a successful IFMIS story

[John Short]

Just completing the fieldwork on a PEFA in a country: it is encouraging to see an IFMIS having been rolled out that has been well specified with business practices and change mangement addressed and as a result improvement in the relevant PEFA indicators beginning to show compared to the benchmark PEFA.

[John Short]

Following a full review of business practices, the IFMIS project was started with installation of initial modules (General ledger, Appropriations, Expenditure controls) and training on applications with addressing change management as a key element.  The underlying philosophy was Think BIG, start small: the project was developed in phases ensuring successful implementation at each stage before moving on to new modules (Purchasing, Revenue, Assets, Inventory) and further ministerial roll out.   That summarises the PR, but it has been a success and the impact can be seen on commitment control and in-year reporting of budget execution and annually, leading to audited accounts within the year.  All of these were unheard of prior to the IFMIS.

[petagny]

There's a strong 'back to basics' movement in PFM reform emerging at the moment, an IFMIS being considered as a later stage of development. It would be interesting to know whether some of the 'basics' were in place before IFMIS development or whether the IFMIS actually helped in driving or cementing some of the basic reforms. Or can the staged implementation described be phased to coincide with progress on other fronts?

[halle]

We (FreeBalance) have noticed recent efforts to schedule implementations to be more aligned with ongoing PFM reforms.  A mundane operational example was in a country where we had to "wait" for the parliament to pass a law allowing digital signatures before we could implement electronic authorizations.  This didn't materially affect the implementation but it did create dependencies.  On a practice model perspective, we have noticed more countries and advisers are now focusing on platform centric approaches for managing and gating their reforms.

Regardless of the necessity and design, a sequenced approach to implementing an IFMIS in lower income countries is absolutely critical given practical capacity constraints.  As well, it reflects good risk management.  Conveying that to the authorities in the early stages have at times been frustrating.  One approach to deal with this resistance is to host a study tour of nearby countries as well as role model countries.

Not sure if specific examples can be given so feel free to contact me if you want me to expand on this.   

Thanks,
Hal

[petagny]

Thanks Hal.

I think we need to hear more about successful IFMIS implementation because there has been an over-emphasis on the failures recently, leading some commentators to suggest that this is too advanced a reform for many countries and turning some donors off supporting it. To me it seems that there are ways of tailoring the reform to local needs and capacities, the mistake is not promoting IFMIS, but promoting 'state-of-the-art' approaches too quickly.

I've heard John talk about some impressive improvements in scores for repeat PEFA assessments as a result of successful IFMIS implementation. I've also just read a draft PEFA which makes it clear that further advances in certain areas will be contingent on full IFMIS implementation and wider coverage. Handled correctly, IFMIS seems like a technical fix that can potentialy deliver results.

[STONE]

As John puts it "following a review of business practices".  That's the key.  But which business practices - the current ones or the desired business practice in the reformed system.

[John Short]

I put together a review of the IFMIS in Kosovo based on Edwin Rodin-Brown Integrated Financial Management Information Systems: A Practical Guide April 2008 USAID and discussions with a consultant who worked in the MOEF Kosovo on the installation and continued implementation of IFMIS. This gives an answer to Stone’s question at least for this example, but one that is working well.

Under the transitional administration led by UNMIK, an administrative structure was put into place pending establishment of Kosovar Government and a basic financial management system was created. This initial system was a simple tracking mechanism in Microsoft Excel, the purpose of which was to get an immediate handle on spending and donor funds disbursement. However, within a few months the need for a more sophisticated, Government-wide IFMIS became apparent. The objective was to integrate the 30 Municipalities and the future line ministries and spending agencies under one central system.

Financing of the IFMIS implementation was jointly undertaken by the Canadian International Development Agency (CIDA) and the Swedish International Development Agency (SIDA), with contributions from USAID including all the administrative instructions and procedures. The basic financial envelope was approximately 11 million Euros.

The IFMIS assistance included an E-Government software system and Free Balance eFinancials, produced by a Canadian firm and also used by the Canadian Government. Having no choice in the software package was slightly unusual, but Free Balance is in itself an unusual package for it can run as well on a notebook as it can on a large-scale network.  Consequently, it was possible to set up a pilot system for the CFA in less than 26 days.  Further modules were purchased by the "Government" (PISG of UNMIK), including assets, revenue and purchasing modules.  Currently, the Kosovo Treasury pays an annual maintenance fee.  Part of the annual maintenance fee covers FreeBalance funded staff who sit in the Treasury and assist with FMIS issues- this arrangement is an element of the success of the project (long term quality trained highly paid staff).

Once the first emergency budget for Kosovo had been developed, the IFMIS implementation team set out to install and configure the basic system architecture. The challenge was to set up a system that would work for the entire future administrative structure.  There was an important decision to be made "is this an IT project or a Treasury project" - the decision made in Kosovo was that it is a Treasury project and this was a key element of success - most of the challenges in implementation relate to aligning system configuration and PFM business processes which is not in the purview of IT staff.

There was an assessment process, identifying available human resources and system requirements. The Steering Committee decided early on that the system should include functionality for full accounting (general ledger), treasury operations, and revenues. The resource assessment showed that many Kosovars had worked under the previous Yugoslav system. Therefore, it was decided to use, wherever possible, procedures and practices that corresponded to the old system.

On this basis, a Chart of Accounts (CoA) was developed, agreed on and adopted using the old Yugoslav model. The CoA conformed both to local requirements, and to international standards, such as those of the IMF and the EU.  This gave the new administrators a familiar set of rules and regulations by which to work, while also meeting international requirements. A further key decision was to connect all Budget Organisations s to a central database with common CoA, including municipalities.  A suggestion had been made that each Budget Organisation should have their own database due to network difficulties which would have created awful problems with running a Treasury Single Account (TSA), allocating funds efficiently and then consolidating data for reporting.  This was rejected and now there is a networked common database, common CoA, TSA and real time consolidated reporting capability.

Much effort was also placed on developing interfaces with commercial banks and the central bank, which was important in enabling all transactions to be made by bank transfer (revenues & expenditures) and thus avoid opportunities for cash funds to be mishandled and enhancing reconciliations.

At this stage in the implementation process, a meeting was organized to present the complete new system to financial and executive personnel of all of the municipal administrations in the parliamentary assembly hall. After the meeting, comments were received, certain minor changes were made, and the system configured.
The program for the Kosovo roll out included:
•   Systems demonstrations to show the functionality to future users;
•   Training workshops for both senior administrators and managers and for general staff; and
•   Change management workshops;
•   Connection of all budget organizations to a central database through a Wide Area Network (initially the UNMIK WAN).

Municipalities were the first users to be connected to the system due to their geographic isolation from Treasury regional offices and the desire for decentralized operation of certain financial management functions.
The data originally entered into the system during the emergency phase were migrated to the new CoA and the system gradually expanded as the different municipalities were brought on line.  As the system expanded, a significant training program was undertaken, with all current users of the system having been trained by local staff, funded predominantly by the Swedish International Development Agency (SIDA).

The other major challenge was political, that of addressing the three Kosovo municipalities that are majority Serb.  The IFMIS implementation team adopted a non-partisan approach, configuring the system and disseminating all documents and information in Albanian, Serbian and English.  As the system incorporated the Yugoslav CoA, the new system allowed Serbian municipalities to extract and communicate their financial data to the Serbian Government in Belgrade, which after the war continued its financial support to the Serbian communities in Kosovo.  So, even reporting and accounting were made easier with the introduction of the new IFMIS, which was configured to work in multiple languages depending on the user’s system profile.
The core system was up and running in all three languages in 2003 and has been expanding in functionality ever since.  The CoA has been progressively updated to reflect the growing sophistication in financial management and reporting and to achieve greater conformity with the IMF’s Government Financial Statistics.  All 80+ Kosovo budget organizations, including existing and new municipalities, have been added to the system and there are over 650 active users.
In addition to the “foundations” general ledger module, “revenue”, “asset management”, “purchasing” and reconciliation modules have been added to the system and a direct interface established with the central bank for electronic payments.  Kosovo has now reached the point where all revenues and expenditures are transacted via bank transfers, completely avoiding the use of cash.

In some cases, Kosovo was the first to implement certain features of FreeBalance and played a significant role in developing workable configurations and solutions that have been copied in other developing countries using FreeBalance.   Kosovo’s Treasury Department and FreeBalance have formed an agreement under which FreeBalance funds local staff who works actively with the Treasury to manage and troubleshoot the system. This relationship has produced mutual benefits: Treasury has gained access to expertise not otherwise available (on government salaries), and FreeBalance has had the benefit of understanding the practical challenges of implementation and the ability to develop solutions in a real-world environment.  The success of this partnership has been such that FreeBalance has established a local office through which Kosovar experts are assisting implementation in other developing countries.

[halle]

John's summary of Kosovo is excellent.  Its covers the entire time range from initial implementation start  to post-implementation / sustainability.

I would add a few more items that some readers may not know:

1)  Kosovo was FreeBalance's first 'international' customer.  In that regard, it is in many respects the most advanced of FreeBalance's customers in terms of procedures, processes and capacity.  The Kosovo maturity level is comparable to FreeBalance's Canadian and US customers and more so in some areas such as Treasury functions.  Some other FreeBalance customers are still in their nascent stage and are on a different timeline to achieve comparable objectives and milestones vis-a-vis Kosovo's.  For example Southern Sudan is running a "basic" configuration and will be for some time; Timor Leste is running a 'medium' and upgrading to 'advanced'; Guyana is running a 'medium-advanced' as a status quo.  .  The important lesson FreeBalance has learned is that although all customers are on the same software, it is vital that the system can support and reflect each customer's specific needs in terms of process and capacity, as well as provide a streamlined mechanism that can progressively grow with the customer as their needs matures and advances.

2)  The original model that FreeBalance had with its customers has evolved and grown over time.  The original approach was a traditional vendor-customer relationship.  It has now been dramatically transformed.  For example, FreeBalance now actively solicit governments to participate in its internal processes.  For example, the (FISC) FreeBalance International Steering Committee is an annual meeting between FreeBalance senior management and senior representatives of each country that has implemented FreeBalance.  One of the customers is voted as head of the group (never FreeBalance).  One outcome from these meetings is a list of product and solution needs for the upcoming year.  This list is controlled, managed and prioritized completely by the customer and is given to FreeBalance at the end of the meeting.  FreeBalance will then report back to the group in the following year on the status of that list.  To add relevance to this annual process, FreeBalance allocates a significant percentage of its development budget to FISC to ensure those priorities are actually implemented.  The FISC meetings are completely separate from the user groups and other meetings that FreeBalance conducts at an operational level.  To our knowledge, no other vendor has transformed and committed themselves in such a manner to their customers.  This approach is typically not feasible for vendors targeting multiple industries and market domains.  However, as FreeBalance solutions are exclusively government centric (all customers are public sector), it is not constricted from following this customer model.

3)  There are country case studies (Kosovo is included) available on the FreeBalance website.

4)  The PEFA scores are an important PFM component.  FreeBalance is now incorporating the scores into its analyses, country case studies, etc.  This will take some time because of the complexity of matching the scores with the actual project timelines for each country, etc.

Thanks,
Hal

[atseacliff]

John/Halle

Interesting piece and supplemental comments -thanks.  Going back to Stone's original question I guess that the specifics of the Kosovo situation dictated that the system was put in place quickly and that there was minimal time available to address some of the procedural and policy weaknesses inherent in the current system - am I right?  I am working in an FSU country at present which has most of the basics in place, reasonably strong upstream PEFA scores and a commitment to decentralised decision making. The development of an IFMIS which which enables budget organizations to manage their resources is seen as the most significant "enabler" to implementing the broader PFM strategy. Current business processes are well documented however the Government is keen to conduct a broader review on the changes to business practices expected of line ministries and budget organizations as a result of the strategy.  This work will be conducted this year and will be an input to the development and installation of the IFMIS; probably in 2012.

One other comment - I'd be interested to be kept posted on Halle's work on PEFA scores.  I suspect that direct collation being the introduction of IFMIS and individual PEFA scores would be difficult to demonstrate however it would be interesting to identify areas where it had been an enabler to improved PFM performance. 

     

[halle]

Hello,

Yes, Kosovo was originally implemented as an emergency system - hence the 1 month timeline.  There was minimal external guidance on the configuration so it was implemented with just the core settings.  The software is designed to be configured via the user interface (it normally doesn't require code level modifications to support changes to business procedures and the like; if a change is required, then FreeBalance development will actually make the change and expose it as a configuration option available to all customers).  So as the situation stabilized and the procedures and processes were documented, the system's configuration was subsequently upgraded to reflect those changes.

A typical major bottleneck that we find some customers have raised is that if the business process is not finalized, and may not be finalized for some time, then the system should be delayed because the effort required to reconfigure the system once the final process is decided upon can be substantial and can be a major drain on morale and available resources, especially given the country context.  As the FreeBalance software model is different based on the exclusive government focus, much of the critical path dependencies on the IFMIS software implementation and configuration has been removed.  This has helped the Kosovo government to devote their energies on the their critical success factors such as capacity, institutional design, administrative absorption, decentralization and the like through the years as the system was upgraded both in functional coverage, usage and versions. 

One thing we have seen is that it is absolutely critical that change agents are in the key positions to make decisions.  Without those champions Kosovo would not be where it is right now.  And one reason why some countries are not going as fast as Kosovo's.

I'll try and update this forum on the PEFA work but it is something that will take us some time to do.  I'm trying to make this post non-partisan but its hard

Thanks,
Hal

[atseacliff]

Thanks for your response to my questions. In my case the process has been drawn out because the government readily recognises the risks you highlighted. I have no direct experience of FreeBalance but have been able to draw plenty of more generic lessons from this tread. I wouldn't worry too much about partisanship.

Regards

[STONE]

I like your  comment about the process being drawn out, if one is re-forming the thing that is government one wouldn't want it to be rushed, and perhaps we do want it to be what Sir Humphrey always sought for new things, for them to come in due course, in the fullness of time, at the appropriate juncture.

[atseacliff]

And once you've made a decision there is always donor procurement practices to fall back on