We do not evaluate our own advice.
Where we have designed or implemented a programme, we decline the evaluation of it and say why. Independence is not a claim. It is a set of refusals.
§ 01 · The firm
Thirteen years of being the people who have to say what actually happened.
Colnag International is an independent advisory firm working on public money: how it is planned, spent, accounted for and evaluated. We were established in South Africa in 2013 and now work across Southern Africa and the Middle East.
§ 01.1Origin
How the firm started.
A national evaluator panel, an incentive that had already spent its budget, and a question nobody had answered.
In late 2012 our founding team was selected onto a national evaluator panel in South Africa and given an implementation evaluation of a services sector incentive. The programme was live, the money was moving, and the department needed to know whether the design was doing what the policy said it would. The final report was approved in 2013. Colnag International was formed around that work the same year.
What followed was a decade of the same discipline applied in wider circles. Evaluations of innovation support, of a presidential service channel, of the national evaluation system itself. Then public finance: expenditure reviews, cost models, budget reform. Then the mandates that only arrive when something has gone wrong, which is to say forensic investigations for boards and regulators, and the wind-down of state-owned enterprises that could not be saved.
In 2013 and 2014 the practice moved north and east. Namibia and Botswana first, through budget support and sector programmes financed by the European Union and the African Development Bank. Then Jordan, where a five-year competitiveness programme and a monitoring and evaluation project put us on the ground for the better part of a decade. Then the Gulf.
We have never marketed a methodology. What we sell is the willingness to reach a finding, write it down, and defend it in the room where it lands.
§ 01.2Chronology
A sequence, not a highlight reel.
Dates mark when a line of work opened, not when it ended. Several of these mandates are still running.
- 2012Selected onto a national evaluator panel in South Africa. The first implementation evaluation begins in October.
- 2013Colnag International is founded in South Africa. First mandates outside the country follow in the same year, beginning with a national health and demographic survey in Namibia.
- 2013First long-horizon programme work: a five-year competitiveness programme and a monitoring and evaluation support project in Jordan.
- 2013Technical assistance opens on a sector policy support programme of 103.4 million euro in Botswana, running to 2018.
- 2016Public finance practice established. First public expenditure and financial accountability assessment. The treasury spending review programme begins in South Africa.
- 2016First mandates in the Gulf, covering sovereign investment restructuring and multi-year assurance on a world expo programme.
- 2018A whole-of-government review of roughly R50 billion a year in business incentives, jointly mandated by the presidency, the treasury and the trade department.
- 2019A national integrated tax administration system goes live in Namibia, replacing the legacy revenue platform.
- 2021Climate and sustainable finance work begins in earnest: nationally determined contributions, biodiversity budgeting and climate fund evaluation.
- 2023Return to national evaluation at scale, with a systems-lens evaluation of a food and nutrition security plan spanning several departments and both spheres of government.
- 2025The ninety-first technical spending review is completed. Appointment to a government forensic audit follows negotiation for specialist data analytics capability.
- 2026Work opens on a national sustainable finance taxonomy with a central bank, a non-bank regulator and a finance ministry.
§ 01.3Method
Five things we hold to.
These are not values on a wall. Each one has cost us work at least once.
Evidence before opinion.
A finding has to survive being read by someone who wishes it said something else. That means the data collection is designed before the conclusion is interesting, the counterfactual is stated even when it is inconvenient, and the limits of what we can claim are written into the report rather than into a footnote. Where the evidence does not reach, we say so.
Built to be used.
Reports that are not used are the sector’s most expensive habit. We design every mandate backwards from the decision it is meant to inform, whether that is a budget bilateral, a board meeting or a cabinet memorandum, and we deliver into that calendar. Recommendations name the institution that has to act, the instrument it will act with, and what it costs.
Confidential by default.
Much of our work concerns money that has gone missing, systems that have failed, or entities being wound down. Clients tell us things on the understanding that we do not publish them. So our public record is deliberately thin: what we did, when, and for what kind of institution. Details go out on request, with the client’s consent, and not otherwise.
Local teams, one method.
Every long-running mandate we hold is delivered by people who live in the country and will still be there when the programme ends. In Jordan, seventeen evaluations and a national population survey were delivered through a local partner structure. What we standardise is method, quality control and independence, not accents, and not the answer.
We stay for the hard part.
The interesting phase of an assignment is not the diagnosis. It is the eighteen months afterwards, when a revenue platform has to actually go live, a procurement regulation has to be drafted, an entity has to be liquidated without destroying value, or a recommendation has to survive contact with a budget process. A large share of our mandates are multi-year for exactly this reason.
§ 01.4Scope
What we do not do.
Being clear about this is what keeps the rest credible.
We do not lobby.
We advise institutions on their own policy and spending. We do not represent commercial interests to the governments we serve, in any market.
We do not pre-agree findings.
Clients see and comment on drafts. That is good practice and it improves accuracy. What a client cannot do is set the conclusion in the terms of reference.