In many countries there is a patchwork of often incompatible and incoherent legacy systems (sometimes within a single department) many of which are either at the end of their lifecycle or out of support completely. Systems are often geared up for compliance and historic reporting with little evaluative or predictive power. They can be hard to interrogate flexibly or reconfigure to meet changing needs.
Often individual departments have modified off the shelf systems or built their own bespoke systems that are then difficult and expensive to support or modify, in some cases making them over-reliant on particular contractors who charge premium rates for their services. In the worst cases opportunities are missed because responsibility for system design and commissioning is given to staff without sufficient relevant expertise or domain knowledge. This can give rise to unrealistic expectations (which contractors charge a premium to meet), or in worst cases a failure to deliver systems of sufficient utility to meet the original objectives. As much of the development activity is outsourced misunderstandings can and do arise as to system capabilities or functionality, resulting in missed opportunities and more limited functionality than is needed.
Lack of high-quality management information
Under-investment in systems is also often accompanied by lower levels of investment in the activities designed to utilise the capabilities and outputs of systems. Most government departments do not produce high quality management information, the task of producing it at a government wide level is practically impossible. This is often due to a lack of coherence in information standards and categories, configuration, operating segments and charts of account. Without a shared taxonomy it is not generally possible to interrogate systems for information regarding relationships between expenditure and activity, other than at the top level.
Without shared information standards and taxonomies government will not be well placed to take advantage of emerging analytical, automation and machine learning capabilities. While modern systems increasingly can operate with unstructured data, dealing with poorly configured information does increase the complexity and development effort required to get utility from them. If government is going to be able to take advantage of these technologies, it will need to improve its basic data management as well as making greater investment.
There are two particular problems that often need to be tackled in order to improve matters:
- Establishing a coherent set of information and configuration standards for financial systems – these should be centrally mandated for all government departments to follow with agreed transition plans for all departments onto a common standard over a defined time period. While there may be some exceptions - this will form the basis of a more coherent set of financial systems and enable. Setting up a central Systems Information Standards Office would be one way to create the standards supported by central funding dedicated to upgrading and conforming systems.
- Maintaining financial systems expertise – the expertise required to establish standards and configuration for financial systems is in short supply in industry. Consequently, rather than disperse it among government departments, it could be recruited and concentrated in a centre of excellence – ideally within the Systems Information Standards Office, described above. This could in theory operate outside the normal constraints of the public service reward and grade frameworks and enable specialist expertise to be concentrated in an environment where the skills were understood and could be sustained.
Given the scale of government and the importance of financial management systems the costs of specialist staff would be justified by the improvement in overall financial systems performance and procurement from having an in-house expert team. All departments would be required to take and act on the advice of this centre of excellence when procuring and configuring financial management systems. Such a centre of excellence could in theory cover its costs by recharging its costs to government departments using its services, but it might be more politic for the cost to be borne centrally.
There is no quick fix however. Both investment and enhanced standards are required, but it will not be a trivial task introducing them across organisations of the scale and complexity of most governments. Quick wins might however be possible for work designed to ask specific questions. Experience from the corporate world shows that faster incremental improvements can be made where systems are aligned and data cleansed to provide a small set of specific analysis. This small set is then grown incrementally to encompass larger amounts of data and potential analysis. The experience of business generally is that attempts to cleanse sufficient data all at once to provide a wide range of analysis, invariably founder under the scale of effort required.