Financial management for security sector
What are core government systems for security sector finance?
This about how expenditures are authorized, and about how payments are executed. Thematically, this about financing payroll, operational expenditures and infrastructure. Functionally, this is about organization, roles and responsibilities, work-flows and technologies.
In most countries, these systems are slightly different form the civilian systems. Security organizations have particular needs.
However, they all integrate with the governments core treasury and payments systems, usually. There are unfortunately countries where expenditure authorization and payments executions are sidelines from the treasury.
There is often a significant focus on issues like fingerprinting, biometrics and ghost workers when discussing public finance, security sector and fragile states.
But without the basic organizational and work-flow capacities fixed, the technology will not resolve the problem.
In fragile and post-conflict countries, there are typically old-school bureaucratic systems. These can be quite clever. But after years of war, the ingenuity of the original design is all gone. Those bureaucratic practices are more or less broken.
Its quite an undertaking to redesign all of this. And its even quite a more of an undertaking to actually implement and rebuild.
Yet, without these basics in place, the security sector will remain dysfunctional. Peace agreements and carefully negotiated security sector agreements in fragile states may collapse as there are no foundational systems to deliver. This is foundational.
Getting the core government systems to work is critical for success in security sector reform.
Fixing parallel systems managed by international donors
There is often considerable international involvement in a post-conflict country security sector. In may places, Somalia and Afghanistan included, this has also focused on building parallel systems, managed buy U.N agencies or other contractors. These parallel systems channel international funding to salaries and to other expenditures.
These systems may have been built, originally, as an emergency response to weak and non-credible government systems. The government systems on their own offered no credible alternative. They may function poorly,. and be uncomfortably opaque with respect to control and accounting.
International partners then build dedicated processes, systems and practices that are operated for purposes that government themselves should be managing, and are, in parallel with tax-payers money. Yet, little investment is made in building the government systems, but in the continued development of the parallel systems.
These phenomenons are less prevalant on civilian side. Concepts for making investments to strengthen core governments systems, and to direct aid spending using government systems, are much more evolved.
The parallel systems may have points were they integrate with government systems. Don’t mistake such partial integrations as the systems being truly aligned. Key questions are about who makes spending and payment decisions, where funds are held, and how payments are executed.
Over time, the complexities of these parallel worlds can be quite mind boggling. In Afghanistan, there are substantial challenges in aligning core personnel registries, that have inconsistent designs, and inconsistent use of a.o ID numbers, creating fundamental challenges for control of security sector personnell.
A similar situation is emerging in Somalia. There were recently three main personnel registries, not reconcilable, but presumably covering the same personnel. Basically the names in the registries could not be matched across the systems. All the registries captured multiple data-points for each soldier.
To reconcile these, we built tests that deployed jacquard-index character matching algorithms developed open source by google. These calculations basically compare a number of variables for different records (in this case soldiers), and assign a probability that they are the same. The computers ran hot for several days calculating probabilities.
The discrepancies were surprising. One of the systems indicated at most high (+90 percent) probability of match for 77 percent of the soldiers in the two registries. The other systems had single digit match probabilities.
It does necessarily mean that the registries are containing names of differentsoldiers, but it does mean that even if the individuals are the same, they cannot be identified across the government and international systems. There was no consistency in name spelling, in practices to ensure quality of data entry, no consistent use of unique ID numbers, and biometric fingerprint information was captured with different algorithms and not compatible, and only one system had captured photos, but of poor quality, that could potentially have been used for face recognition. And a series of other technical issues related to the design of the information fields and how they had been imputed. The devil is in the detail.
Government eventually decided it needed to rebuild completely its own system and that the international systems could not be used.
At best such solutions are highly inefficient. Whether these parallel systems provide additional benefits, like enhanced control, is questionable. They may create a veil of legitimacy, but are unlikely to offer substantially higher value benefits.
The strategic alternative is to invest in building and strengthening government systems.
While expensive investments, they are sustainable. And those costs will shortly be off-set by reduced operating costs of running parallel systems managed by UN agencies and contractors.
Huge differences in personnel registries by international agencies and government, presumably to pay salaries to the same soldiers
We have also written about strategies for security sector finance. Thats about ensuring that plans, visions, peace accords and other strategic processes are founded on facts and can be sustained financially.
