How money reaches a firm from a multilateral agriculture loan

Local firms lose multilateral work for a reason that has nothing to do with capability. They pitch the funder. The funder does not hire anyone. On a sovereign loan, the borrowing government hires, through a project management unit, under rules the funder wrote but does not administer. This note sets out the chain, the documents that matter, and the two numbers that decide whether a local firm competes alone or partners.

The one structural fact

All three of the major agriculture financiers say the same thing in their own documents, and they say it plainly.

The World Bank’s guidance for borrowers splits the roles explicitly: the borrower prepares the procurement strategy and plan, invites offers, evaluates them, awards and signs the contract. The Bank reviews, approves and advises. The Bank does not sign contracts with consultants.

IFAD is blunter. Its Project Procurement Guidelines require the borrower to “take full responsibility for the implementation of project procurement activities”, and state that IFAD’s review “does not exempt borrowers from their obligations”. The Project Procurement Handbook adds an indemnity holding IFAD and its officials harmless from claims arising out of project procurement.

The African Development Bank’s own consultancy guidance states that “the borrower selects and contracts with consultants”, and that the executing agency administers the process and the award.

So the client is the project management unit. Everything else in this note follows from that.

The chain, from board to bank account

1. Board approval. The financier’s board approves the loan or credit. This is the moment the press release appears. No procurement has happened.

2. Financing agreement. Signed between the financier and the sovereign, usually through the finance ministry. World Bank disbursement letters for Kenyan credits are addressed to the Cabinet Secretary at the National Treasury. The Treasury signs. A line ministry implements. Those are different offices, and the gap between them is where projects lose time.

3. Effectiveness. Conditions precedent are met. Until effectiveness, nothing can be spent. Kenya’s NARIGP took from August 2016 to July 2017, about eleven months. A later restructuring that moved the project from one ministry to another cost a further twelve months on its own.

4. The PMU is established, and permission to spend arrives. IFAD’s own country strategy evaluation for Kenya identifies the bottleneck that sits between effectiveness and any firm being paid: “the slow issue of Authority to Incur Expenditures”. The same evaluation notes that staff capacity was insufficient because of high turnover and under-staffed project coordination units, and that projects able to reuse established management infrastructure avoided start-up delays. A brand new PMU is the slowest thing in the system.

5. Accounts open. A designated account is opened, typically at the central bank. To take a real example, Credit 7918-KE authorised three segregated accounts denominated in euro at the Central Bank of Kenya, with ceilings of EUR 3 million, EUR 3 million and EUR 2 million, monthly reporting, and a minimum application value of USD 10,000 for direct payment.

6. The procurement plan is agreed and published. More on this below. It is the single most useful public document in the whole chain.

7. Notices, proposals, evaluation, award, contract. And only then does money reach a firm.

How long is all that? There is no published dataset giving average elapsed time from board approval to first consultancy contract. What is published: about eleven months from approval to effectiveness on NARIGP; 13.67 months from approval to first disbursement as IFAD’s 2021 average, against a 12 month target and a 17 month 2016 baseline. Add procurement plan clearance, advertising, a minimum of 30 business days for international proposal preparation, evaluation, a 10 business day standstill and negotiation. Eighteen to thirty months from board approval to first substantive consultancy contract is a reasoned composite, not a published statistic. Plan cash flow on that basis.

How you actually get paid

There are four disbursement methods under World Bank investment project financing, and IFAD’s four are functionally the same:

• ◆ Advance, into the borrower’s designated account, from which the PMU pays you
• ◆ Direct payment, where the financier pays you at the borrower’s request
• ◆ Reimbursement, where the borrower pre-finances and claims back, IFAD requiring the claim within 90 days
• ◆ Special commitment, essentially a letter of credit

Under every one of them the PMU raises the withdrawal application. Its authorised signatories, its discipline on statements of expenditure, and the timeliness of its interim financial reports determine when you are paid. On IFAD projects the statement of expenditure threshold and the initial advance ceiling are set per project in the Letter to the Borrower. This is why late payment on multilateral work is almost never the funder’s doing, and why due diligence on the PMU’s financial management record is worth more than due diligence on the funder.

Where the work is actually advertised

One important correction first. UN Development Business closed on 31 March 2025, after 46 years. It is no longer a channel. The World Bank Beginner’s Guide, the IFAD handbook and the AfDB consultancy FAQ all still instruct borrowers to advertise there, so the funders’ own documentation is out of date on this point.

The live channels are:

• ▪ World Bank, Projects and Operations procurement, filterable, with email alerts
• ▪ IFAD, project procurement opportunities
• ▪ AfDB, general and specific procurement notices on afdb.org
• ▪ Kenya: the Public Procurement Information Portal at tenders.go.ke, and the e-GP system launched by the National Treasury in April 2025, where supplier self-registration is reported to be a condition of responding to tenders
• ▪ Uganda: the PPDA e-procurement portal, egpuganda.go.ug
• ▪ Tanzania: NeST at nest.go.tz. It replaced TANePS, which was retired for new tenders at the end of 2023. If your firm is still registered only on TANePS, you are not registered.
• ▪ Rwanda: Umucyo at umucyo.gov.rw

The AfDB matters more than its share of money suggests here, because its 2015 procurement policy makes the borrower’s own procurement system the preferred first choice. AfDB financed work is therefore more likely than World Bank work to run through the national portal under national rules.

The document nobody reads

The procurement plan is published, and under World Bank rules it is legally binding, incorporated by reference into the legal agreement. It names the assignments, their estimated values, the selection method for each, the evaluation approach and the timetable. It exists before any expression of interest is advertised.

That makes it the best open source business development intelligence available to a local firm. A firm that reads the procurement plans of every active agriculture project in its country knows what will be tendered, at roughly what value, by roughly when, six to eighteen months before the notice appears. Most local firms first hear about an assignment when the REOI runs in a newspaper, with fourteen days to respond.

The two numbers that decide your strategy

USD 300,000: the national shortlist ceiling. Under World Bank rules a consulting shortlist may be composed entirely of national firms where the assignment falls below a ceiling set in the project’s procurement plan. In both Kenyan agriculture projects we examined, NARIGP and KCSAP, that ceiling is USD 300,000. Below it, a Kenyan firm competes against other Kenyan firms. Above it, expect an international shortlist and start thinking about partnership. These ceilings are country and project specific, set by regional procurement managers, so read the plan for the project you are targeting rather than assuming the Kenyan number applies in Kampala or Dodoma.

USD 500,000: the prior review threshold. Under the Bank’s March 2024 procedure, consulting contracts above USD 500,000 in a high risk country require prior review. Substantial, moderate and low risk thresholds are USD 1 million, USD 2 million and USD 4 million. Kenya’s climate smart agriculture project applies exactly the high risk row.

Read those two together and the picture is stark. Almost every consultancy a local East African firm can realistically win is both nationally shortlisted and post review. The PMU decides alone. The financier sees the file later, on a sample basis. There is no appeal to Washington or Rome that will change an award.

What the PMU is actually scoring

There is a two stage asymmetry here that decides most outcomes, and very few firms organise around it.

At expression of interest, you are judged as a firm. Shortlisting criteria in the World Bank’s own sample format are corporate: core business, years in business, relevant experience, technical and managerial capability. AfDB shortlists on technical qualifications, capability and experience. Shortlisting carries no standstill period and no automatic right to a debrief, so a firm that is not shortlisted has no protected window and no explanation.

At proposal, you are judged on your people. Under the World Bank’s standard request for proposals, a full technical proposal is scored out of 100 with the firm’s own specific experience worth only 0 to 10 points. Methodology and work plan carry 20 to 50. Key experts carry 30 to 60, rising to 60 to 80 in a simplified technical proposal. Within the key expert score, “adequacy for the assignment” is weighted at 60 to 80 per cent. Participation by national experts scores 0 to 10. Minimum qualifying technical scores typically sit between 70 and 85, and quality and cost based selection commonly weights technical at 80 per cent.

So a local firm is filtered out on corporate track record and then wins on individual CVs. The practical implication: your corporate reference list is what gets you through the gate, and it must be built deliberately, in the right value bands, in the right sectors, with completion certificates you can actually produce. Your CVs are what wins the contract, and a generic biography scores badly against a criterion that is explicitly about adequacy for this assignment.

The preference trap

This one costs firms real money. National preference and set-aside regimes are creatures of national law:

• ▪ Kenya: AGPO reserves 30 per cent of procurement spend for enterprises owned by youth, women and persons with disabilities
• ▪ Uganda: 15 per cent of the annual procurement plan reserved for registered associations of youth, women and persons with disabilities, with relaxed turnover and experience criteria
• ▪ Rwanda: a 15 per cent margin of preference for Rwandan registered companies on consultancy services under Law 62/2018

Under World Bank rules, there is no domestic price preference for consulting services at all. The Regulations provide domestic preference for goods and for works only. The mechanism that helps national firms in consultancy is the national shortlist, not a price margin. A firm that builds its bid strategy around AGPO status and then meets a World Bank rules request for proposals will find that status buys it nothing on price.

Which regime applies is settled in the financing agreement and the procurement plan. Read those two documents before you decide how to bid. IFAD allows a margin of preference only exceptionally, with its agreement, and only under international competitive bidding, and it will not accept bidding conditions requiring mandatory joint ventures. AfDB’s guidance prefers regional and national consultants where qualifications are equal.

What gets a good firm thrown out

Verified grounds, from the rules themselves:

• ▪ Offering alternatives for a key expert position. One CV per position, or the proposal is rejected.
• ▪ Failure to confirm key expert availability before negotiations, or availability that lapses inside the validity period.
• ▪ Submitting more than one proposal, individually or as a joint venture partner. All of that firm’s proposals are disqualified.
• ▪ Subcontracting the whole of the services.
• ▪ Debarment by the financier.
• ▪ Conflict of interest through a business or family relationship with client staff involved in preparing the terms of reference, selection or supervision.
• ▪ In Kenya, an expired KRA tax compliance certificate. It is valid for twelve months and requires eTIMS compliance.
• ▪ In Rwanda, bidding outside the scope written on your registration certificate. Article 88 of Law 62/2018 is explicit, and it catches firms whose registration is narrower than their capability.

How a local firm should organise itself

Six things follow, and they are all organisational rather than technical.

1. Track projects, not tenders. Build a watchlist of every active and pipeline agriculture operation in your countries, and read each procurement plan as it is published and updated.

2. Get registered before you need to be. National e-procurement portal, tax compliance, sector registrations, audited accounts, professional indemnity. Every one of these takes weeks and none of them can be done inside a fourteen day EOI window.

3. Build the reference list on purpose. Take assignments that put completed contracts of comparable value in the sectors you intend to bid in, and collect the completion certificates at the time, not two years later.

4. Maintain an expert bench, not a CV folder. Named people, confirmed availability, CVs written against typical terms of reference rather than as biographies.

5. Decide the USD 300,000 question deliberately. Below it, bid alone and build a national track record. Above it, be a substantive partner. Note that shortlisted firms may associate with non-shortlisted ones provided the shortlisted firm leads, which is a genuine route in for a strong local firm that missed the shortlist.

6. Do diligence on the PMU. Its disbursement record, its audit history, its staff turnover and its payment behaviour will affect your cash flow more than the identity of the financier.

---

Sources

World Bank, Procurement Regulations for IPF Borrowers, 7th edition, September 2025 · World Bank, Bank Procedure: Procurement, March 2024, prior review thresholds · World Bank, Standard Request for Proposals, Consultants, January 2020 · Kenya NARIGP procurement plan and restructuring paper · IFAD, Project Procurement Guidelines · IFAD Independent Office of Evaluation, Kenya Country Strategy and Programme Evaluation · IFAD, Report on IFAD’s Development Effectiveness 2022 · AfDB, acquisition of consultancy services and the 2015 procurement policy · UN Development Business closure notice · PPRA Tanzania on NeST · PPDA Uganda on reservation schemes · Rwanda Law 62/2018 · AGPO Kenya · KRA tax compliance.

Thresholds and shortlist ceilings are project specific. The USD 300,000 figure is verified in two Kenyan agriculture procurement plans and should not be assumed for other countries or projects. The eighteen to thirty month estimate is a composite of published components, not a published statistic. Kenya’s e-GP mandate and its interaction with donor funded tenders should be confirmed against National Treasury Circular NT/PPD 1/3/14 Vol. VI and PPRA Circular 02/2025 before you rely on them.