4.8 The results illuminated some important aspects of the information technology market and the partnership's relative position within it. The main conclusions were that:
■ prices paid by EDS for capital and infrastructure were competitive;
■ the information available did not allow the comparative efficiency of the Department's infrastructure to be clearly assessed;
■ EDS charges for staff had increased at a slower rate between 1995 and 1997 than salaries in the information technology industry generally; and
■ EDS software development productivity had been relatively low on some major projects although it was improving at a faster rate than in the industry comparators.
Details of the benchmarking work are shown at Appendix 4.
4.9 The evidence on staff cost inflation and software productivity was provided by commercial benchmarking organisations or established benchmarking groups. We believe that it is sound and supports the conclusions drawn. In contrast, the Department had considerable difficulty in gathering evidence on hardware and infrastructure costs directly from other organisations in the private and public sectors, with the result that the sample was possibly unrepresentative. We therefore believe that only limited assurance can be drawn from the conclusions based on this information.
4.10 The findings on software productivity were not unexpected, given that, when it outsourced the service, the Department recognised that this was one of the areas where performance needed to improve under EDS and that this would take some time to achieve. The Department and EDS explained the variation by pointing to the unique difficulties faced in developing complex taxation software to tight timetables. However, the findings have encouraged them to take further initiatives to control and reduce costs, including greater emphasis in proposals on the factors affecting cost, and the employment of consultants to improve the use of software development tools and to provide training.