Medium Term Expenditure Framework > IT systems applied to PFM?

IFMIS - blessing or a curse?

(1/2) > >>

Martin Johnson:
I have been fortunate to have spent the last two weeks in Zambia with the moderator of this PFM topic, Mr Smallman. In the process I have had the opportunity to understand better the very real benefits that a functioning IFMIS can bring, if and when it is implemented well and if and when the capacity is there to operate it ... including the enhanced control it can offer (without the requirement to stifle budget management), improvements in management information, security, information for audit purposes and more besides.

But our time here has also brought home very clearly the very real costs that can be incurred if decisions are rushed, ownership is weak and if implementation follows without being clear about the business processes that the IT system should be built around. And these can be expensive mistakes in terms of cash (lots of it) and in terms of the attention it diverts from ensuring sound regular government business and progress in other areas of reform.

After having seen examples of negative experiences in this country and elsewhere, I sense the beginnings of a change in attitude internationally about the wisdom of selling the idea of sophisticated IFMIS solutions in environments where a more organic approach to constructing IT solutions around sound business processes over time and integrating when appropriate.

Does anyone else sense the same thing, or am I living in a fool's paradise? Otherwise, it would be good to hear experiences from people in countries where there have been good experiences in developing and implementing IFMIS-type systems. What were the critical success factors? What are the benefits?

petagny:
I'm afraid I haven't witnessed many IFMIS successes directly myself. John Short and others talk positively about the FreeBalance system set up in Kosovo.

Bill Dorotinsky put some advice on the IMF PFM Blog a while back, including this, which seems sensible:


'FMIS versus IFMIS. Much of the work in automating PFM systems has focused on implementing an integrated financial management information system, including general ledger, accounts payable, accounts receivable, procurement, payroll, asset management, debt management, budgeting, etc. This approach might be too large to implement effectively, in a timely fashion, or to achieve results. It might be better to think of automating some core part of the system, such as general ledger, and accounts payable and receivable, with an eye to adding-on or replacing the system within a few years. Other points in this regard were to start from where you are in terms of PFM system development, rather than from where you want to be, and also to recognize that not everything may need to be automated. As PFM systems evolve, the needs will change, so the scope of the automation can be expanded. Given the rapid change in technology, it may not be feasible to plan all of these potential needs or IT options in advance.'


I attach some useful documents on the subject, including Sara Fyson's paper on what went wrong in Ghana, which is enlightening not just from the perspective of IFMIS, but also on how the donor-consultant-government relationship works.

John Short:
USAID Introducing Integrated Financial Management Information Systems Fiscal Reform and Economic Governance Best Practice Notes: www.fiscalreform.net http://pdf.usaid.gov/pdf_docs/PNADK595.pdf

and

Edwin Rodin-Brown Integrated Financial Management Information Systems a Practical Guide April 2008 USAID
http://www.fiscalreform.net/images/ifmis_rodinbrown_080425.pdf

are also useful

petagny:
The USAID assessment of the situation in Vietnam is consistent with what I have seen:

‘The Socialist Republic of Vietnam has been working on introducing a new IFMIS since 2001…

Vietnam offers an interesting example of “too much, too fast.” The Government has embarked on drastic reforms to come into compliance with international standards and agreements and to become an accepted player in the international community. Unfortunately, while the leadership races forward, the people and systems in the government administration are not capable, or not willing, to keep up with the feverish pace of reforms…

…the IFMIS strategy is too ambitious and will likely have to be scaled down if it is to have any chance of success. The system chosen requires, as in Kazakhstan, a strong communications backbone. This does not exist and it is not feasible in the short term. All these problems might have been foreseen and could have been addressed had a proper assessment been done in advance of the tender. It is so tempting to try to drive forward faster than resources can follow.’

greggpavitt:
Here in Afghanistan, I worked with a local national programmer to produce a fairly simple IFMS system for local governments. The interface and initial functionality is simple, and mimics the paper world they live in. Underneath we put enough structure is there so more sophisticated reports can be generated down the live. We just finished implementation in our core municipalities (here in the tough southern area), and will be rolling out elsewhere.

Navigation

[0] Message Index

[#] Next page