1,005 explicit concession instruments and 12 firm-specific status instruments.
Published researchKevin L. Michel · 25 August 2026
Publication dataset and policy report · 2004–2025
Incentives by OrderA Census of Published Firm-Specific Tax-Concession Instruments in Saint Lucia, 2004–2025
Abstract
This study constructs a census of published, firm-specific tax-concession instruments in Saint Lucia from 2004 through 2025. It enumerates the National Printing Corporation’s annual statutory-instrument indexes, downloads and hashes a broad candidate corpus, classifies instruments using fixed inclusion rules, and reconciles the resulting inventory against regular and extraordinary Gazette evidence. The headline population comprises 1,017 core instruments: 1,005 with an explicit concession effect and 12 whose tax consequences are activated through a firm-specific statutory status. Forty-one duty-free-premises authorizations are reported separately. Gazette reconciliation adds 41 core instruments absent from the annual indexes, a 4.2 per cent increase over the index-only baseline. Tourism-linked regimes account for 701 core instruments (68.9 per cent), while grants account for 927 (91.2 per cent). Body-level evidence is available for 984 core records (96.8 per cent); 33 are supported only by official publication listings. Among 979 non-anomalous core records with made and publication dates, the median publication lag is 17 days. The paper treats instrument counts as measures of legal-publication activity, not fiscal value. It proposes a machine-readable public incentives register, annual archive reconciliation, standardized order content, and administrative-data linkage as foundations for tax-expenditure reporting and evaluation.
The evidence in one minute
The published legal record is large, concentrated and reconstructable.
The study counts published legal instruments. It does not estimate whether relief was used, what it cost, or whether it produced investment and jobs.701 core instruments across five tourism-related legal routes.
927 core instruments were coded as grants.
Duty-free-premises authorizations are retained but kept outside the headline core.
Regime concentration
Tourism-related laws account for more than two thirds of the instrument count.
This is a composition of publication events, not tourism’s share of revenue forgone. A narrow one-year concession and a long tax holiday each count once.Counts describe legal-publication activity. Fiscal exposure can only be estimated after linking the published record to tax and customs data under an explicit benchmark.
115 core instruments, including 105 tied to Cricket World Cup accommodation incentives.
The largest single regime, representing 35.0% of the core census.
The second-largest regime, representing 24.2% of the core census.
Tourism Development instruments appeared within the 2024–2025 end of the study window.
Index and Gazette reconciliation
One official web index did not reproduce the full recoverable population.
Gazette reconciliation recovered 41 numbered core instruments beyond the annual statutory-instrument index. Each had official publication evidence; the finding is about information-system coverage, not secrecy or legal defect.core records recovered from the annual statutory-instrument index pages.
a 4.2% increase over the index-only baseline.
984 core instruments have a full or official consolidated body.
core records have official publication listings but no located operative body.
What the reconciliation proves
- A dual-source method materially improves recovery.
- A maintained register can reduce repeated forensic reconstruction.
- Original records, corrected metadata and a public correction log can coexist.
What it does not prove
- An index omission does not mean an instrument was secret.
- A listing-only record is not evidence of misconduct.
- Residual undercoverage may remain because the source systems are not a perfect expected-population manifest.
Made-to-publication lag
Among 979 core instruments with usable, non-anomalous dates, the median publication lag was 17 days.
Timing is a transparency and administrative-control signal. Defined provisions allow retrospective operation in some circumstances, so delay alone does not establish illegality or impropriety.core instruments with usable made and publication dates.
with an interquartile range of 11 to 27 days.
230 of the 979 non-anomalous dated records.
11 records; the observed non-negative range was 0 to 376 days.
SLPA implementation proposal
Build one public register, one protected monitoring layer and one evidence-led review cycle.
This operating design develops the paper’s recommendations. It is an SLPA proposal, not announced Government policy. The time targets are indicative and depend on staffing, legal review and source validation.Public legal register
Publish each instrument, beneficiary and project as written; legal authority; relief stated; made, publication, commencement and term dates; conditions; status; and every amendment or revocation link.
Protected administrative record
Use the stable identifier to connect actual claims, imports, taxable income, investment commitments and compliance across Inland Revenue, Customs and responsible sector agencies.
Public accountability outputs
Release annual coverage and correction reports, tax-expenditure estimates under a stated benchmark, and proportionate evaluation findings while protecting confidential taxpayer information.
Assign ownership · indicative 30-day start
The ministry responsible for finance should name one accountable register owner, while legal-publication custodians and award, tax and sector agencies become named data suppliers.
Validate a baseline · indicative first 90 days
Use the 1,017 research records to seed a provisional baseline, then validate identity and current status against authoritative records. Retain the 41 duty-free records as a visibly separate series and publish the schema, update date, versioned download and correction route.
Close source gaps · indicative 3–6 months
Seek bodies for the 33 listing-only records, replace 29 flagged truncated titles from operative citations and connect amendments and revocations to stable target identifiers.
Standardize future orders · indicative 3–6 months
Subject to each enabling statute and Attorney General or drafting review, define action-specific minimum fields. Grant orders should state the beneficiary, project, authority, relief, dates, applicable conditions, reporting agency and review point; later actions should cite the stable target identifier.
Link administration · indicative 6–12 months
Connect the public identifier to protected tax, customs, investment and sector records under lawful access and confidentiality rules. The public register remains the crosswalk, not the fiscal estimate.
Evaluate by exposure · renewal or sunset
Prioritize reviews by estimated fiscal cost, duration, renewal history, policy importance and available outcome data, not by the raw count of instruments.
Evidence boundary
The census establishes the publication layer, not fiscal cost, compliance or impact.
The limits narrow the claim and identify the administrative data needed for the next stage.Revenue forgone is not measured
Instrument counts cannot substitute for actual claims, a benchmark tax system and a declared estimation method.
No causal outcome estimate
The paper does not establish investment, employment, output, spillovers, value for money or the counterfactual without relief.
No allegation by counting
The census does not support conclusions about fairness, favouritism, impropriety or corruption.
Detailed aggregates withheld
Rate, duration, condition-clause and unique-beneficiary totals require further instrument-level verification.
Michel, K. L. (2026). Incentives by Order: A Census of Published Firm-Specific Tax-Concession Instruments in Saint Lucia, 2004–2025. Saint Lucia Policy Analysis. https://doi.org/10.5281/zenodo.22094482