Reported work / May to July–August 2020
37.9 → 47.2%A larger share reported work.
The national profiles show the share working in the previous week. Their survey population and weights need to stay attached to the figures.
Work / essential needs / dependable support
A job returns.
Has the household recovered?
A return to work is one sign of recovery. A household also needs steady resources for food and bills, and room to cope with the next setback. Saint Lucia’s historical evidence helps us ask the next questions.
01 / The finding in a minute
Work improved.
Household strain remained visible.
A two-round phone-survey study offers a historical case for reading work, food, modelled consumption poverty and assistance together. Each signal answers a different question. 1
Food / July–August 2020
25.7%Running out of food was reported.
The national profile asked whether any household member had experienced this during the preceding 30 days.
Modelled poverty / historical starting point
About 24%The baseline was reconstructed.
Point estimates rose to 33.25% in May and eased to 30.33% in July–August. The source contains two baseline versions.
Keep the dates and people clear. These are national 2020 aggregates. They show work improvement and food strain in the same body of evidence; the published data do not trace the same households through both experiences. Current conditions require fresh evidence.
02 / An interactive reading guide
Read the signals.
Keep their meanings clear.
Select a signal to see the historical evidence, its limits and the practical question to carry into a current review.
Historical evidence / 2020
Work is one part of the recovery story.
The national survey respondent profiles show 37.9% reporting work in the previous week in May 2020 and 47.2% in July–August. This is a useful sign of improvement across the two profiles. 1
Figures 4–5, page 9. Weighted respondent profiles; confidence intervals are not supplied. Source 1
Read it carefully
The figures describe weighted phone-survey responses. They are not an official labour-force employment rate, and the profiles do not establish which individuals returned to work. The report gives no confidence intervals for these shares.
What to check next
Ask about paid hours, take-home earnings, regularity and the costs of working. Track whether available resources cover food, travel, care and other essentials over a clearly defined period.
Has the household gained dependable resources after the costs of working?
Historical evidence / 2020
Ask what happened at home.
In the national round-two profile, 25.7% reported a household member running out of food, 17.7% being hungry without eating, and 12.5% going a whole day without eating during the previous 30 days. 1
Figure 3, page 8. Separate questions; experiences can overlap. Read each percentage on its own. Source 1
Read it carefully
These are separate household questions. Experiences can overlap, so adding the percentages would double-count people. The source does not fully document the household/person denominator and weighting for each plotted share, or provide a calibrated overall severity measure.
What to check next
Keep the same question wording and reference period in repeat monitoring. Follow food access alongside earnings and essential costs, with confidential routes to seek help when an urgent need is reported.
Are essential needs being met consistently as work and income change?
Historical evidence / 2020
Read the estimate with its uncertainty.
Annex 1 reports modelled poverty of 23.86% before COVID, 33.25% in May 2020 and 30.33% in July–August. The point estimates show a sharp rise followed by a smaller easing. 1
- Pre-COVID, reconstructed23.86%Reported 95% interval 18.62–29.10%
- May 202033.25%Reported 95% interval 26.40–40.11%
- July–August 202030.33%Reported 95% interval 23.66–37.01%
Read it carefully
The starting point was reconstructed from phone responses. Annex 2 gives a different pre-COVID estimate, 24.04%, without explaining the difference. Reported confidence intervals are broad, and the joint uncertainty needed to test changes is not supplied.
What to check next
Use quality-assured consumption evidence for a new population baseline. Require a documented model, weights, validation and uncertainty before presenting a short survey as a new poverty estimate.
What does the estimate measure, and how firmly can the change be established?
Historical evidence / 2020
Follow assistance through to use.
The 2020 round-two national profile reports cash transfers at 6.6%, food assistance at 8.5% and other in-kind transfers at 0.4%. The report says the survey preceded a later programme expansion. 1
Proposed reporting stages
- Assessed
- Waiting or approved
- Payment issued
- Receipt confirmed
- Service received
Use the stages supported by each programme’s actual records. Identify the responsible office and unresolved cases at every handover.
Read it carefully
The categories may overlap. Programme names, amounts, frequency and a clear recall window are absent. These shares cannot show whether support was sufficient, arrived on time or caused a change in household welfare.
What to check next
For today’s case review, distinguish a completed assessment, a budget-constrained wait, approval, payment and confirmed receipt. Track referrals through to a service received and record unresolved problems with a named owner.
Did the household receive the assistance or service it was meant to receive?
03 / The poverty estimate and its limits
A sharp rise.
A smaller easing in the estimates.
The study estimated consumption statistically from survey responses. Its pre-COVID comparison was reconstructed after the outbreak. Annex 1 supplies the following estimates and reported uncertainty. 1
Pre-COVID, reconstructed23.86%18.62–29.10% interval
May 202033.25%26.40–40.11% interval
July–August 202030.33%23.66–37.01% interval
The pre-COVID estimate is reconstructed. Annex 2 gives an alternative baseline of 24.04%. Testing change between periods requires their joint uncertainty. Source 1
Read the exact figures in a table
| Reference period | Point estimate | Reported 95% confidence interval |
|---|---|---|
| Pre-COVID, reconstructed | 23.86% | 18.62–29.10% |
| May 2020 | 33.25% | 26.40–40.11% |
| July–August 2020 | 30.33% | 23.66–37.01% |
Keep the alternative baseline visible.
Annex 2 reports 24.04%, with a 17.73–30.35% interval, for the reconstructed pre-COVID baseline. The report does not explain the difference. “About 24%” is the appropriate headline precision; exact figures retain their table labels.
Separate arithmetic from inference.
May to July–August falls by 2.92 percentage points in the published estimates. Establishing the precision of that change needs the joint uncertainty across periods. Overlapping intervals alone cannot settle that question. 1
Why the earlier poverty figures are kept separate
The first-round report’s 26.22% was a preliminary reconstruction of pre-COVID poverty. The later methodological experiment used a phone-owning subset of the 2016 Household Budget Survey. Those estimates have different roles. Multiplying either by the two-round paper’s rounded index would manufacture an unsupported rate. 23
The paper also contains ambiguous 85%/64% probability statements. Its probability calculation and one comparison are insufficiently specified. This feature uses the annex estimates with their published intervals and leaves those probabilities out of the recovery assessment.
04 / The decision point in October 2026
A current review
makes the question practical.
Saint Lucia is reviewing assistance records while developing its wider social protection system. This is an opportunity to make the path from changed circumstances to a useful response easier to follow.
PAP / effective 15 September 2026
Recertification is under way.
The Public Assistance Programme review covers current beneficiaries and waiting-list applicants. The notice also invites recipients suspended after the March–May review to participate. Home visits use the Saint Lucia National Eligibility Tool, SLNET. 4
Verified: announced review process
Government and WFP / 8 October 2026
Readiness is being assessed.
A validation workshop reviewed findings for a Shock Readiness Index baseline. The notice documents assessment activity. The full baseline, any numeric score and evidence of operating performance remain records to obtain. 6
Verified: assessment and workshop
Information system / announced May 2026
Follow the delivery timetable.
The planned Social Protection Information System covers intake, assessment, payment and frontline work. The government announcement expected operation in the first quarter of 2027. Service improvements could proceed once the responsible institutions confirm authority and capacity. 9
Status: expected future delivery
Population evidence / checked 20 October 2026
A new baseline needs a verified release.
The CSO describes the 2025/26 living-conditions and budget survey as in its final phase. New poverty results were not verified in this review. The October World Bank brief still identifies 2015/16 as the latest available poverty period. 78
Status: historical poverty reference remains
A missed contact
leaves a question open.
The ministerial statement recognises changing addresses, household composition and economic circumstances. Reporting should distinguish failed contact, pending evidence, budget-constrained waiting and a completed decision. A smaller recipient list can have several causes; the reason for each change matters. 54
05 / An honest source audit
Keep the useful finding.
Show how it was produced.
The supplied report was checked against the official PDF. Their bytes match. Every page was read and the figures and annexes were visually checked. The downloadable data transcribe published aggregates.
Telephone responses and modelled poverty do different jobs.
Households reported conditions by phone. The study then used a consumption model trained on earlier survey evidence to estimate poverty. Food and employment variables also enter the post-COVID model. Differences between model-poor groups therefore partly reflect the classification method, so they cannot independently validate it. 1
Coverage and changing samples need attention.
The first round had 1,093 respondents; the second had 900. Adjusted weights address observed imbalances under assumptions. The report lacks the full diagnostics needed to establish how unreachable households and nonresponse were represented. A later 2016-survey experiment supports the method’s promise, while leaving pandemic-period model stability to be tested. 13
Question labels can change the story.
Round one asked three food questions; round two asked eight. “Pre-COVID” in a food chart labels a poverty-classification group, rather than the date of a food observation. Employment stoppage categories also use different periods. This feature favours the national second-round food items and the more comparable worked-last-week question. 1
What a full replication would require
Request questionnaire versions, the variable dictionary, final survey weights and calibration targets, model specifications, imputation and variance procedures, and the intended baseline version. The exact denominator for each chart also needs clarification. The public report alone cannot reproduce the model or household transitions.
The document prints no publication date. The 2021 citation follows its PDF metadata; observations concern 2020. Its 2024 upload directory does not change either date. This research introduces no 2026 poverty estimate.
06 / From evidence to a useful response
Give each evidence stream
a clear job.
SLPA proposes a small recovery evidence process alongside existing work. Population research, repeat monitoring and individual case review need distinct purposes, access rules and outputs.
Population statistics
CSO’s quality-assured consumption and living-conditions evidence establishes population patterns, with methods and uncertainty. Aggregate findings can guide resources and outreach.
Output: Dated population estimates with clear definitions and publication controls.
Repeat recovery monitoring
A short, CSO-reviewed module could follow paid hours, dependable resources, food, essential bills and coping. Consistent wording and reference periods matter; report attrition and missing responses.
Output: Separate trends in household conditions between major surveys.
Programme case records
Record intake, evidence, decisions, waiting, payments, referrals and corrections. Individual decisions follow authorised programme rules, reasons and review arrangements.
Output: A dated service record showing the next action and responsible office.
Protect the purpose of each contact. A statistical interview should have clear confidentiality safeguards. Seeking assistance should have an accessible intake route. Any person-level sharing needs a verified lawful basis and explicit controls.
Start by inspecting what already exists.
The PAP notices do not publish SLNET’s full questions or scoring specification. Check the actual instrument before adding fields. Test which extra evidence changes a useful decision enough to justify the burden on households and staff. 4
Keep entry and correction open.
Include new applicants, changed circumstances and re-entry. For a proposed change linked to employment, examine earnings, work costs and continuing needs through the proper review. Separate contact problems from demonstrated improvement and give people usable reasons.
07 / Applying the SLPA frameworks
Diagnose. Trace. Test.
Then choose the commitment.
Use IIR to understand current functions, RIPPLE-4 to examine consequences, IIR’s five gates to compare readiness, and SOT to choose and stage the next step. 111213
The starting functions are reliable support, fair assessment, new intake, correction and trusted statistics. Compare a bounded rehearsal with finishing the current exercise first. If equivalent work is already under way, join it and use its evidence.
RIPPLE-4 / testable pathways
Follow the effects beyond a better form.
Direct effects
A clearer record may resolve a case faster. Extra contact also takes household and officer time.
People’s responses
Fear of losing support may change participation or answers. Test notices and assisted access.
Wider effects
New intake can increase demand for assessments, funding, payments and appeals. Follow every queue.
Lasting effects
Better methods can endure. So can sensitive data obligations and system dependence. Specify retention and recovery.
| IIR gate | Evidence required | Proposal status |
|---|---|---|
| Bounded causal account | Named workflow, intended decisions, selected cases and comparison with current work. | Conditional |
| Legitimacy and rights separation | Authority, lawful data use, clear notices, reasons, correction and review. | Open |
| Dependency and alternative symmetry | Staff, finance, contact and payment capacity; the same tests for continuation and change. | Conditional |
| Function coverage and transition integrity | Continuing payments and intake, offline access, funded responsibilities and a fallback. | Open |
| Learning and recovery | Failure signals, authority to pause, restore procedures, correction resources and review date. | Conditional |
SOT: make the first
commitment useful.
Begin with a shared evidence dictionary, dated case stages and a narrow rehearsal. Learn whether the bottleneck is contact, evidence, staff, funding or payment. Preserve working service routes and household options. Income can return after savings or productive assets have been used up. Check the household’s ability to cope with another setback. Count the harm of delay as well as the cost of action. 13
SLPA proposal / subject to lawful authorisationTest the path
08 / A 90-day evidence and reconciliation rehearsal
Test the path
through to usable support.
Use existing approved tools where they are sufficient. Keep ordinary assistance and intake running. The rehearsal should identify a practical repair and the evidence needed for the next decision.
- Days 1–30
Map current work and protect its capacity.
The programme owner maps review, new intake, decisions, payment and correction. CSO defines the statistical boundary; Finance identifies costs and funding limits. Verify lawful data use, staff capacity and existing equivalent work before selecting test cases.
Output: Scope, data dictionary, baseline queues, protected service plan and named owners.
- Days 31–60
Rehearse the exceptions.
Use synthetic or properly authorised records to test a wrong address, unavailable phone, assisted application, budget-constrained wait, failed payment and correction. Record staff time, errors and handovers. Test restoration and confirm how an unresolved case reaches its next owner.
Output: Case-stage trail, exception log, resource record and revised instructions.
- Days 61–90
Make a reasoned next-step decision.
Consider a tightly bounded live workflow only after its readiness conditions are met. Publish redacted findings and unresolved issues. Decide whether to continue, redesign or end the exercise, with reasons, responsible offices and a review date.
Output: A reviewable decision; a live component can remain deferred while identified repairs proceed.
Fund the work.
Protect continuity.
Cost officer time, backfill, assisted access, contact, statistical and legal work, secure handling, payment tests and error correction. Separate rehearsal costs from any future allocation for assistance. This proposal creates no appropriation or change in eligibility.
Set acceptance tests before live use. Pause the affected trial function if it threatens payments, privacy or fair handling, while lawful assistance continues. A May 2025 simulation already exposed contact and coordination lessons; use that record to avoid repeating work. 10
09 / Downloads, methods and sources
Read the paper.
Inspect the evidence.
The next contribution is a clearer connection between a dated signal, a fair review and a dependable response. SLPA’s wider Household Resilience proposal sets out the broader system.
Research edition / 20 October 2026
A job returns. Has the household recovered?
The paper contains the analysis and proposed rehearsal. The CSV and JSON contain published aggregate estimates with source references and audit notes. They contain no household microdata.
Poverty projections and profiling from St Lucia High Frequency Phone Surveys using a SWIFT-COVID19 package
Chen and co-authors, cited as 2021 from document metadata. Historical 2020 evidence: sampling p. 6; poverty p. 7 and Annexes 1–2, pp. 12–14; food p. 8; work p. 9; assistance p. 10. The 2024 upload folder is not the study date.
The Socio-Economic Impact of the COVID-19 Pandemic in Saint Lucia
CSO and World Bank, 5 August 2020. First-round survey. Annex A1, p. 24, labels its 26.22% poverty projection preliminary and pre-COVID.
Correcting Sampling and Nonresponse Bias in Phone Survey Poverty Estimation Using Reweighting and Poverty Projection Models
Zhang, Takamatsu and Yoshida, Policy Research Working Paper 10656, December 2023. Saint Lucia experiment uses phone owners within the 2016 HBS; it tests a different question from pandemic-period prediction.
Welfare Unit to reassess PAP beneficiaries as recertification begins in September
Government, 21 September 2026. Exercise effective 15 September; current beneficiaries, waiting-list applicants and participation by suspended recipients are described.
Ministerial statement on the 2026 Public Assistance Programme recertification exercise
Government, 28 September 2026. Addresses changes in contact details, household composition and economic circumstances, with home visits and public communication.
Saint Lucia and WFP assess social protection readiness for emergencies
Government, 8 October 2026. Validation workshop and Shock Readiness Index baseline work. The announcement supplies no numeric score or completed operational certification.
2025/26 Survey of Living Conditions and Household Budgetary Survey
CSO project page checked 20 October 2026. Describes a 1,524-household selected sample and planned May 2025–April 2026 fieldwork. Quality-assured new poverty results were not established by this page.
Saint Lucia Poverty & Equity Brief
World Bank, October 2026 edition verified on 20 October. Identifies 2015/16 as the latest period with available poverty data. The URL is mutable; preserve the edition date.
World Bank mission supports major modernization of Saint Lucia’s social protection system
Government, 18 May 2026. Describes intake, registration, assessment and payment reform, with an integrated information system expected in the first quarter of 2027.
Integrating Anticipatory Action for Hurricanes into Social Protection Systems: Saint Lucia Case Study
WFP, published 11 March 2026. Reports a May 2025 simulation, including contact, Treasury coordination and data-quality lessons.
Beyond Chesterton’s Fence: Institutional Intervention Readiness in Complex Policy Systems
Kevin L. Michel, 11 August 2026. Diagnosis, comparison, essential functions, transition integrity and learning/recovery gates.
Beyond the First Ripple: RIPPLE-4
Kevin L. Michel, 7 August 2026. Recursive Impact Pathways, Politics, Leverage, and Exposure.
The Sovereign Option: A Lewisian Theory of Decisive Action in Complex Policy Systems
Kevin L. Michel, 5 August 2026. Public purpose, usable options, commitment, learning and the costs of delay.