Introduction

Solomon Islands' tax system still runs on instruments most of its Pacific neighbours have moved past: a dual-rate Sales Tax Act from 1990 charging 10% on locally manufactured goods and 15% on imports, in place of a broad VAT, and a round-log export duty calculated not on the price the exporter actually receives but on a "determined value" the Minister of Finance sets by gazette order, last materially revised for inflation in 2010 and periodically thereafter. The gap between the two prices is not a technical footnote. Transparency Solomon Islands has stated plainly that exporters have reasons to want the determined price low — to cut the duty owed, to understate the income subject to business tax by structuring the sale through a related overseas buyer, and to reduce what must be shared with the customary landowners and agents who arranged the harvest — and that the losses involved have run to SBD 1-2 million a week for years.

That revenue matters more now than it did a decade ago, for two reasons. First, logging itself is declining: the accessible commercial forest is being exhausted, and the IMF has noted "declining log exports" as a structural driver of the country's fiscal and current-account deficits since at least 2024. Second, gold mining has grown quickly enough to become a named driver of 2025's above-forecast growth, and the IMF's own conditional language, that mining's expansion is an upside risk "if supported by robust resource management and limited fiscal incentives," is a direct warning that the fiscal-incentive and valuation problems documented for logs should not simply migrate to the new resource.

Both issues sit on top of a genuinely near-term liquidity problem. Public debt has nearly tripled since before the pandemic, government cash reserves have been significantly depleted, and the IMF's 2026 Article IV consultation flagged "significant budget implementation risks arising from depleted cash reserves, revenue uncertainty, and limited investor capacity to absorb government bonds" as an immediate, not eventual, concern. A VAT Bill that would replace the fragmented sales tax with a single 15% rate has been drafted and was expected to reach Parliament by April 2025, but remains undebated.

The research question this paper takes up is therefore threefold: how much of the log-pricing gap is recoverable with a defensible, data-driven benchmark; how the same discipline should be built into mining-sector valuation and incentives before the pattern repeats; and how Treasury should manage the next several budget cycles given depleted cash and a thin domestic bond market, while a VAT transition adds its own implementation risk on top. As in the other papers in this series, the underlying identity splits the revenue gap into a policy component and a compliance component, here applied specifically to the export-duty base:

R^{pot}_{export,t} - R_{export,t} = \underbrace{\left(R^{pot}_{export,t} - R^{law}_{export,t}\right)}_{\text{policy gap: an under-set determined price}} + \underbrace{\left(R^{law}_{export,t} - R_{export,t}\right)}_{\text{compliance gap: volume and grade mis-declaration}}

Current tax issues

2.1 A price set by order, not by market

The round-log export duty is charged at 25% of a "determined value" fixed by the Minister of Finance under the Customs and Excise Act, by species and grade, through a gazette order that is revised at irregular intervals — the schedule sat unchanged from 2003 until a roughly 30% catch-up revision, and further orders have followed in the years since. The duty is paid on this determined value regardless of the FOB price the exporter actually agrees with the overseas buyer. Transparency Solomon Islands has been explicit about why an exporter would want that determined value low: it cuts the export duty directly; it can be paired with an artificially low sale price to a related overseas buyer to shift income (transfer pricing) and reduce business income tax; and it reduces the amount the exporter must share with the customary landowners and the agents who arranged the harvest. TSI's own estimate of the resulting loss, SBD 1-2 million a week, has stood for years without an official rebuttal that this paper is aware of.

2.2 A structurally declining base

Log exports have been on a declining trend that the IMF has linked directly to Solomon Islands' fiscal and current-account deficits since at least its 2024 Article IV work, as accessible commercial forest is exhausted. Taxes on exports still made up over 12% of total tax revenue as recently as 2022 (World Bank data), so the decline is a first-order fiscal event, not a rounding error.

2.3 Gold mining, growing quickly, with the same risks unaddressed

Gold production and agricultural output were named by the IMF as the main drivers of 2025's above-forecast 3.5% growth, and further mining expansion is flagged as an upside risk for 2026-27 — conditional, in the Fund's own words, on "robust resource management and limited fiscal incentives." That conditional phrasing is a warning: nothing in the public record suggests Solomon Islands has yet built the valuation and incentive discipline for mining that it visibly lacks for logs, and a resource transition that inherits the old sector's governance gaps forfeits much of its fiscal benefit before it is collected.

2.4 Cash reserves depleted, debt nearly tripled, bond market thin

Public debt has nearly tripled since before the pandemic. The IMF's 2026 Article IV consultation names "significant budget implementation risks arising from depleted cash reserves, revenue uncertainty, and limited investor capacity to absorb government bonds" as a live concern for the 2026 fiscal year, not a medium-term projection. This is a materially different starting point from Fiji's paper in this series, where debt is elevated but reserves and market depth are not simultaneously named as an immediate constraint; here, cash management itself is the binding near-term problem.

2.5 A VAT transition that has stalled

A VAT Bill proposing a single 15% rate to replace the fragmented 10%/15% Sales Tax Act was expected to reach Parliament by April 2025; as of the most recent public reporting it remains pre-legislative. A VAT that eventually passes will need an implementation runway most administrations underestimate, and Solomon Islands' own delay is itself evidence that runway has not yet begun in earnest.

2.6 Political instability as a fiscal risk, not only a governance one

The IMF's 2026 assessment names "domestic political instability" alongside geopolitical tensions as a downside risk to the outlook. For this paper's purposes, the relevant implication is narrower than the general governance literature: instability raises the premium on revenue instruments that do not depend on sustained ministerial discretion being exercised well, which is the same argument this paper makes for replacing the determined-price schedule with a formula.

Models

3.1 Model A: log export reference pricing

Model A replaces the gazetted determined value with a benchmark built from data the ministerial schedule ignores: destination-market import statistics. China has for years been the dominant buyer of Solomon Islands round logs; Chinese customs publish import unit values by species-equivalent HS code and origin. For species s, grade g and month t, the reference value blends this mirror data with the peer distribution of Solomon Islands' own exporters, using the same robust construction as the customs models elsewhere in this series:

V^{ref}_{s,g,t} = \alpha \cdot \tilde{u}^{mirror}_{s,t} + (1-\alpha) \cdot \tilde{u}^{peer}_{s,g,t}, \qquad \tilde{u}^{peer}_{s,g,t} = \mathrm{median}\left(u_{i} : i \in \text{exporters of } s,g \text{ in } t\right)

alpha is set higher where mirror coverage is reliable (a small number of destination markets take most volume) and lower for species with thin or diverted trade. The duty-at-risk from using the old gazetted value V-det instead of the reference value:

Y_{s,g,t} = \tau_{export} \cdot Q_{s,g,t} \cdot \max\!\left(0,\; V^{ref}_{s,g,t} - V^{det}_{s,g,t}\right)

where Q is exported volume in cubic metres and tau-export is the 25% duty rate. Publishing V-ref alongside V-det, updated quarterly rather than by irregular ministerial order, converts a discretionary act into a data update — the same shift from officer discretion to published score that Tonga's paper in this series makes for exemption certificates, applied here to a ministerial rather than an officer-level discretion.

3.2 Model B: resource-revenue transition and incentive-leakage score

Model B tracks the fiscal handover from logging to mining explicitly, so the transition is measured rather than assumed, and scores mining agreements for the same leakage patterns documented in the log trade. The revenue-mix shift:

\text{ResourceShare}_t = \frac{R_{log,t} + R_{mining,t}}{R_{total,t}}, \qquad \text{TransitionRatio}_t = \frac{R_{mining,t}}{R_{log,t} + R_{mining,t}}

tracked over time to show whether mining revenue is genuinely replacing logging revenue or merely offsetting its decline in nominal terms while the resource-dependence share stays flat or rises. Each mining agreement or concession is then scored on the same three risks documented for logs — under-valued export/sale price, related-party sale structuring, and fiscal-incentive generosity relative to comparable Pacific and Melanesian jurisdictions:

I_m = \beta_1 \cdot \text{RelatedPartySaleShare}_m + \beta_2 \cdot \max(0, \text{IncentiveGenerosity}_m - \text{PeerBenchmark}) + \beta_3 \cdot \mathbb{1}[\text{no independent price verification clause}]

A high I_m does not imply wrongdoing; it flags a concession or exporter for the same reference-price and related-party scrutiny Model A applies to logs, before the pattern has years to compound the way it has for timber.

3.3 Model C: cash-buffer liquidity early warning and VAT-transition readiness

Where Fiji's Model C (this series) manages a buffer against infrequent disaster shocks, Solomon Islands' near-term problem is ordinary cash-flow risk against depleted reserves and a bond market too thin to absorb a shortfall. A rolling liquidity indicator, updated weekly from Treasury's own cash position and known near-term obligations:

LQ_t = \frac{\text{CashOnHand}_t + \text{ExpectedRevenue}_{t:t+k} - \text{CommittedExpenditure}_{t:t+k}}{\text{CommittedExpenditure}_{t:t+k}}

over a forward window k (illustratively four to eight weeks), with a bond-market absorption constraint layered on top: planned domestic issuance should not exceed a share of recent auction bid-to-cover ratios, since issuing into a market that cannot absorb it either fails outright or crowds out the banking sector Solomon Islands' small financial system can least afford to lose:

\text{MaxIssuance}_t = \gamma \cdot \overline{\text{BidCover}}_{t-12:t-1} \cdot \text{AverageAuctionSize}_{t-12:t-1}

Alongside this, a VAT-transition readiness checklist tracks the specific implementation prerequisites — registration-threshold design, input-credit system build, refund-processing capacity, and a public education timeline — against the Bill's own provisions, so that if and when it reaches Parliament, the administrative runway is not starting from zero.

Algorithms and illustrative results

4.1 Algorithm 1: log reference-price update and duty-at-risk

QUARTERLY:
FOR each species s, grade g:
  u_mirror <- median destination-market import unit value for s (China customs data,
              adjusted for species-equivalent HS mapping), trailing 3 months
  u_peer   <- median FOB unit value declared by Solomon Islands exporters of s,g,
              trailing 3 months, MAD-based outlier trim
  alpha    <- reliability weight (higher where destination trade is concentrated
              in few, well-reported markets)
  V_ref    <- alpha * u_mirror + (1-alpha) * u_peer
  V_det    <- current gazetted determined value for s,g
  PUBLISH V_ref alongside V_det with the gap in percentage terms
FOR each export declaration e of species s, grade g, volume Q:
  Y_e <- 0.25 * Q * max(0, V_ref - V_det)
  FLAG e for post-clearance review if Y_e exceeds a materiality threshold
ANNUALLY: Minister reviews V_ref series as the basis for the next gazette order,
  replacing ad hoc revision with a standing, data-linked update cycle

4.2 Algorithm 2: liquidity early warning

WEEKLY:
  cash        <- Treasury cash position, all accounts
  rev_forecast <- expected revenue over forward window k, by source
  commitments  <- committed expenditure over forward window k, by category
  LQ_t         <- (cash + rev_forecast - commitments) / commitments
  IF LQ_t < warning_threshold: alert Treasury and Ministry of Finance leadership
  IF LQ_t < critical_threshold: trigger the contingency sequence (expenditure
     reprioritisation, donor liaison, emergency financing options)
FOR each planned bond auction:
  bid_cover_avg <- trailing 12-auction average bid-to-cover ratio
  max_issuance  <- gamma * bid_cover_avg * trailing average auction size
  FLAG planned issuance exceeding max_issuance for review before the auction is called

4.3 Illustrative arithmetic: the log pricing gap

Transparency Solomon Islands' own figure, SBD 1-2 million a week, annualises to roughly SBD 52-104 million a year, or in the order of USD 6-12 million at typical exchange rates — a figure this paper treats as the headline claim to be tested by Model A, not as a verified number of its own. Applying a conservative reading (the lower end of TSI's range, and only the portion attributable to price rather than volume under-declaration) still implies annual recoverable revenue on the order of SBD 50-100 million once the determined-price schedule tracks a published reference value rather than an irregularly revised gazette order. Against a country where taxes on exports were over 12% of total tax revenue as recently as 2022, a gap of this size is not a rounding error in the national accounts.

4.4 Illustrative results: liquidity early warning

Using an illustrative forward window of six weeks and a stylised cash/commitment path consistent with the IMF's description of a widening 2026 deficit (4.1% of GDP projected) against depleted reserves, a warning threshold of LQ_t < 0.10 (cash plus expected revenue covering less than 110% of committed six-week expenditure) would have triggered multiple alerts across a simulated FY2025-26 path, most acutely around the months the IMF specifically flagged for "significant budget implementation risk." The exercise is illustrative of the indicator's behaviour, not a reconstruction of Solomon Islands' actual cash position, which only Treasury's own data can supply.

Solutions, software, data science and AI models

5.1 Issue-to-solution map

5.2 Model catalogue

5.3 Reference architecture

flowchart TD
  A[Sources: customs log export entries, China customs mirror data, mining agreements, Treasury cash position, bond auction results] --> B[Ingestion and data quality]
  B --> C[Warehouse: bitemporal, entity linked]
  C --> D[Reference-price service: mirror/peer blend]
  D --> E[Duty-at-risk flags]
  E --> F[Post-clearance review]
  C --> G[Resource transition and mining incentive scoring]
  C --> H[Liquidity early-warning service]
  H --> I[Treasury alerts]
  C --> J[VAT-readiness tracker]
  D --> K[Public reference-price register]

The liquidity service is the only component needing weekly, near-real-time operation; the reference-price and transition models run quarterly.

5.4 Software stack

5.5 Governance of AI use

• Publish the reference price, not just the flag. Making V_ref public alongside V_det is itself an accountability mechanism, independent of any enforcement action it triggers.
• No automated duty reassessment. Y_e selects entries for post-clearance review; only that review, with documentary evidence, supports a reassessment.
• Separate the mining-incentive score from any single official's sign-off. I_m is computed and logged independently of the negotiating team for a given agreement, so a high score cannot be quietly dropped.
• Treasury alerts go to more than one office. A liquidity alert routed to a single point of contact defeats the purpose of an early-warning system in a small administration; route to both the Ministry of Finance and an independent oversight body.

5.6 Team and indicative effort

A small joint Customs, IRD and Treasury cell — one lead, one data analyst covering trade and mining data, one Treasury liaison for the liquidity model — is proportionate to Solomon Islands' scale. The trade-data component (Model A) is inexpensive to start, since it uses public mirror data; the liquidity model (Model C) can be built directly from Treasury's own existing cash records with a modest reporting-automation effort. External support of four to six months is a reasonable estimate for the initial build. Cost and staffing figures are estimates.

Limitations

• The SBD 1-2 million weekly figure is Transparency Solomon Islands' own public estimate, not an independently verified government figure; Model A's purpose is to generate a defensible, ongoing figure to replace it, not to validate it in advance.
• China's dominance as an export destination makes mirror-data reference pricing workable for logs specifically; it will not transfer directly to gold or other minerals, where trade is more diffuse and Model B's incentive-score approach is more appropriate than a mirror-price benchmark.
• The liquidity early-warning thresholds in section 4.4 are illustrative; real thresholds need calibration against Solomon Islands' own historical cash-crisis episodes, which only Treasury holds.
• The bond-issuance ceiling assumes a reasonably informative bid-to-cover history; with very few auctions, the trailing average will be noisy and should be treated as indicative rather than a hard cap in the model's early operation.
• This paper does not have access to the VAT Bill's actual text beyond public reporting of its headline 15% rate; the readiness checklist in section 3.3 should be built directly against the Bill once available, not against this paper's assumptions about its likely content.

References

• International Monetary Fund. IMF Executive Board Concludes 2026 Article IV Consultation with Solomon Islands. Press Release 26/237, 7 July 2026.
• International Monetary Fund. IMF Executive Board Concludes 2024 Article IV Consultation with Solomon Islands. Press Release 25/042, 25 February 2025.
• International Monetary Fund. IMF Staff Completes 2024 Article IV Mission to Solomon Islands. Press Release 24/430, 21 November 2024.
• Transparency Solomon Islands, as reported in Solomon Times, "Setting the determined price for logs is not a job for MPs: TSI."
• Solomon Islands Customs and Excise Act (Cap. 121), Determination of Duties on Log Exports orders, various years, PacLII.
• World Bank. Taxes on exports (% of tax revenue), Solomon Islands, World Development Indicators, accessed via tradingeconomics.com, March 2026.
• vatcalc.com. "Solomon Island VAT implementation: Slow progress on 15% VAT to replace Sales Tax." 2026.
• Methods: as in the Fiji and Tonga papers in this series — Keen, M. (2013), The Anatomy of the VAT, IMF Working Paper 13/111; Hutton, E. (2017), The Revenue Administration Gap Analysis Program, IMF Technical Notes and Manuals 17/04; WTO Agreement on Customs Valuation. Method citations are from memory; verify before publication.