Electronic receipting and invoicing (EFRIS)

EFRIS was expected to increase tax revenue by reducing the scope for evasion, such as claiming input tax credits several times on one invoice, invoice mills and under-reported sales.

¶InterpretationInterpretation: A reading of the evidence, by a named source or by this register. [WIDER-2025-112, printed page 10, PDF page 12]

Marked as interpretation: the source states this objective, but either does not tie it to this measure in so many words or is not the Ministry that proposed it.

Policy objective as classified by the register: Compliance. Intended revenue direction: raise.

1. Rung 1

   Observed change

   ○Observed: A value or event as an official source records it, copied without calculation.
   • Collections for the tax heads affected are drawn on the story page, from URA's annual data books. Before and after is never presented as this amendment's effect.
2. Rung 2

   Official forecast

   →Official forecast: What MoFPED or URA said a measure would raise, or why it was made, before the year began.
   • URA: UGX 170 bn expected in FY2019/20 (Electronic Fiscal Devices).[URA-ARR-1920, printed page 53, PDF page 60, Annex 3, item A1.0]
3. Rung 3

   Reported yield

   ←Reported yield: What URA attributes to a measure after the year. An official attribution, not a causal estimate.
   • URA reports UGX 0.64 bn for FY2019/20, 0.38 percent of the forecast (computed).[URA-ARR-1920, printed page 53, PDF page 60, Annex 3, item A1.0]
4. Rung 4

   Static or modelled estimate

   ≈Modelled estimate: A mechanical or model calculation, with every assumption shown.

   ∅ Missing. No estimate made. A static estimate needs the taxable base before the change (volumes or values by product), which the register does not hold.

5. Rung 5

   Causal estimate

   ⇒Causal estimate: An effect identified by a stated research design against a comparison group.
   • Synthetic difference-in-differences (Arkhangelsky et al. 2021) on a balanced monthly panel of EFRIS users. Treatment: The April 2022 instruction that about 5,000 larger firms move to system-to-system EFRIS integration. Population: VAT-registered firms using EFRIS; treated firms are those named in the April 2022 mandate. Period: January 2021 to June 2023.[WIDER-2025-112, PDF page 2]
   • Reported purchases: minus 43 percent.
   • Reported VAT liabilities: nearly +150 percent.
   • Reported sales: minus 39 percent.
   • Caveat: It estimates the effect of an administrative mandate made under the e-invoicing provisions, not of the statutory amendment itself.
   • Caveat: Treated firms still had negative VAT liabilities on average at the end of the data period.
   • Caveat: The fall in reported sales is read by the authors as indicative of sales suppression.
6. Rung 6

   Wider impact

   ¶Interpretation: A reading of the evidence, by a named source or by this register.
   • Firms mandated to integrate their systems with EFRIS also reduced their reported sales, which the authors read as indicative of sales suppression. The estimates are on the Revenue tab.[WIDER-2025-112, PDF page 2]
   • The introduction of e-invoicing coincided with faster growth in the number of monthly VAT filers, many of whom previously filed corporate income tax returns.[WIDER-2025-112, PDF page 2]

What is known, what is uncertain, and what would settle it

• Known: official forecast, reported yield, causal estimate, wider impact.
• Uncertain: static or modelled estimate; and 1 open point below.
• Would settle it: Request to the Clerk to Parliament for certified copies of the Excise Duty, Income Tax, Value Added Tax, Stamp Duty and Tax Procedures Code (Amendment) Acts of 2015 to 2025, with their dates of assent and Gazette notice numbers. Held by Uganda Printing and Publishing Corporation (Uganda Gazette); Parliament of Uganda Library; ULII.
• Would settle it: Request to the Uganda Law Reform Commission for the tables of comparative section numbers accompanying the Revised Laws of Uganda 2023 for Caps 336, 338, 339, 343 and 344. Held by Uganda Law Reform Commission.

From the amendment to its wider consequences. Select a link to read what supports it. No money is drawn along the chain, because no flow along it is measured.

1. amendment: verified. The provision is read in the ULII consolidated text.
2. mechanism: verified. A change in how tax is assessed or enforced, not in what is owed.
3. implementation: verified. Dated implementation events are recorded.
4. response: verified. Mandated firms reported fewer purchases and lower sales.
5. outcome: verified. Higher reported VAT liabilities among mandated firms: a causal estimate, not a reported yield.
6. wider: verified. More taxpayers filing VAT monthly, many previously corporate income tax filers; an association, not a causal estimate.

Graded against ten criteria. No composite score: a reform that raises money can still be regressive or costly to comply with, and one that loses revenue can serve a legitimate aim. Revenue effectiveness is graded by a stated rule; the rest only where a source speaks to them.

Open or not evaluated

• Legal clarity and certainty

  Section 73A(1) of the 2018 amendment says a taxpayer may issue e-invoices, while s. 73A(2) lets the Commissioner make them mandatory by gazette. Retailers argued that a notice covering every VAT-registered taxpayer goes beyond the Act; the High Court refused interim relief without deciding the point.[ULII-UGHCCD-2020-154, printed page 7, PDF page 7]

  ◌ Inconclusive

• Administrative feasibility

  No source read for the register speaks to this criterion.

  · Not evaluated

• Compliance cost

  No source read for the register speaks to this criterion.

  · Not evaluated

• Efficiency

  No source read for the register speaks to this criterion.

  · Not evaluated

• Investment and competitiveness

  No source read for the register speaks to this criterion.

  · Not evaluated

• Vertical and horizontal equity

  No source read for the register speaks to this criterion.

  · Not evaluated

• Distributional effect

  No source read for the register speaks to this criterion.

  · Not evaluated

• Unintended consequences

  The causal study finds mandated firms cut reported sales while their VAT liabilities rose, which the authors read as possible sales suppression.[WIDER-2025-112, PDF page 2]

  ◌ Inconclusive

Consequences on record

• ⇒Causal estimateCausal estimate: An effect identified by a stated research design against a comparison group.Firms mandated to integrate their systems with EFRIS also reduced their reported sales, which the authors read as indicative of sales suppression. The estimates are on the Revenue tab.[WIDER-2025-112, PDF page 2]
• ~AssociationAssociation: Two things moved together. Not evidence that one caused the other.The introduction of e-invoicing coincided with faster growth in the number of monthly VAT filers, many of whom previously filed corporate income tax returns.[WIDER-2025-112, PDF page 2]

Dated events

• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2020-05-01 (month only): EFRIS pilot with a small number of taxpayers (implementation).[WIDER-2025-112, printed page 11, PDF page 13]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2020-07-01: All VAT-registered firms instructed to register for EFRIS (implementation).[WIDER-2025-112, printed page 11, PDF page 13]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2021-01-01: VAT-registered firms expected to issue e-invoices for all sales (implementation).[WIDER-2025-112, printed page 11, PDF page 13]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2021-11-01: URA begins active enforcement of e-invoicing (implementation).[WIDER-2025-112, printed page 11, PDF page 13]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2022-04-07: About 5,000 larger firms told to move to system-to-system integration (announcement).[WIDER-2025-112, printed page 12, PDF page 14]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.2022-06-30: Deadline for system-to-system integration (implementation).[WIDER-2025-112, printed page 12, PDF page 14]

Later changes, cases and practice

• §Verified lawVerified law: Wording or a date read in a statutory text, with the version and page named.Section 93 of the revised Act refers to taxpayers 'specified under section 80(2)'. Section 80 concerns tax stamps fixed on wrong goods; section 92(2) is where taxpayers are specified. The cross-reference looks like a revision slip and is flagged for verification.[ULII-TPC-2024, PDF page 43, s. 93]
• !UnverifiedUnverified: Not yet checked against a primary source. Shown so the gap is visible.The working paper reports that the fine for not issuing an e-invoice was initially a fixed sum per invoice and was amended in the 2025/26 tax bills; that amendment is outside the ULII version read.[WIDER-2025-112, printed page 12, PDF page 14]
• ○ObservedObserved: A value or event as an official source records it, copied without calculation.URA piloted the e-invoicing system with 58 taxpayers from May 2020, then on 23 June 2020 gazetted SI No. 82 of 2020 and General Notice No. 595 of 2020, making e-invoicing mandatory for every VAT-registered taxpayer from 1 July 2020. Five supermarkets sought an injunction, arguing that s. 73A(1) of the 2018 amendment leaves e-invoicing to the taxpayer's choice; the High Court refused it on grounds of public interest, leaving that question to the judicial review.[ULII-UGHCCD-2020-154, printed page 4 to 14, PDF page 4]

Before the Tribunal and the courts

• Capital Shoppers Ltd and 4 Others v Uganda Revenue Authority Miscellaneous Application No. 265 of 2020, [2020] UGHCCD 154

  High Court, Civil Division, 10 July 2020

  Injunction refused, with no order as to costs. The court held that public interest weighs against restraining URA from collecting or managing revenue save in exceptional circumstances, and that the gazette notice already bound the VAT-registered applicants. Whether the notice goes beyond s. 73A of the Act was left to the pending judicial review.

• ULII-TPC-2024

  Tax Procedures Code Act (Chapter 343), version of 23 December 2024

  Uganda Legal Information Institute (ULII) and Laws.Africa, from the Law Reform Commission revision. 2024-12-23. Covers Text in force 23 December 2024 to 30 June 2025.

  Desktop/Personal/Tax Legislation and URA Returns/Tax Procedures Code Act Cap 343.pdf

  Status: verified. Pages confirmed for this record in the source audit: 43 (PDF).

  Read in full with pdftotext. Lists amending Acts on PDF page 6.

• WIDER-2025-112

  The panopticon taxman: the impact of e-invoicing on VAT compliance in Uganda (WIDER Working Paper 2025/112)

  UNU-WIDER; authors Adrienne Lees, Maria Jouste, Nicholas Musoke and Joseph Okello. 2025-12. Covers VAT returns January 2021 to June 2023.

  https://doi.org/10.35188/UNU-WIDER/2025/671-1 · Desktop/Personal/URA DATA BOOKS AND REVENUE REPORTS/wp2025-112-panopticon-taxman-impact-einvoicing-VAT-compliance-Uganda.pdf

  Status: verified. Pages confirmed for this record in the source audit: 2, 12, 13, 14 (PDF).

  Published working paper, read with pdftotext: abstract (PDF page 2) and institutional background (PDF pages 12 to 14).

• URA-ARR-1920

  Annual Revenue Report FY 2019/20

  Uganda Revenue Authority. 2020. Covers FY2019/20.

  Desktop/Personal/URA DATA BOOKS AND REVENUE REPORTS/Annual Revenve Report FY 2019-20.pdf

  Status: verified. Pages confirmed for this record in the source audit: 60 (PDF).

  Table 20 (PDF page 50, printed 43) and Annex 3, the measure-level table (PDF pages 60 to 62, printed 53 to 55), read from page images. The file name misspells Revenue.

• URA-ARR-2223

  Annual Revenue Report FY 2022/23

  Uganda Revenue Authority. 2023. Covers FY2022/23.

  Desktop/Personal/URA DATA BOOKS AND REVENUE REPORTS/Annual Revenue Report FY 2022-23.pdf

  Status: verified. No page of this source is confirmed for this record by the audit; it is cited for context.

  Table 22 (PDF page 62, printed 55) and the Digital Tax Solution section (PDF page 58, printed 51) read.

• ULII-UGHCCD-2020-154

  Capital Shoppers Ltd and 4 Others v Uganda Revenue Authority, Miscellaneous Application No. 265 of 2020, [2020] UGHCCD 154 (ruling of 10 July 2020)

  High Court of Uganda, Civil Division (Ssekaana J), as published by ULII. 2020-07-10. Covers Ruling on a temporary injunction.

  https://ulii.org/en/akn/ug/judgment/ughccd/2020/154/eng@2020-07-10

  Status: verified. No page of this source is confirmed for this record by the audit; it is cited for context.

  Read in full from the PDF published by ULII (14 pages).

Corrections to this record

• 2026-10-21 Capital Shoppers Ltd and 4 Others v URA [2020] UGHCCD 154 read in full; the case marked read, the afterlife extended with the 2020 pilot and gazette notice, and a clarity assessment added. The ruling records the pilot, SI No. 82 of 2020 and General Notice No. 595 of 2020, and the dispute over s. 73A.