Beyond the award. Follow the business.

Saint Lucia Policy Analysis

Small business / productive capability / public value

Beyond the award.
Follow the business.

A grant, a new machine or a training place can help. The public question is what the business can do afterwards, what still holds it back, and how the next support should change.

Test the headline Read the paper (PDF)

Saint Lucia Policy AnalysisEdited by Kevin L. Michel · 21 October 2026

Three questions after support

01
Support received
What reached the firm?
Award, funds available, equipment or assistance
02
Productive use
What can it now do?
Operate, fulfil an order, keep records, get paid
03
Business change
What improved, and for whom?
Comparable sales, cash, work and resilience

Let the finding change the next support.

Give the action an owner, a deadline and a recheck.

SLPA’s proposed evidence-to-action sequence. Each step needs its own proof; these are not measured conversion rates.

01 / The finding in a minute

Count the support.
Understand the result.

Saint Lucia has financed micro, small and medium-sized enterprises (MSMEs), offered training and specified monitoring work. The next useful step is a clear account of the same firms over time: what they received, what became operational, what changed and what action followed.

Keep the business in view.

An award and a firm are different units. Follow repeat recipients without double counting, retain closed businesses and show where outcomes remain unknown.

Ask what became possible.

Did equipment work, an order get fulfilled, a record become usable or a customer pay? A practical capability connects finance with the business’s daily work.

Close the action.

A finding needs an owner, a suitable response and a recheck. The value of evidence includes a repaired process or better support while it can still make a difference.

The document behind this contribution

This study develops ideas from Kevin L. Michel’s Lean MSME Evidence Architecture, a July 2026 executive brief and technical annex in the original SLPA source register. Those independent designs used synthetic demonstrations. Here, the public question is developed through fresh official evidence and SLPA’s policy methods.

02 / What the public record establishes

Build on the progress.
Make the next result clearer.

Administrative reach, programme design, independent evaluation and a monitoring plan answer different questions. Reading them together gives a stronger starting point.

6 February 2026Official programme account

Finance has reached a substantial group.

Commerce reported access to finance for more than 700 MSMEs and announced marketing and digitalisation training. This establishes reported reach and intended assistance; a firm-level result requires further evidence. 1

21 May 2026Independent evaluation

Progress and capability gaps coexist.

The CDB country evaluation records contributions to MSME capability and financial resilience. It also identifies business-management and application weaknesses, constrained government support capacity and gaps in portfolio outcome monitoring. Its scope covers earlier activities through December 2024. 4 5

10 June 2026Programme design

The support mix changed.

The new EC$1 million grant-only facility allowed previous Loan-Grant recipients to apply. Its June application window has passed. Distinguish this funding window from earlier support; link any confirmed subsequent award to the same enterprise where appropriate. 2

July 2026Monitoring specification

Follow-up is already part of the task.

The public TOR specifies distinct cohorts, field verification and business-performance monitoring, with an August 2026–January 2027 schedule. Its objective of at least 5% growth in productive and/or sales output is a target to examine. This review did not establish contract award, actual fieldwork or results. 3

Read the cohort and interpretation notes

The TOR lists 735 first/second-call awardees, 60 extended-phase beneficiaries and an estimated 150 grant-scheme awardees, with different coverage requirements. These are scope inputs. Reconcile enterprises, people, awards, duplicates and overlap before publishing a combined unique-business total. Preserve the contract’s full requirements.

The CDB evaluation is a wider, mainly qualitative contribution assessment, with stated data limitations. It supplies meaningful findings without providing a current grant-cohort impact estimate. No national survival, profit or job-creation rate is inferred here.

03 / Inside the business

The oven is installed.
Can the bakery earn reliably?

Consider an invented small bakery. A new oven works and sales rise. Flour and fuel cost more, some customers pay late, and the owner works longer. The business may have greater capacity while its cash position remains tight.

Follow a useful function.

Compare output in consistent units, similar trading periods, input costs, payment timing and paid and unpaid hours. Record interruptions, maintenance and repeat orders. An increase in sales value can reflect prices; productivity needs an output-and-input measure.

SalesWhat was soldMarginWhat remains after relevant costsCashWhat is available when bills fall due

For a cooperative or service firm, choose a function suited to its work. Keep a short common core, then add only the measures needed for its particular constraint.

The equipment is idle

What prevents productive use?

Check installation, power, suitable inputs, operator skill and customers. The next useful intervention may be a repair or connection, rather than another purchase.

Sales rise, cash tightens

Where does the money go?

Examine prices, input costs, margins, invoice collection and repayment dates. More turnover can coexist with a business that cannot pay its next bill.

Records are incomplete

What can be established now?

Keep the outcome unknown where necessary, help the firm keep proportionate records and revisit. Separate a documentation gap from a conclusion about performance.

The business has closed

What happened, and what remains?

Retain the firm in the cohort. Understand timing, reasons, remaining obligations and the owner’s experience. A closure is part of the account, even when it complicates the story.

04 / A headline worth checking

Before the percentage,
ask “of which businesses?”

A comparable observation may be missing without the business disappearing. OECD’s evaluation guidance reinforces clear objectives, sales, employment and survival measures, and attention to firms that do not survive. 6

Invented example / 20 businesses / no programme data

The same businesses.
Different questions.

Imagine a cohort with 12 comparable sales increases, four comparable observations showing no increase, and four businesses without comparable observations. Choose the question you want to answer.

All 20 businesses stay in view.

12 Sales increased
4 Sales did not increase
4 No comparable observation

75%

12 ÷ 16 comparable observations

Sales increased in 12 of the 16 businesses with comparable observations.

This describes the observed group. Four businesses still lack a usable comparison. Keep them visible in the cohort account and follow-up queue.

Next action: investigate the missing comparisons and report how those businesses differ.

All three interpretations

75%: 12 ÷ 16 comparable observations. Sales increased in 12 of the 16 businesses with comparable observations.

This describes the observed group. Four businesses still lack a usable comparison. Keep them visible in the cohort account and follow-up queue. Next action: investigate the missing comparisons and report how those businesses differ.

60%: 12 ÷ 20 businesses in the cohort. A sales increase is documented for 60% of this cohort.

Another 20% have comparable observations showing no increase; 20% remain unknown. The documented share is not a complete programme success rate. Next action: publish the three groups together, with evidence dates and missingness reasons.

Still open: What would have happened without the support?. These counts cannot establish how much change the programme caused.

A credible comparison must address selection, seasonality, prices, other support and effects on nearby firms. Before-and-after records alone do not answer that question. Next action: design an appropriate evaluation before claiming attributable impact.

This illustration teaches denominator choice. It estimates neither Saint Lucia’s programme performance nor the reliability of any actual sample. “Comparable” also requires a stated period, consistent measure and adequate evidence.

05 / The public results statement

Evidence should lead
to a useful response.

Publish a concise aggregate account that a reader can follow from the cohort to the decision. Keep protected case details inside the authorised process. Reuse existing records where permitted and measure the time owners spend reporting.

Who and what?: Define the award cohort and support type. Link a stable enterprise identifier to separate people and awards. Show award, disbursement and productive-use dates.
What happened to the firm?: Report operating, temporarily paused, closed and unconfirmed status. Show attempted contact, completed assessments and reasons for missing observations.
What changed?: State the baseline, follow-up period, units and comparable denominator for each outcome. Separate paid and unpaid work, hours and seasonal jobs. A supported job is not automatically a newly created job.
How strong is the evidence?: Label documentary or observed evidence, corroborated reports, self-report and unknowns. Keep reasons such as missing, declined or not applicable distinct. Retain corrections and release versions.
What did the result require?: Count programme resources, business-owner time, assistance and maintenance. Cost per assessment or resolved action answers a different question from cost per additional job.
What changed next?: Show actions due, completed and overdue, their responsible roles, the evidence of closure and any recheck. Record whether referred assistance actually became available.

Evidence to check

Verify the claim.

Resolve a material discrepancy with appropriate corroboration and human review.

Business support to arrange

Address the constraint.

Confirm a suitable adviser or service, capacity to receive the case and the follow-up.

Records to repair

Make the account usable.

Correct a duplicate, inconsistent date or missing field while preserving the history.

These are separate work queues. A need for assistance does not decide a compliance case. Any adverse decision belongs to the applicable authorised process, with evidence, review and remedy.

06 / RIPPLE-4

Look beyond
the supported firm.

Follow the proposed change through four causal orders. The pathways below are hypotheses to investigate, not forecasts of programme effects. Their timing can overlap. 8

Direct incidence

Finance and time both count.

The firm gains capital or assistance. The owner, field officer and adviser also spend time. Track what became usable, the full resource cost and which constraint remained.

Who received support, when did it become usable, and what work did reporting displace?

Participant adaptation

Measures shape behaviour.

Clear feedback can encourage better records. A narrow sales target can also favour easy cases, reward price increases or discourage owners from reporting difficulty.

Are selection, missingness and negative experiences visible alongside the headline?

System propagation

A stronger firm affects others.

Orders may reach suppliers, imports may rise and credit needs may change. Several programmes can draw on the same public staff. A supported firm can also take customers from an unsupported neighbour.

Did capability or total demand grow, or did existing local business shift between firms?

Inherited state

What lasts after the support?

Usable skills, portable records and trusted relationships can endure. So can debt, maintenance liabilities, an unsupported digital tool or dependence on repeated awards.

Can the business maintain the capability, recover from disruption and choose its next step?

A better firm result can still leave a national question.

Growth may come from new capability, a tourism season, higher prices or demand moving from another local business. Trace household income, firm costs, government resources, external payments, finance, institutional trust and natural or infrastructure assets. A positive recipient story alone cannot settle the whole public-value account.

07 / Institutional Intervention Readiness

Improve the follow-up.
Protect the useful functions.

SLPA proposes a targeted redesign of the evidence-to-action link. Preserve legitimate verification, trusted support relationships and existing contractual duties. Assess the five readiness gates separately for the proposed workflow. 7

Bounded causal account

Define the firm’s constraint and the programme mechanism. Establish a baseline, exposure period and outcome. Separate evidence for a management action from evidence for a causal impact claim.

Conditional

Legitimacy and rights separation

Confirm lawful access, clear notice, fair selection, confidential handling and a correction route. Keep support requests, optional publicity and adverse compliance decisions in their proper processes.

Open

Dependency and alternative symmetry

Compare the proposed follow-up with the strongest feasible alternatives using the same cost, burden, rights and learning horizon. Map reliance on field staff, advisers, connectivity and existing records.

Conditional

Function coverage and transition integrity

Assign owners for assessment, support, verification and unresolved cases. Protect current services and contractual coverage. A new reporting form cannot substitute for available assistance.

Open

Learning and recovery

Agree correction, version control, review dates and authority to pause or redesign. Retain recoverable files and fund later follow-up before expanding the workflow.

Conditional

Disposition: prepare and rehearse, then release a bounded operational test only when authority, rights, resources and continuity are secured. These are separate requirements for an SLPA proposal, rather than an official readiness rating.

Compare the strongest feasible alternatives

A one-off compliance check

Can establish a defined use-of-funds question with limited repeated burden. It provides little evidence about the business’s later trajectory.

Repeated reporting from every firm

Offers broad coverage if well resourced. Repeated demands can strain owners and staff, and complete forms may still contain incomparable measures.

A lean follow-up panel with targeted depth

Can combine a short common core with deeper review where needed. Representative estimates require a proper selection and nonresponse design; risk-selected cases need separate treatment.

A complementary impact-evaluation layer

Addresses the distinct question of attributable change through a credible comparison. It complements routine monitoring and needs suitable design, data, timing and resources.

08 / Sovereign Option Theory

Earn the next commitment
with a working cycle.

The mission is durable business capability and better public decisions. Protect business continuity, rights and support capacity. Start with preparation and a small probe; let evidence determine a wider pilot, modular scale and eventual institutionalisation. 9

Weeks 1–2
Reconcile and prepare.
Choose one award cohort. Map existing records, forms, contracts and assistance routes. Agree measures, timing, access, correction rules and full costs. An authorised Commerce/SBDC programme owner would confirm responsibilities with SLDB where relevant and any appointed evaluation team.
Rehearse with invented records while critical gates remain open. Integrate the work into the existing authorised process where its scope permits.
Weeks 3–4
Test the workflow on 12 cases, if ready.
SLPA proposes a small operational probe spanning different record conditions and operating states. Test contact, evidence review, owner burden, referral capacity and recheck. Twelve is a design choice for learning how the workflow functions.
This probe is not representative and does not replace the TOR’s required coverage. Fieldwork begins only with authority, funding and resolved critical safeguards.
Weeks 5–6
Repair, recheck and decide.
Resolve inconsistent denominators, examine overdue actions and confirm whether assistance arrived. Publish a safe aggregate learning record. Fix or stop a weak process before wider follow-up; release the next stage against actual staff and support capacity.
The first cycle tests a way of working. It cannot establish long-term business impact in six weeks.
After the first probe assessment
Return at three and six months.
Schedule follow-up three and six months after each firm’s first probe assessment, with a 12-month same-season check where relevant. Preserve any historical pre-support baseline separately and record time since disbursement. Track closures, unknowns and support needs; fund later work from the start.
Retain portable definitions, records and local staff capability so the responsible institution can adapt its tools and delivery arrangements.

Viability and cost

Fund the whole loop.

Include staff, field travel, devices, data, secure hosting, support referrals, owner time, verification, independent review, maintenance and later follow-up. Match the scope to funded capacity. A low reporting bill can conceal an expensive burden on businesses.

Sensing and responsibility

Give each signal a decision owner.

The programme owner carries overdue assistance and scope decisions; the evidence lead carries comparability and correction; the responsible safeguarding role carries rights and data incidents. Confirm these roles and authority before launch.

No lawful access or accountable owner

Hold fieldwork. Resolve the authority, access and responsibility before releasing a case.

Incomparable outcomes

Hold the affected metric and repair the comparison. Keep the business in the cohort and continue appropriate support.

Assistance remains unavailable

The programme owner restores receiving capacity and limits further reporting or referral burden. Recheck before expanding.

A rights, data or repeated workflow failure

Contain affected processing, use the applicable remedy route and redesign before restart. Preserve lawful essential duties.

Keep one PITONS decision record

Record the problem and public purpose; intervention readiness; trade-offs and total consequences; options under uncertainty; navigation to functioning service; and scrutiny, evaluation and learning. Include the strongest rejected alternative, unresolved evidence, dissent and what would change the decision. 10

Keep three gates separate: L, legality, rights and legitimacy; R, resilience and system coupling; F, future options, recovery and learning. A favourable result on one cannot compensate for a failure on another.

09 / Paper, evidence and connections

Make support useful.
Make the result intelligible.

The contribution is a practical public account: a defined cohort, a useful business function, comparable evidence and a response that is carried through. It gives the next funding and support decision something firmer to rest on.

Research paper / 21 October 2026