution and storage of goods on site.

The Prison Service's procurement of goods and services has been fragmented and inefficient in the past, and it has required large numbers of employees to manage and process transactions and stockholdings. In 1998 an internal review criticised its performance and called for change. In the consequent shake-up, the Prison Service sought to reduce the cost of goods and services. Price negotiations and the wider use of centrally negotiated contracts enabled it to report savings of L6.3 million in 2000-01 and a further L5.8 million in 2001-02.

As part of the National Audit Office's (NAO's) role of scrutinising public spending on behalf of Parliament, we recently examined how much progress the Prison Service has made towards improving its procurement activities. The findings were published in April in the report Modernising Procurement in the Prison Service.

We found that the Prison Service had made a concerted effort to improve the performance of its central purchasing team and had taken forward a number of other initiatives. But our review still highlights several areas where further improvements can be made.

For one thing, there are significant differences in the cost of procurement in different prisons, which suggests that more savings can be achieved. At HMP Blundeston, for example, the cost of staff procurement time in 2001-02 was L56,000, compared with L244,500 at HMP Parkhurst, which is a similar size and type of prison. We recommended that each prison should designate someone, other than the governor, at local level with clear responsibility and authority for the effective management of all procurement. Such a person would, for example, decide where to strike a balance between central and local purchasing.

Most large organisations have to decide whether the best procurement deals can be achieved by centralising activities and using central contracts to buy goods and services, or by giving responsibility for procurement to local managers. The Prison Service needs to strike the optimum balance between relying on central contracts and allowing local discretion and flexibility.

The main benefit of central purchasing is that it enables organisations to make best use of their purchasing power - ie, a large firm can negotiate discounts from suppliers because it requires large volumes of goods and services. The Prison Service has a central contracts team that has negotiated deals for many items. Prisons spent L177 million on these contracts in 2001-02. Our study suggested that there was still too much emphasis on local procurement. We estimated that prisons spent L74 million on locally bought items that could have been obtained from the centrally negotiated national contracts. Only four of the 74 centrally negotiated contracts available to prisons were mandatory. The Prison Service estimated that the grocery contract, if made mandatory for all prisons, could save the organisation up to L450,000 a year.

All of the institutions we visited bought fruit and vegetables locally and their annual spending on these items ranged from L20,000 to L45,000. Many of the prisons used a regular fruit and vegetable supplier, and staff informed us that their providers often delivered to neighbouring prisons as well. Even if it wasn't feasible for the prisons to use the central contract, co-ordinating these purchases locally could allow the service to reduce its costs.

But it can be difficult to persuade local procurement staff to use central contracts. Our discussions with prison procurement staff showed that many were unaware of the range of items that could be bought centrally. And many believed the prices of items on central contracts to be uncompetitive. Locally bought items seemed, on the face of it, to be cheaper. But the local managers did not take into account the extra cost of the time they spent purchasing items locally. Previous studies by the NAO across government departments have found that the cost of staff time taken to purchase an item often outweighs the cost of the item itself. This finding is borne out in the case of the Prison Service, which typically incurs L33 in staff costs every time a prison makes a purchase.

Prisons could also reduce the time spent ordering goods and services by using the government procurement card, a charge card available in central government. Consultants to the Prison Service have estimated that it can save L12 with each card transaction. This would amount to an annual saving of L3.8 million if the service were to use the card for all of its 450,000 transactions a year.

But our visits to a selection of prisons showed that use of the card was limited. Managers were concerned that that it might be misused (although with proper safeguards in place cards are no less secure than any other procurement method). Staff also expressed scepticism about their ability to reconcile monthly card statements with the records of what they bought. This was surprising, given that the card gives free access to aggregate management information covering transactions, suppliers and commodities.

Security requirements mean that movement across prison sites is restricted, and this almost inevitably means that goods are being-stocked in many different places across establishments. For example, stores are held not only in a central warehouse near Burton on Trent, but also in 547 different stock rooms in individual prisons. Holding goods across this number of locations impedes efficiency and, although there are some sites where this would pose practical problems, limiting the number would improve the Prison Service's ability to control stocks.

Another issue is the level of stock held by the service. Although it has reduced its holdings significantly over recent years, our investigations have shown that further reductions are possible. A fundamental problem is that there is no central information on the quantities of stock lines held by each prison. They do not have the information they need to compare their stockholdings with those of similar establishments, nor are they held to account for stockholdings significantly above or below the norm.

Of the 547 main stock rooms in prisons, only 149 have independent stock records kept by finance or procurement staff. Our calculation of how long the L13.6 million of clothing held in prisons might last found wide variations. Stocks in women's prisons, for example, could last from less than one month to more than six.

One of the consequences of the lack of central management information that would allow prisons to monitor stock consumption is that some are buying large amounts of routine items. For instance, the Prison Service issued 430,000 pairs of boxer shorts in 2001-02 at a cost of L870,000. On average, six pairs were issued for each prisoner place, although one young offenders' institution received 37 pairs for each prisoner place. While some of this variation could be attributed to differences in prisoner throughput, much of the variation couldn't be explained.

We were able to calculate from central stocktaking records the values of each major commodity. We estimated that reducing overall stock levels in prisons to a maximum of three months of supplies, for example, would achieve a further stock reduction worth L12 million.

Another way to establish whether prisons manage the demand for supplies properly is to look at what they discard. The Prison Service gets rid of large volumes of waste each year: in 2001-02 its total expenditure on waste disposal exceeded L2 million. The service has now set up a waste management team to review its practices. One of its first reviews, at HMP Elmley, found that it was possible to save L66,000 a year. The savings included L36,800 from reduced disposal costs and L29,200 from recovering clothing and equipment that would otherwise be lost.

The Prison Service recognises that it requires better management information to monitor progress. At present it has an ageing computer system that cannot provide basic communication across the whole organisation. It is therefore introducing a networked system that will enable it to improve the way it monitors performance and to use an integrated purchase order processing and stock control system. A service-wide IT framework would have the additional benefit of enabling prisons to share good procurement practice. This would enable the poorer-performing prisons to learn from the best.