Capacity Development

01 The argument

Development finance has a long, well-documented record of capital that arrived and evaporated: projects built and not maintained, funds disbursed and not absorbed, systems installed and never operated, reforms legislated and never implemented. Post-mortems on half a century of such failures converge on a single finding — the missing ingredient was almost never money. It was the institutional capacity around the money: the organisation with the mandate and staff to own the asset, the professionals to run it, the systems to operate it, and the data to know whether it was working.

The coalition’s founding design decision follows directly: capacity development is not a side programme, a training budget, or a goodwill gesture appended to real work. It is underwritten into every material commitment — the Investment Committee will not approve capital without seeing the capacity plan — sequenced with the money rather than after it, funded from inside the deal, and measured with the same seriousness as repayment.

This is also, candidly, the coalition’s answer to the fair question of what distinguishes it from a well-capitalised lender. Money is fungible; the machinery on this page is not. It is slow to build, unglamorous to run, and nearly impossible to counterfeit — which is precisely why it compounds.

02 What “capacity” actually means here

The word is used so loosely across the industry that it needs pinning down before anything else on this page can mean something. The coalition uses it for four specific, inspectable things — and an engagement plan must name which of the four it is building, because “capacity” as an aspiration funds workshops, while capacity as a specification funds institutions.

The four interlock, and the ordering matters. Skills without an institution to house them emigrate. An institution without systems improvises until it fails an audit. Systems without data run blind. Engagements are therefore scoped against all four even when the funded work concentrates on one — the plan must show where the other three stand, or the one being built will not hold.

• Institutions — organisations with the mandate, staffing, and budget to own an asset or function after handover
• Skills — the professional depth to run it: supervision, treasury, procurement, engineering, statistics
• Systems — the operating machinery: registries, payment rails, reporting lines, maintenance regimes
• Data — baselines, monitoring, and the analytical capability to know whether anything is working

03 Who the work serves

Capacity engagements target the institutions that decide whether capital compounds — a deliberately short list, because spreading capacity work thin is one of the classic failure modes this page later catalogues.

Delivery institutions absorb the largest share of engagements — nearly two-fifths — because the last mile is where capital most often fails without support. A power plant is only as durable as the utility that dispatches it; a rural connectivity programme only as good as the agency that maintains it in year six, after the ribbon-cuttings. The exhibit below shows the current distribution.

Central-bank and regulator engagements are fewer but longer and deeper, because their effects are economy-wide: one supervisory framework protects every bank depositor; one modernised payment system carries every transaction. These are the engagements with the highest ratio of patience to visibility, and the coalition protects them from the pressure to show quick wins.

• Central banks — monetary operations, supervision, payment-system oversight, reserve management
• Regulators — the market rules that let private capital follow public capital safely
• Debt-management offices — the sovereign borrowing capability that keeps terms honest across every future negotiation
• Delivery institutions — the ministries, utilities, and agencies that run programmes after handover

04 The three delivery channels

How the work reaches the institution matters as much as the content — most failed capacity programmes taught the right material through the wrong channel. The coalition uses three, usually braided together within one engagement.

Residence is the channel the coalition weights hardest, and the job description is deliberately humble: the advisor works inside the counterpart’s hierarchy, on the counterpart’s files, at the counterpart’s pace. Advisors who arrive with a model to install are recalled; the ones who succeed leave behind colleagues who no longer need them, which is written into their terms of reference as the objective.

Peer exchange exploits a truth consultancies cannot replicate: authority transfers between peers. A deputy governor will hear from another deputy governor — who has run the same crisis with the same political constraints — what no external expert could tell her. The coalition’s role is logistical and financial: matching institutions, funding the secondment, and staying out of the room.

The structured programmes carry the volume — over 1,300 officials to date — and their distinguishing feature is the examination. Attendance certificates are worthless as signals and everyone in the industry knows it; coalition programmes are examined, failure is possible, and the pass is consequently worth listing in a career. Cohorts mix institutions across regions deliberately, seeding the professional network that later makes peer exchange self-organising.

• Resident advisors — senior specialists embedded in the counterpart institution for one to three years, working the actual desk, not presenting slides
• Peer exchange — secondments between member institutions, because a central banker learns fastest from another central banker
• Structured programmes — multi-year curricula in supervision, treasury, statistics, and procurement, co-taught with member institutions and examined, not merely attended

05 Sequencing: capacity with the capital

The industry habit is to finance the asset first and train people later, which is how assets fail: the operating institution meets its power plant the way a new driver meets a truck already rolling downhill. The coalition inverts the sequence structurally, not rhetorically.

The capacity plan is part of underwriting. Before approval, the Investment Committee sees who will operate the asset, what skills exist today, what the gap-closing plan is, and what it costs — the plan sits in the credit file next to the cash-flow model, weighted in the durability score. A commitment with excellent economics and no credible operator scores a durability one, which is a decline.

Disbursement then ties to institutional milestones as well as physical ones. A tranche may release on the utility certifying its maintenance team, not merely on the contractor pouring concrete. By the time capital is fully deployed, the institution that must carry the asset already can — that is the design, and the disbursement schedule is where it is enforced.

06 The curriculum, concretely

Abstractions about capacity are cheap, so here is what an engagement actually contains. A typical central-bank modernisation runs eighteen to thirty-six months across four streams — monetary-policy operations, supervisory frameworks, payment-system oversight, and data infrastructure. Each stream carries a resident advisor, a peer-exchange partner institution, and a defined exit exam.

The exit exam is set by the institution itself, with the coalition holding the standard: operate the function for two consecutive quarters without external support, at defined quality thresholds — supervision cycles completed on time, settlement incidents below target, statistical releases on calendar. The institution graduates itself; the coalition merely verifies.

Delivery-institution engagements mirror the structure for different content: procurement integrity, maintenance regimes, programme management, and community-facing delivery. A utility engagement, for instance, pairs its resident engineer-advisor with a maintenance-planning curriculum and a twinning arrangement with a member utility two development stages ahead — close enough to be credible, far enough to be worth following.

07 What it costs and who pays

Capacity work is funded by a grant component sized inside each commitment — typically two to five percent of commitment value — for two reasons that took the industry decades to learn. Charging struggling institutions consulting fees defeats the purpose: the institutions most in need are precisely those least able to pay. And burying the cost in general overheads hides it from governance: inside the deal, the Investment Committee sees exactly what institution-building each commitment carries and can weigh it against the durability score it supports.

The Preparation Fund finances capacity work that must precede a commitment entirely — the readiness engagements covered on its own page. And a small central budget funds the cross-cutting programmes, examined curricula, and the rotation infrastructure that no single commitment could justify but every commitment draws on.

For scale: across the current portfolio the embedded capacity components sum to a nine-figure programme — one of the largest institution-building efforts in development finance, financed without a single standalone appeal, because it rides inside the capital it protects.

08 Measuring it: the exit test

Capacity is measured by independence, not attendance — the sentence is repeated deliberately, because it is the whole methodology. Every engagement defines its exit test at the start: the function the counterpart must run alone, for how long, at what quality. Certificates, workshop counts, and advisor-days are recorded as inputs and carry no evaluative weight.

Verification is independent and delayed. The Independent Evaluation Panel re-inspects engagements two years after the advisors leave — long enough for the counterfeit version of capacity to have collapsed — and asks one question: does the institution still run the function, alone, at standard? The two-year verified pass rate is published in the annual report alongside the financial results, at the same level of prominence, because the coalition regards the two numbers as the same kind of claim.

Failures are published too, with cause. The commonest is counterpart churn — the trained team promoted away or poached — which is why engagements now contract for named-role continuity and why the structured programmes deliberately over-train beyond the minimum team. The methodology improves the way the portfolio does: by publishing what went wrong.

09 The failure modes we design against

The coalition studied why capacity programmes fail before building its own, and the design answers each documented failure mode explicitly. The list is kept in the open because every one of these failures still walks the industry, wearing new acronyms.

The parallel-unit prohibition deserves the elaboration, because it is the most tempting failure of all. Donor-funded project units — better paid, better equipped, reporting to the funder — deliver beautifully for exactly as long as the funding lasts, while hollowing out the ministry next door by hiring away its best people. The coalition’s rule is absolute: the work happens inside the real institution, at its pay scales, in its reporting lines, however much slower that is. Slow and real compounds; fast and parallel evaporates.

• Fly-in training — replaced by residence and peer exchange; nobody learns treasury from a hotel ballroom
• Counterpart churn — engagements contract for named-role continuity, and programmes over-train beyond the minimum team
• Donor dependence — budgets taper on a published schedule; the institution absorbs costs as it absorbs skills
• Parallel units — the coalition strengthens the real institution, never builds a shadow one beside it
• Template curricula — every engagement is scoped from the institution’s own gap analysis, not a global syllabus
• Unmeasured outcomes — the exit test, verified independently at two years, is non-negotiable

10 Where it connects

Capacity development is the thread through the coalition’s entire method rather than a department off to the side. It is stage three of Our Approach, delivered in step with the capital. It is the groundwork the Preparation Fund finances when readiness must precede commitment. It is the durability test inside the underwriting guidelines — the reason a deal with no credible operator is a decline, not a discount. And it is what makes the exit question — does it run without us? — answerable with yes.

The engagements themselves are visible in the portfolio under Commitments, tagged by institution type and stage, and the programme’s aggregate results are published annually beside the financial statements. For the roles that do this work, the Capacity Development office currently lists positions under Join the Fraction.