Developing countries have been encouraged to implement Integrated Financial Management Information Systems, but there are risks associated with this approach and there is little evidence of success

Late last year Paolo de Renzio published an interesting report reviewing the progress made by 16 aid-dependent countries in improving public financial management over the period 2001 to 2007.  Despite the significant levels of technical advice provided, only half of these countries showed a significant improvement.  This raises the question of whether the advice provided is appropriate.

One of the standard reforms introduced in many developing countries over the last couple of decades has been an Integrated Financial Management Information System (IFMIS).  This is a concept that I had not come across until I went to Africa.

An IFMIS is a common computerised financial system used in government offices. It often covers all ministries, departments and agencies and may be extended to include regional and local governments. It usually comprises the core modules of budgeting, cash and debt management and accounting, and in some cases ancillary components such as payroll and procurement.   However, as with any complex computer system, the risks involved in adopting an IFMIS are very high.

In this country we know this to our cost.  To quote but one example, the NHS’s National Programme for IT was scrapped last year following years of delays and technical failures.  As an example of the costs involved, earlier this month it was announced that an agreement had been reached with Computer Science Corporation to release more than £1bn back into the health service.

Similarly in France, the national audit office issued a major report in August last year on public financial management reforms.  On the French IFMIS system, Chorus, initiated in 2006, the audit report stated:

‘The project experienced delays in development…. In addition, the system is complex for managers to use as it involves multiple data entry and reduces their responsibilities. Finally, as it currently exists, Chorus does not allow the development of true cost accounting nor effective management cost control.’

In developing countries, the risks associated with major IT projects are even greater. According to the eGovernment for Development Information Exchange (co-ordinated by the University of Manchester’s Institute for Development Policy and Management) e-government initiatives are rarely successful – 35% are total failures, 50% are partial failures and 15% are successes.

One of the reasons for this is that the economics of IT investment in the Global South is different from that in OECD countries – labour is relatively cheap and IT is not available locally and so is more expensive to purchase and to maintain. Average public sector wage costs in Africa can be one-tenth or less than those in the West; average ICT costs can be two to three times higher.  E-Governance and automation using modern IT technology therefore results in replacing cheap civil servants with costly IT.

At a World Bank presentation in late 2003, Bill Dorotinsky provided a useful overview of the Bank’s experience of providing over $1bn to finance IFMIS projects over the previous decade. The average time for the completion was over nine years for African projects and the average cost of each of the 34 projects worldwide was $12.3m.

If success is defined on the basis of budget, time and deliverables, then only 21% of these projects were considered to be successful. An even gloomier view was provided by assessments of the same projects by the World Bank staff. In only 6% of cases were these projects considered to be sustainable.

More than a decade earlier, the United Nations warned that, ‘attempts to computerise an entire government accounting system within a few years were doomed either to failure or to dramatic cost and time overruns’.

However, despite these warnings, many developing countries are still being encouraged to implement further IFMIS systems.  In November last year, the Federal Government of Nigeria, for example, signed a $29m contract for an IFMIS with only just over two months to go before the ‘go live’ date of 1 January 2012.

In Ghana, the government is currently involved in implementing an ambitious $65m IFMIS project, which, according to one local public financial management expert, ‘is going to fail’.  This is despite the Ghanaian government having spent up to $50m on a previous IFMIS project from which, in 2006, the Deputy Controller and Accountant General suggested ‘there has as yet not been one Cedi [local currency] benefit from it. I have not used [the IFMIS] to generate one report yet’.

The appropriate lessons are there for those who wish to learn.  The IFMIS in Tanzania is often quoted as being successful.  But it was arguably successful because it was an initially just payment system (excluding payroll) based in a single office, rather than being a comprehensive integrated financial management system.

In contrast, at the recent annual conference of the East and Southern African Association of Accountants General (ESAAG) a leading World Bank public financial management expert argued that because most African governments now had an IFMIS, any move to accrual accounting would be much cheaper – but that is another story.

Andy Wynne is editor of the International Journal of Governmental Financial Management and has been a consultant for the Federation of Accountants and Auditors General of West Africa, the Commonwealth Secretariat and the International Budget Project