Method and Sourcing

Every figure below carries a paragraph and page citation from the substantive 2026 Budget Statement and Economic Policy ("Budget") and the 2026 Mid-Year Fiscal Policy Review presented to Parliament on 23 July 2026 ("MYR"), or from the Budget's Appendix 4A (MDA Expenditure Allocation) or the MYR's Appendix 2C (Economic Classification of Central Government Expenditure - H1).

Where the two documents, or a document and its own appendix, do not agree, that is stated as a discrepancy requiring clarification rather than asserted as an error, since Budget and Mid-Year appendices are structured differently (the Budget's Appendix 4A is Ministry-by-Ministry; the MYR's appendices are by economic classification, not Ministry), which can make figures look inconsistent when they are in fact drawn from different accounting bases.

Cross-Sector Tracker: Budget Commitment vs. H1 2026 Reported

Sector / Line item	2026 Budget commitment	H1 2026 reported (MYR)	% of annual	Flag
Big Push Infrastructure Programme	GH¢30.0bn (Budget 1144, p.142)	GH¢6.5bn paid (MYR 720, p.82); 87 projects commenced, 74 trunk roads/bridges, 10 urban, 3 feeder (MYR 563, p.66); 13 of 87 reached ≥50% completion, incl. 6 past 75%, and a further 15 past 25% (MYR 564, p.66)	21.7%	Narrative figure conflicts with Appendix 2C's Big Push CAPEX sub-lines (below), needs reconciliation, not asserted as an error
Big Push, Appendix 2C sub-lines	"MDAs CAPEX ABFA (Big Push)" GH¢15.334bn; nested "Big Push Capex" GH¢14.666bn (App. 2C, p.93)	Provisional H1: GH¢8.142bn and GH¢7.830bn respectively (App. 2C, p.93)	53-55%	These appendix figures are 20-25 points higher than the GH¢6.5bn narrative figure for the same programme, the two should be reconciled before either is relied on alone
Ministry of Roads & Highways (direct CAPEX)	GH¢4.301bn CAPEX (Budget App. 4A, Infrastructure Sector-GOG, p.164), matching the GH¢4.3bn narrative figure (Budget 1142, p.142)	No cedi figure in the MYR's Roads narrative (673-685, pp.76-77); not identifiable in MYR's economic-classification appendices, which report by expenditure type, not Ministry	Unverifiable	Confirmed absent from the mid-year account despite being a clean, exact figure in the Budget's own appendix
Road Maintenance Trust Fund (166 constituencies)	GH¢3.007bn (Budget 1143, p.142; App. 2C Budget line, p.93)	GH¢1.7bn paid (MYR 720, p.81); Appendix 2C provisional actual GH¢1.708bn (p.93)	56.8%	Ahead of the 46.0% programmed pace — on track
Energy Sector shortfalls (IPPs)	GH¢15.2214bn (Budget 1140, p.141; App. 2C Budget line, p.93)	Appendix 2C provisional actual GH¢6.933bn (p.93); narrative cites a separate, undifferentiated GH¢7.1bn "to the energy sector for stable power supply" (MYR 720, p.81)	45.6%	Modest lag against the 54.1% programmed pace; narrative and appendix figures are close but not identical and are not explicitly reconciled
Legacy IPP debt	GH¢4.8bn (Budget 1141, p.141)	No separately identifiable H1 actual found in Appendix 2C or the MYR narrative	Unverifiable	May be embedded in the combined GH¢7.1bn or GH¢6.933bn figures above; no line isolates it
Farmer Service Centres (50 centres)	GH¢690m CAPEX (Budget 1136, p.141; App. 4A, p.163)	GH¢551m placed in escrow at the Bank of Ghana for a Letter of Credit to procure 1,840 machinery units (MYR 599-600, p.69; 720, p.82)	79.9% (escrowed, not spent)	Escrowed for procurement, not yet spent on operating the centres; no centre reported operational
Ghana Buffer Stock Company / NAFCO	GH¢200m per the Budget's own narrative (1137, p.141); but GH¢105m per the Budget's own Appendix 4A CAPEX line for the same entity (p.163)	20,000+ MT of grains purchased Jan-June 2026, targeting 32,000 MT by year-end, volume only, no cedi figure (MYR 629-630, p.72)	Unverifiable in cedis	GH¢95m unreconciled gap inside the Budget document itself, between its own narrative and its own appendix, structurally the same pattern as the KG/JHS3 textbook mismatches already found in education
Oil Palm finance window	GH¢6.9bn domestic window, 2026-32, initial 2026 rollout (Budget, 1134, p.141)	"US$500 million Oil Palm Development Finance Facility", land-bank progress reported (270,000+ ha reviewed in Western Region); no cedi figure tying it to the GH¢6.9bn line (MYR 589-598, p.69)	Unverifiable	Same programme name (Integrated Oil Palm Development Policy) but a different, foreign-currency facility; no statement of whether it replaces, supplements, or is unrelated to the GH¢6.9bn domestic window
Agricultural enclave roads	GH¢828m for 1,000km (Budget 1135, p.141)	"US$523 million" to rehabilitate 1,050km, described as the same-titled "1000 kilometres Agricultural Enclave Roads Programme" (MYR 584-588, p.68)	Unverifiable	Same programme title, different currency and funding source, no cedi figure against the GH¢828m Budget line, and no cross-reference given
NHIS (claims, medicines, vaccines, Free PHC)	GH¢9.0bn (Budget 1157, p.143)	GH¢4.5bn paid (MYR 720, p.81)	50.0%	Exactly on pace
NHIS, Appendix 2C cross-check	"National Health Insurance Fund" grant line: GH¢11.297bn Budget (App. 2C, p.93)	Provisional H1: GH¢3.481bn (App. 2C, p.93)	30.8%	This appendix line is GH¢1.02bn (22.6%) below the GH¢4.5bn narrative figure, flagged for verification; the Fund's earmarked-levy transfer and the Scheme's total budget support may be different accounting bases, so this is a question, not a confirmed contradiction
MahamaCares (Ghana Medical Trust Fund)	GH¢2.3bn (Budget 1158, p.143)	GH¢1.1bn paid (MYR 720, p.81); Fund also reports mobilising GH¢19.82m in donations on top of its budget allocation (MYR 694, p.78)	47.8%	On pace
Hospital construction (3 regional hospitals, 7 upgrades, 10 Agenda 111 completions)	GH¢600m + GH¢79m + GH¢100m = GH¢779m combined (Budget 1160-1162, p.143)	Not mentioned anywhere in the MYR (independently verified: zero matches for "regional hospital," "Agenda 111" as a hospital programme, or "CHPS" in the full 107-page document, the sole "Agenda 111" mention, 298 p.32, is an unrelated debt-repayment pun)	Unverifiable	Entire combined line absent from the mid-year account; the only mid-year hospital news is MahamaCares-funded cardiology centres at Korle Bu, Komfo Anokye and Tamale (MYR ¶695, p.78), a different, smaller initiative
LEAP (Livelihood Empowerment Against Poverty)	GH¢1.1bn (Budget 1163, p.143)	GH¢485m paid (MYR 720, p.82)	44.1%	On pace
District Assemblies Common Fund (DACF)	GH¢8.9bn (Budget 1168, p.143); App. 2C Budget line GH¢8.770bn (p.93)	GH¢4.4bn paid (MYR 720, p.81); Appendix 2C provisional actual GH¢4.161bn (p.93)	47.4-49.4%	Narrative and appendix differ by about GH¢240m (5.4%), smaller than the GETFund gap but the same pattern
Assembly Members' allowances	GH¢100m (Budget 1167, p.143)	GH¢93m paid (MYR 720, p.82)	93.0%	Sharply ahead of pace, same pattern flagged for No Fees Stress in education
GETFund	GH¢9.872bn (Budget 1145, p.142; App. 2C Budget line, p.93)	GH¢4.2bn per narrative (MYR 720, p.81) vs GH¢3.481bn per Appendix 2C provisional actual (p.93)	35.3-42.5%	GH¢718.78m (17.1%) gap between the Review's own narrative and its own appendix, see the dedicated education report
No Fees Stress Policy	GH¢537.4m (Budget 1148, p.142)	GH¢537m paid (MYR 720, p.82)	99.9%	Full year's budget spent in H1, see the dedicated education report
GSLIP + basic-school construction + textbooks	GH¢1.1bn + GH¢2.0bn + GH¢3.0bn = GH¢6.1bn combined (Budget 1147, 1150-1151, p.142)	Zero mentions anywhere in the 107-page MYR (independently re-verified: no matches for "GSLIP," "E-Block," "mono-desk," or "textbook"). Replacement: US$300m World Bank financing, 210 interventions, incl. upgrading 20 Category B schools to Category A, up from 10 in the Budget (MYR 602-604, p.70; Budget 1038, p.131)	Unverifiable	Entire GH¢6.1bn programme absent, see the dedicated education report
Ministry of Education (overall headline allocation)	GH¢33.3bn (para 1145, p.142); GH¢33.76bn GoG-only / GH¢39.23bn all-sources per Appendix 4A (pp.165, 183)	Not reported anywhere in the Mid-Year Review, no ministry-wide total or MDA-level appendix exists	Unverifiable	The single largest number in the entire education budget has no mid-year counterpart at all; only sub-programme lines are reported.

Key highlights:

1. The Budget's single largest education commitment, "we have allocated GH¢33.3 billion to the Ministry of Education for its programmes and activities in 2026" (para 1145, p.142), never reappears in the Mid-Year Review in any form, and no MDA-level appendix equivalent to the Budget's own Appendix 4A (pp.165, 183) exists in the Review. Every education figure reported (para 720, pp.81-82) is a sub-programme line. It is possible to check individual programmes; it is not possible to check the Ministry as a whole.

• Four separate capital lines, GSLIP and basic-school construction (education), hospital construction (health), the Ministry of Roads & Highways' own GH¢4.301bn CAPEX line, and the Oil Palm and agricultural enclave roads windows (agriculture), have no mid-year cedi figure that ties back cleanly to the Budget line that funded them. In three of the four agriculture/roads cases, the Review does report activity under a similarly or identically named programme, but financed in US dollars rather than the Budget's cedi allocation, with no statement of whether the two are the same facility, a successor to it, or something additional. That ambiguity, repeated across three different sectors in almost the same form, is now a pattern rather than a single anomaly.

• The Ghana Buffer Stock Company/NAFCO case is the clearest instance: the Budget's own narrative promises GH¢200 million (paragraph 1137) while the Budget's own Appendix 4A allocates GH¢105 million to the same entity (p.163), a GH¢95 million gap with no footnote reconciling it. This is structurally identical to the kindergarten and JHS3 textbook mismatches already documented in the education report, and it shows the pattern of internally inconsistent figures is not confined to the Ministry of Education.

• The National Health Insurance Fund's Appendix 2C transfer (GH¢3.481bn provisional H1) sits GH¢1.02bn below the GH¢4.5bn NHIS figure quoted in the same paragraph 720 that contains the GETFund figures; proportionally, an even larger gap (22.6%) than GETFund's 17.1%. The DACF shows a smaller version of the same thing (GH¢240m, 5.4%). None of this proves double-counting on its own; the Fund-level appendix transfer and a Ministry's total budget support can legitimately differ, but it means the GETFund finding is now the third instance of the same shape of discrepancy, not an isolated one, and all three deserve the same explanation from Government.

• The mid-year review is explicit that it sought no supplementary estimate and instead undertook a "strategic realignment of expenditures within the existing appropriations" (MYR 14, p.1). Specific reallocations are named: GH¢350m from the Contingency Vote to the Ministry of Interior's Disaster Management Committee for flood relief (461, p.50), GH¢226m reallocated within the Ministry of Works, Housing and Water Resources for flood mitigation (462, p.50), GH¢400m reallocated for high-occupancy buses for Metro Mass Transit and STC (475, p.51), and a GH¢3.0bn downward revision to foreign-financed capital expenditure tied to GH¢5.0bn newly allocated to reserve accumulation under GANRAP (486-488, p.54). That is at least GH¢3.976bn in named, disclosed reallocations. What is still missing is a consolidated table showing, for every sector, how much of its original Budget ceiling was reduced to fund these and any other in-year reallocations; without that, it remains impossible to rule out that some of the unreported capital lines above were among the sources.

• The Budget allocated GH¢6.9 billion to the Oil Palm Development Finance Facility (para 1134, p.141) and GH¢828 million for 1,000km of agricultural enclave roads (para 1135, p.141). The Review reports the same two programmes as a "US$500 million Oil Palm Development Finance Facility" (p.69) and a US$523 million project covering 1,050km (para 585, p.68), neither figure reconciled to its original cedi allocation, and the road-length target shifted without comment.

• The GH¢961 million paid to GNPC (para 720, p.81) does not correspond to any Section 7 resource-allocation line in the Budget (paras 1129-1189, pp.140-144). This is the mirror image of the GSLIP problem: a substantial reported payment with no visible originating commitment to check it against.

Questions unanswered?

1. The Minister of Finance has to reconcile the Big Push Infrastructure Programme's GH¢6.5bn narrative payment figure (720, p.82) against Appendix 2C's GH¢8.142bn and GH¢7.830bn Big Push CAPEX sub-lines (p.93); which figure is the actual H1 disbursement?

• Provide the actual H1 2026-cedi disbursement against the Ministry of Roads & Highways' own GH¢4.301bn CAPEX line (Appendix 4A, p.164), absent from the Review's Roads narrative (pp.76-77).

• Reconcile the Budget's own GH¢200m narrative figure for the Ghana Buffer Stock Company (1137, p.141) against its own Appendix 4A allocation of GH¢105m for the same entity (p.163).

• Clarify whether the US$500m Oil Palm Development Finance Facility and the US$523m Agricultural Enclave Roads Programme are the same instruments as the Budget's GH¢6.9bn Oil Palm window (1134) and GH¢828m enclave roads line (1135), successors to them, or additional to them, and report the cedi disbursement against the original Budget lines either way.

• Report actual H1 2026 disbursement against the combined GH¢779m hospital-construction lines (3 regional hospitals, 7 upgrades, 10 Agenda 111 completions; 1160-1162, p.143), none of which appear anywhere in the Review.

• Explain the GH¢1.02bn gap between the GH¢4.5bn NHIS figure and the GH¢3.481bn National Health Insurance Fund transfer in Appendix 2C (p.93), and the smaller GH¢240m DACF gap, using the same standard already owed on GETFund.

• Publish a consolidated reallocation table showing every Budget line reduced to fund the GH¢350m, GH¢226m, GH¢400m and GH¢3.0bn reallocations already disclosed (461-462, 475, 486-488), and any others not yet named.

Overall Assessment

The core findings show that capital and infrastructure commitments across at least four sectors, education, health, roads, and agriculture, either go unreported at mid-year or reappear under a different, foreign-currency facility with no stated link back to the original Budget line, while at least three narrative figures (GETFund, NHIS, DACF) sit above their own appendix's provisional actuals by amounts too large to ignore. Education's GH¢6.1bn absence remains the single largest and clearest instance, but it is no longer an outlier; it is the sharpest example of a pattern now visible in the government's own numbers, across its own two documents.