## 1zmbea2020001

## Source details

**Canonical URL:** [1zmbea2020001](https://www.imf.org/-/media/files/publications/cr/2020/english/1zmbea2020001.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2020/english/1zmbea2020001.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2020/english/1zmbea2020001.pdf.json)

---

### Mission summary and context
- A Government Finance Statistics (GFS) mission visited Lusaka, Zambia, during November 19–30, 2018, under the United Kingdom’s Department for International Development (DFID) Enhanced Data Dissemination Initiative (EDDI) 2 project.
- Main counterpart: Zambia Central Statistical Office (CSO) Public Finance Unit (PFU); engagements also with Ministry of Finance Accountant General, Budget, Investment and Debt Management, and Economic Management departments.
- This was the third mission under DFID EDDI2 (following March 2016 and November 2017).
- Objective: review progress against previous GFS TA recommendations and assist in improving the quality of compiled and disseminated GFS.

### Key findings on implementation and data quality
- Progress on 34 prior recommendations:
  - 3 were completed,
  - 2 were partly completed,
  - 13 were ongoing,
  - 18 were not done.
- Constraints and risks:
  - Lack of resources at the CSO (staff and financial) limited progress.
  - Different fiscal aggregates produced by the Ministry of Finance and the CSO despite using essentially the same source data; some differences unexplained.
  - Inaccuracy or incompleteness of fiscal data could adversely affect policy making.

### Coverage gaps and scale of non-budgetary entities (preliminary, end‑2016 samples)
- Broadening coverage beyond budgetary entities is needed to understand true government size and fiscal vulnerabilities.
- Preliminary figures:
  - Local government entities (representative sample): surplus of a little less than 0.1 percent of GDP.
  - Social security funds: surplus of about 1.7 percent of GDP.
  - EBUs: deficit of about 0.1 percent of GDP.
  - Social security funds held financial assets of about 7.1 percent of GDP.
  - Sample of public corporations (financial and nonfinancial), including the Bank of Zambia: combined assets close to 60 percent of GDP.
- Recommendation: collect representative data beyond budgetary entities to assess fiscal risks.

### Priority recommendations (text preserved)
- Target Date: March 2019
  - Finalize terms of reference of the GFS technical working group and hold the first meeting to review and adopt the action plan.
  - Responsible Institutions: CSO, MOF
- Target Date: Immediate and continuous
  - Finalize and adopt a public sector institutional table for consistent use across all macroeconomic statistics; regularly review and update.
  - Responsible Institutions: CSO, BOZ, MoF, IDC, OAG
- Target Date: Immediate and continuous
  - Investigate and address (or explain) differences in fiscal statistics from the Budget Office at the Ministry of Finance, and the CSO’s PFU.
  - Responsible Institutions: CSO, MOF
- Target Date: Continuous
  - Continue collection and compilation of data for local government and EBUs missing from coverage; finalize compilation of GFS for all subsectors of general government; also collect data for public corporations.
  - Responsible Institutions: CSO

### Institutional and process weaknesses; suggested institutional actions
- Weak institutional capabilities and processes for sustainable compilation of quality fiscal data.
- Efficiency gains possible via coordinated work and automation (e.g., accessing source data directly rather than manual transcription from PDFs).
- Recommended institutional actions:
  - Strengthen PFU staffing (including increase in interns; temporary staff relocations during slow periods).
  - Support CSO staff participation in regional and online courses on GFS and PSDS.
  - Review, finalize, disseminate, and ensure consistent use of the preliminary public sector institutional table.

### Status of GFS compilation and dissemination
- CSO PFU compiles annual BCG GFS from the Ministry of Finance’s annual Financial Report (FR); revised BCG series for 2007–2017 were compiled and submitted to the GFS Yearbook (GFSY).
- No high-frequency GFS compiled by the CSO for the BCG; dissemination to GFSY limited to annual series and stock positions limited to debt liabilities.
- National Summary Data Page (NSDP) dissemination lapsed; last government operations and debt data posted dated back to 2016.
- Follow-up GFS TA mission tentatively scheduled towards the end of 2019 with objectives to improve subsector data, initiate consolidation for central government, investigate MOF–CSO differences, and validate/complete EBU data for 2011–2017.

### Staffing, process, and automation issues (BCG compilation)
- CSO PFU uses MOF Financial Report (FR) as source; FR include statements of cash receipts and payments, budget execution, comparison of authorized provision and actual expenditure, and comparison by functional classification; FR submitted to the Auditor General within three months after each financial year.
- Process weaknesses:
  - Staffing insufficient and staff inadequately trained.
  - Manual processing of large Excel files is time consuming and error prone (expenditure file for 2017 had 120,000 rows).
  - Revenue source file presentation changed for 2017 creating three separate series: for 2007 to 2010, for 2011 to 2016 and for 2017.
  - Suggested option: automate manual mapping directly into IFMIS at MOF to provide summarized reports for CSO GFS compilation.
- Noted large transaction: change in government currency and deposits in 2017 for 18.3 billion Kwacha (7.5 percent of GDP) — understood as adjustment to closing balances in 2016; mission recommended investigating given the size.

### Classification issues and required reclassifications (Budget Office)
- Budget Office compiles high-frequency and annual fiscal statistics but classifications could be improved toward GFSM 2014 consistency.
- Key issues:
  - Statement of operations includes transactions in financial assets and some liabilities above the line; financing limited to borrowing liabilities.
  - Change in balances is computed above the line as a residual; recommended to record change in balances below the line as transactions in financial assets and record any residual as a statistical discrepancy below the line.
  - Ensure DCS institutional coverage is consistent with fiscal statistics (DCS limited to budgetary central government).
- Specific transactions requiring reclassification or better treatment:
  - Receipts/sales from strategic reserves: treat as decrease in inventories and increase in cash; should not be recorded as revenue.
  - Transactions in financial assets/liabilities above the line: reclassify below the line.
  - Farmers’ Input Support Program (FISP): distinguish cases where repayment is expected (record as loan financial asset) versus not expected (record as other expense current transfer).
  - Food security pack: reclassify as social assistance benefits or other expense—current transfers not elsewhere classified.
  - Fuel payment arrears converted to debt: recognize as increase in debt securities below the line when issued; service as reductions in debt securities.
  - ZESCO electricity subsidy: record separately as subsidies.
  - Ordinary grants: check whether funds provided to public corporations should be recorded as subsidies, capital transfers, loans, or equity.
  - Social benefits: reclassify payments to pension fund that are grants or financing-gap coverage; employer social contributions for current period should be recorded as compensation of employees (social contributions) if applicable.
  - On-lending and repayments: ensure borrowing and on-lending flows and interest/principal treatments are fully captured.
  - Empowerment fund: record as loan financial asset if repayment expectation exists; otherwise record as another expense.

### Recommended actions for MOF / Budget Office
- Investigate recording and adjustment to cash and cash equivalents in MOF Financial Reports for financial years 2015–2017 pertaining to overstatement of IMF account balances.
- Investigate and address (or explain) MOF Budget Office vs CSO PFU differences to achieve consistent data for policy and macroeconomic statistics.
- Improve classifications in Budget Office’s statement of operations to align with GFSM 2014.

### Local governments: source data, compilation, automation, and key statistics
- No formal mechanisms at CSO for local government data collection; 109 institutional units may qualify as local government entities: provinces (10), city councils (5), municipal councils (13), town councils (91).
- CSO PFU began collecting local government data since prior TA mission; no specific template and some data non-electronic.
- Ministry of Finance established an Intergovernmental Relations (IGR) unit in August 2018 to collect quarterly data; templates developed and an application planned with GIZ support; IGR had received 22 reports for the first quarter of 2018 as of end‑2018 and plans a “Yellow Book” for local government starting 2020.
- Compilation using available data (2015–2018 aggregated sample) produced preliminary GFS with manual transcription and reclassification of financial transactions:
  - Total revenue averaged about 824 million Kwacha a year (about 0.36 percent of GDP).
  - Expenditure averaged 705 million Kwacha a year (about 0.31 percent of GDP).
  - Net lending (surplus) averaged 118 million Kwacha a year (about 0.05 percent of GDP).
- Recommendations:
  - Validate preliminary local government data for 2015–2018, collect missing entities, and disseminate upon validation.
  - Establish mechanisms to automate data collection (standardized templates, then data portals).

### Social security funds: data coverage, GFS compilation, and fiscal positions (2014–2017)
- Identified social security funds: NAPSA and WCFB.
- Data availability:
  - WCFB data cover 2013–2016.
  - NAPSA data cover 2014–2017.
  - Consolidated data for ZNPF and NAPSA available only starting fiscal year 2016; 2014–2015 aggregates for ZNPF and NAPSA were combined by the mission.
- Compilation notes:
  - Financial statements transcribed from print into Excel (income statement, statement of financial position, statement of cashflows); notes were not transcribed.
  - Transcription errors and sign conventions are important risks; adjustments were made where claims/pensions paid recorded with negative signs.
  - A GFS classification assistant was used; guidance provided in Appendix III.
- Preliminary fiscal and balance-sheet positions:
  - Aggregate positive fiscal balances over 2014–2017.
  - WCFB had net borrowing (deficit), small relative to large net lending of NAPSA (NAPSA’s assets in 2016 were seven times those of the WCFB).
  - Average surplus in subsector over 2015–2017: 3.3 billion Kwacha (about 1.6 percent of GDP).
  - Total assets at end 2017: roughly 23 billion Kwacha (about 9.3 percent of GDP).
  - Largest asset component: financial assets about 8.2 percent of GDP (mostly debt securities).
  - Debt liabilities at end 2017: 535 million Kwacha (about 0.2 percent of GDP), entirely payables.
  - No available details on net implicit obligations for future benefits.
- Statistical discrepancy:
  - Preliminary statement of operations includes a statistical discrepancy averaging 5.3 percent of net lending/net borrowing (2015–2017).
  - Recommendation: investigate and reduce/eliminate the discrepancy and residual in the integrated balance sheet.

### Extrabudgetary units (EBUs): coverage, preliminary aggregates, and issues
- Financial statements collected for about 29 EBUs; possible total EBUs up to 224 upon confirmation (coverage represents 13 percent of possible EBUs and accounts for 10.5 percent of BCG grants to EBUs).
- Data span 2011–2017 but inconsistent; inclusion of one large entity only for 2017 (NRFA) distorts 2017 aggregates.
- Preliminary 2017 aggregates (only 10 entities):
  - Revenue: about 32.6 billion Kwacha (about 13.2 percent of GDP).
  - Expenditure: about 42.6 billion Kwacha (about 17.3 percent of GDP).
  - Net borrowing (deficit): about 10 billion Kwacha (about 4.1 percent of GDP).
- Issues: incomplete coverage, high statistical discrepancies, manual data collection errors.
- Recommendations:
  - Review and validate EBU data for 2011–2017 and collect additional data for the most representative missing EBUs.
  - Establish regular data collection mechanisms from EBUs.

### Public corporations: data collected, asset coverage, and next steps
- Financial statements collected for 22 public nonfinancial corporations out of possible 50, and all known 6 public financial corporations for 2010–2017.
- Combined assets covered:
  - Public nonfinancial corporations: about 75.7 billion Kwacha (about 34.9 percent of GDP).
  - Public financial corporations: about 50.7 billion Kwacha (about 23.4 percent of GDP).
- Mapping to GFS categories not yet completed.
- Recommended priority: begin GFS compilation for largest corporations (e.g., ZESCO) and collect missing financial statements and missing years; establish regular data collection.

### Consolidation and dissemination
- An experimental consolidation of available subsectors (BCG, EBU, LG, SSF) into general government GFS is feasible in the very short run despite imperfect coverage; guidance in Appendix IV.
- Recommended dissemination actions:
  - Start disseminating GFS for broader government coverage (Central Government and General Government) as consolidated data become available.
  - Resume data dissemination on the NSDP.

### Follow-up technical assistance
- Objectives for the tentatively scheduled end‑2019 follow-up mission:
  - Improve data for all subsectors of general government.
  - Initiate consolidation for BCG and EBUs to produce preliminary central government GFS and explore consolidation with local governments.
  - Investigate and explain BCG data differences between CSO and MOF.
  - Validate and complete EBU data for 2011–2017 by collecting additional data for significant entities.

### Appendix I — status of implementation of prior TA (selected items, implementation status as of November 2018)
- Coordination and institutional actions:
  - Recommendation 1 (Establish Coordination Committee MoU): Target Date 1,2 June 2018 — Status: Partly completed (draft MoU exists).
  - Recommendation 2 (Submit Strategy of National Development of Statistics): Target Date 2 December 2018 — Status: Partly completed (NSDS2 launched July 2018).
  - Recommendation 3 (Submit Statistics Act for approval): Target Date 2 December 2018 — Status: Completed (Bill approved by Cabinet; still to go to Parliament).
  - Recommendation 4 (Seek Public Finance Act approval): Target Date 2 June 2018 — Status: Completed (Act in place since June).
- Source data and systems:
  - Recommendation 17 (GFS compilers access to IFMIS/FMIS): Target Date March 2018 — Status: Ongoing (letter drafted).
  - Recommendation 20 (Reconcile GRZ Financial Reports vs IFMIS): Target Date Immediate and ongoing — Status: Not done (requires TWG meetings).
  - Recommendation 21 (Reconciliation of deposits data): Target Date Immediate and ongoing — Status: Not done.
- Compilation and reporting:
  - Recommendation 26 (GFS time series FY 2011–2016 verification and submission): Target Date Immediate — Status: Completed.
  - Recommendation 29 (Review GFS high-frequency data and compile same as annual): Target Date Immediate and ongoing — Status: Not done.

### Appendix II — Terms of Reference and objectives for GFS Technical Working Group (TWG)
- Background: CSO mandated by the Statistics Act to collect and compile macroeconomic statistics including GFS; recent TA identified deficiencies in coverage, comprehensiveness, and timeliness.
- Current challenges summarized: limited coverage to budgetary entities; lack of comprehensive asset/liability stock data; classification gaps vs GFSM 2014 and PSDSG 2013; only annual GFS with lag > 11 months.
- TWG objectives (numbered):
  1. Foster initiatives to promote efficiency in timely GFS production across public-sector units.
  2. Promote data sharing among institutions.
  3. Harmonise GFS input data from different institutions to avoid inconsistencies.
  4. Facilitate interagency coordination in GFS production.
  5. Check completeness of scope and coverage of GFS data in Zambia.
  6. Ensure comprehensive GFS coverage for rebasing of national accounts for 2017.
  7. Help validate the institutions register (public sector institutional table).

### Appendix III — Key guidelines for compiling GFS from financial statements (high‑level)
- Primary sources: income statement, balance sheet, cash flow statement for EBUs, social security funds, public corporations, and local governments.
- General process steps:
  a) Prepare bridge tables for income statement, balance sheet, and cash flow statement.
  b) Calculate changes between closing and opening balance sheet values for each asset and liability category.
  c) Prepare derivation tables for major GFS aggregates.
  d) Compile GFSM 2014 detail tables and summary statements.
  e) Verify accuracy of the process.
- Classification and adjustments:
  - Classify items per GFSM 2014; identify and exclude other economic flows (price changes/revaluations, other changes in volume).
  - Adjust for items treated differently in financial statements (e.g., disposals/acquisitions of nonfinancial assets; loan disbursements/repayments; profits/losses on asset sales).
- Verification checks:
  - Major GFS aggregates should correspond to derivation tables; iterate between bridge, derivation, detail/summary tables.
  - Net lending/borrowing should be consistent by alternative calculations; differences recorded as a statistical discrepancy.
  - Closing balance sheet values should equal opening plus transactions and other economic flows.
  - Ensure net change in financial asset Currency and deposits in Statement of Government Operations equals net change in cash in Statement of Sources and Uses of Cash.
- Specific treatments and cautions:
  - Provisions: not treated as liabilities; treat as other change in volume (proxy for loss).
  - Depreciation: use consumption of fixed capital from national accounts where possible; depreciation only as proxy if necessary.
  - Pay careful attention to sign conventions to avoid discrepancies.

### Appendix IV — Consolidation of General Government GFS (principles and practical guidance)
- Consolidation purpose: present statistics for a group of units as a single unit by eliminating transactions and reciprocal stock positions among units being consolidated.
- Aggregation vs consolidation: consolidation follows aggregation and eliminates internal interactions to avoid double counting.
- Priority items to identify and eliminate:
  - Flows: transfers between government units; transactions in financial assets and liabilities; interest income/expense.
  - Stocks: loans; debt securities.
- Practical rules of thumb for reconciliation:
  - "The 'one-side' rule": if one party provides convincing evidence, impute to the other side.
  - "The 'top-down' rule": rely on budgetary/central government accounts when more complete/timely.
  - "The 'creditor is right' rule": for financial assets/liabilities normally rely on creditor records.
- Accounting practice: include counterparty identification in transaction codes where possible; if absent, revise chart of accounts to facilitate consolidation and automation.
- Impact on aggregates: consolidation changes magnitudes by eliminating internal interactions but applied symmetrically does not alter balancing items in the GFS statements.

*Source: 1. Priority Recommendations, Zambia — International Monetary Fund (chapter/section content).*

### 1. Priority Recommendations _______________________________________________________________________5

### 1. Priority Recommendations

### Summary of mission outcomes and context
- A Government Finance Statistics (GFS) mission visited Lusaka, Zambia, during November 19–30, 2018, as part of the United Kingdom’s Department for International Development (DFID) Enhanced Data Dissemination Initiative (EDDI) 2 project.
- The mission worked primarily with staff at the Zambia Central Statistical Office (CSO) Public Finance Unit (PFU) and met staff from the Ministry of Finance’s Accountant General, Budget, Investment and Debt Management, and Economic Management departments.
- This was the third mission to Zambia under the DFID EDDI2 project, following the March 2016 and November 2017 missions.
- The mission reviewed progress against recommendations of previous GFS technical assistance (TA) missions and assisted with improving the quality of GFS compiled and disseminated.

### Key findings on implementation and data quality
- Out of 34 recommendations from previous missions:
  - 3 were completed,
  - 2 were partly completed,
  - 13 were ongoing,
  - 18 were not done.
- Owing to lack of resources at the CSO (both staff and financial), progress implementing previous recommendations has been limited.
- The CSO compiled GFS for the Budgetary Central Government (BCG) and submitted these for publication in the GFS Yearbook (GFSY). Revised BCG series for 2007–2017 were compiled.
- Preliminary GFS were compiled for local governments, extrabudgetary units (EBUs), and social security funds.
- Different sets of fiscal data are produced by the Ministry of Finance and the CSO; while source data are essentially the same, aggregates are not consistent across institutions. Some differences can be explained, others cannot.
- Inaccuracy or incompleteness of fiscal data could adversely affect policy making and must be addressed.

### Coverage gaps and fiscal scale of non-budgetary entities (preliminary)
- Broadening coverage beyond budgetary entities would give a better sense of the true size of government operations.
- Preliminary data as of end 2016 (representative samples where indicated):
  - Local government entities (for a representative sample of local governments) had a surplus of a little less than 0.1 percent of GDP.
  - Social security funds had a surplus of about 1.7 percent of GDP.
  - EBUs had a deficit of about 0.1 percent of GDP.
  - Social security funds held financial assets of about 7.1 percent of GDP.
  - Data for a sample of public corporations (both financial and nonfinancial), including the Bank of Zambia, indicated combined assets of close to 60 percent of GDP.
- Collecting data for representative entities beyond budgetary entities would help understand fiscal vulnerabilities and risks to fiscal policy.

### Institutional and process weaknesses
- Institutional capabilities and processes for sustainable compilation of good quality fiscal data are weak.
- Resource constraints are significant but certain efficiency gains could be achieved through coordinated work and automation (for example, accessing source data directly from providing agencies rather than manual typing from PDF documents).
- The mission recommended a detailed action plan with priority recommendations (see Priority Recommendation table below) to make headway in improving GFS.

### Table 1. Zambia: Priority Recommendations (textual content preserved)
- Target Date: March 2019
  - Priority Recommendation: Finalize the terms of reference of the GFS technical working group and hold the first meeting to review and adopt the action plan.
  - Responsible Institutions: CSO, MOF
- Target Date: Immediate and continuous
  - Priority Recommendation: Finalize and adopt a public sector institutional table for consistent use across all macroeconomic statistics (national accounts, monetary and financial statistics, external sector statistics, and government finance statistics); and regularly review and update the table.
  - Responsible Institutions: CSO, BOZ, MoF, IDC, OAG
- Target Date: Immediate and continuous
  - Priority Recommendation: Investigate and address (or explain) differences in fiscal statistics from the Budget Office at the Ministry of Finance, and the CSO’s PFU.
  - Responsible Institutions: CSO, MOF
- Target Date: Continuous
  - Priority Recommendation: Continue the collection and the compilation of data for local government and EBUs that are currently missing from the coverage and finalizing compilation of GFS for all subsectors of general government; and also collect data for public corporations.
  - Responsible Institutions: CSO

### Recommended institutional actions to support implementation
- Strengthen the PFU’s staffing, including through among others, an increase in the number of interns allocated to the unit; and temporary staff relocation from other units within the CSO during slow periods in those units.
- Support CSO staff participation to regional and online courses on GFS and PSDS.
- Review and finalize the preliminary public sector institutional table and disseminate it afterwards; and ensure consistent use across all macroeconomic accounts.

---

### Status of GFS compilation and dissemination

#### Implementation of recent GFS TA missions’ recommendations
- The mission reviewed progress against the 34 recommendations of previous GFS TA missions (April 2016 and November 2017).
- Seven recommendations relate to institutional operational capacities of CSO including training in GFS; 27 relate to methodologies in the compilation of GFS.
- Out of 34 recommendations:
  - 3 were completed,
  - 2 were partly completed,
  - 11 were ongoing (of which 3 have since been achieved),
  - 18 were not done.

#### Staffing and resource constraints at CSO PFU
- Inadequate staffing and resources at the CSO’s PFU constitute a significant risk to improving GFS compilation.
- Recommended actions:
  - Strengthen the PFU’s staffing (increase interns, temporary relocations).
  - Support staff participation in regional and online GFS and PSDS courses.

### Institutional coverage of the public sector
- A revised preliminary public sector institutional table identified 454 institutional units.
  - General government comprises 398 institutional units, including:
    - 2 social security funds,
    - 109 local government entities,
    - 287 central government entities.
  - There are 56 public corporations, 50 of which are nonfinancial public corporations.
- Reclassification focused on social insurance schemes:
  - Two social security schemes identified: National Pension Scheme Authority (NAPSA) and Workers Compensation Fund Board (WCFB).
  - Public Service Pension Fund (PSPF) and Local Authorities Superannuation Fund (LASF) were reclassified within general government (employment related social insurance schemes; unfunded defined benefits).
  - Pension and Insurance Authority (PIA) reclassified as a public financial corporation (regulatory entity).

### GFS compilation for the Budgetary Central Government (BCG)

#### Source data and classifications used by CSO
- The PFU at CSO uses the Ministry of Finance’s annual Financial Report (FR) to compile annual BCG GFS.
- The FR are consolidated financial statements covering Ministries, Provinces, and Spending Agencies (MPSAs) included in the national budget, based on financial statements prepared and submitted by MPSAs to the Ministry of Finance.
- The FR include statements of:
  - (i) cash receipts and payments,
  - (ii) budget execution,
  - (iii) comparison of authorized provision and actual expenditure by MPSAs,
  - (iv) comparison of budget and actuals by functional classification.
- FR are submitted to the Auditor General every year, within three months after each financial year.
- The FR’s statements, notes, and appendices constitute source data used by the PFU to compile annual BCG GFS.
- The CSO does not compile high-frequency GFS for the BCG. GFS compiled by the PFU are disseminated to the IMF’s Statistics Department (STA) for publication in the GFSY.

#### Process weaknesses and automation opportunities
- Staffing at the CSO’s PFU is insufficient and staff are not adequately trained.
- Manual processing of data from the Ministry of Finance is error prone and time consuming:
  - A GFS compilation file is used to assemble transactions and COFOG classifications, but data must be manually processed in large Excel files (the expenditure file for 2017 had 120,000 rows of data).
  - The revenue source file’s presentation changed for 2017 requiring changes to GFS compilation file mapping (there are now three separate series on revenue—for 2007 to 2010, for 2011 to 2016 and for 2017).
- Current approach to GFS compilation is not sustainable; an option is automating manual mapping directly into IFMIS at the Ministry of Finance to provide summarized reports for CSO GFS compilation.
- Ultimately, only one set of fiscal data needs to be produced and used by both the Ministry of Finance and the CSO.

#### Source data and classifications from the Ministry of Finance
- The Budget Office at the Ministry of Finance also compiles fiscal statistics, but aggregates and balances differ from those compiled by the CSO.
- Data used by the Budget Office presumably comes from the same source as the FR, but differences (significant across categories of revenue, expenditure and financing) cannot be explained by end-of-year adjustments alone.
- Differences in institutional coverage and the basis of recording expenditure transactions could explain some differences but this remains to be investigated.
- Focus areas for investigation should include differences in the recording of taxes, grant revenue, interest expense, and transactions in financial assets and liabilities.
- A large transaction recorded in the FR and reflected in the CSO’s BCG GFS: change in the government currency and deposits in 2017 for 18.3 billion Kwacha (7.5 percent of GDP). The mission understood this transaction was an adjustment to closing balances in 2016 that were erroneously overstated by inclusion of the IMF account balances; no further details were provided but the mission recommended investigating the transaction given the size.

*Source: 1. Priority Recommendations, Zambia — International Monetary Fund (chapter/section content).*

### 16. The Budget Office compiles both high-frequency and annual fiscal statistics, but

### 16. The Budget Office compiles both high-frequency and annual fiscal statistics, but classifications could be improved

### Classification issues in the Budget Office’s GFS
- Classifications in the GFS compiled are somehow aligned with the guidelines of the GFSM 2001, although:
  - The presentation of the statement of operations includes transactions in financial assets and some transactions in liabilities above the line.
  - Financing in the statement of operations is limited to transactions in liabilities pertaining to external and domestic borrowing.
  - The statement of operations includes transactions that, for consistency with the GFSM 2014 guidelines, would be recorded differently.
- The change in balances is currently computed and recorded above the line and used as a residual to equate transactions above the line with transactions below the line (which are limited to a subset of transactions in liabilities). A more appropriate recording would:
  - First record the change in balances (understood as the change in cash and deposits) below the line, as part of transactions in financial assets.
  - Record as the change in cash only amounts deemed so, i.e., amounts included in the Bank of Zambia’s Depository Corporation Survey (DCS) as the change in government deposits plus any cash holding not deposited in the banking system.
- Any residual difference between transactions above the line and below the line would be recorded as a statistical discrepancy below the line.
- Care should be exercised to ensure institutional coverage in the DCS is consistent with that of fiscal statistics; in this case the institutional coverage is limited to the budgetary central government.

### Specific transactions requiring reclassification (as described)
- Receipts of sales of crops from the strategic reserves:
  - These receipts should not be recorded as revenue.
  - When crops are purchased by government for storing in the strategic reserve: record an increase in inventories nonfinancial assets (goods for resale) and a decrease in cash (currency and deposits).
  - When crops are sold from the strategic reserves: record a decrease in inventories nonfinancial assets (goods for resale) and an increase in cash (currency and deposits).
- Transactions in financial assets and liabilities above the line:
  - These ought to be reclassified below the line and should not enter the calculation of the fiscal balance (net lending/net borrowing).
- Grants and other payments currently include various expense items and some transactions in assets and liabilities that should be distinguished to reflect their true economic nature. Specific examples discussed:
  - Farmers’ Input Support Program (FISP):
    - Repayments currently recorded as revenue (“non-tax revenue”) should be reclassified depending on whether repayment is expected.
    - For transactions where repayment is expected:
      - Record FISP payments to eligible farmers below the line as an acquisition of a financial asset in the form of a loan, with a corresponding reduction in another financial asset (currency and deposits) or an increase in accounts payable if not paid in cash at recording time.
      - When farmers repay, record a reduction in loan financial assets and an increase in currency and deposits.
    - For transactions where repayment is not expected:
      - Record FISP payments as another expense (current transfer not elsewhere classified) with a corresponding reduction in currency and deposits (if paid in cash). No subsequent financial asset entries.
  - Food security pack:
    - Currently recorded under “grants and other payments” but are payments directly to households and should be reclassified either as social assistance benefits or as other expense—current transfers not elsewhere classified.
  - Fuel payment arrears:
    - The stock of arrears for fuel payments converted into a debt instrument should be recognized in the statement of operations in the year and month the debt was issued as an increase in debt securities below the line; subsequent servicing recorded as reductions in debt securities below the line.
  - ZESCO electricity subsidy:
    - Included as “grants and other payments” but ought to be recorded separately as subsidies.
  - Ordinary grants:
    - May need checking to ensure funds labeled as grants are not provided to public corporations; if provided to public corporations they should be recorded as subsidies, or as other expense (capital transfers) for recapitalization, or as a financial asset (loan) or equity if government acquires an effective financial claim.
  - Social benefits:
    - Current social benefits include payments to the pension fund that are actually grants or funds to cover the pension fund financing gap; these transactions should be reclassified and not recorded as social benefits.
    - Only social assistance benefits should be recorded under social benefits given available information.
    - If payments are employer social contributions for the current period, record under compensation of employees as social contributions; if not current period contributions, classify elsewhere.
    - Transfers to the pension fund as another level of government should be recorded as grants to other levels of government; recurrent payments to cover a financing gap should also be recorded as grants.
  - On-lending and repayments to corporations:
    - When government borrows and on-lends, borrowing is recorded as an increase in loan liabilities and corresponding increase in cash, then on-lending recorded as decrease in cash and increase in loan financial assets.
    - Repayment of interest recorded as property income (interest) revenue; principal repayments recorded as negative (decrease) in loan financial assets.
    - Current statement of operations does not fully capture all on-lending transactions.
  - Empowerment fund:
    - If realistic expectation of repayment exists, record as loan financial asset (below the line). If not (no evidence of repayments), record as another expense (current transfers).

### Recommended Actions (Budget Office / Ministry of Finance)
- Investigate the recording, and adjustment to cash and cash equivalents in the Ministry of Finance’s Financial reports for financial years 2015–2017 pertaining to overstatement of IMF account balances.
- Investigate and address (or explain) differences in fiscal statistics from the Budget Office at the Ministry of Finance, and the CSO’s PFU with the aim to have consistent data being disseminated and used for policy purposes, and compilation of other macroeconomic statistics.
- Improve classifications in the Budget Office’s statement of operations to align to the GFSM 2014.

---

### D. GFS Compilation for Local Governments

### Source data and coverage
- There are no formal mechanisms for data collection for local governments in place at the CSO.
- Available information indicates a total of 109 institutional units that may qualify as local government entities, that include:
  - provinces (10)
  - city councils (5)
  - municipal councils (13)
  - town councils (91)
- The PFU at the CSO has started collecting data for GFS compilation from local governments only since a previous GFS TA mission; no specific template exists and data are not always in electronic format.
- Staff accessed revenue and expenditure data from all five city councils and all municipal councils; some data were received in non-electronic format and manually entered into Excel following what appears to be a standardized classification, though not used consistently across councils.
- The data do not include stock positions and no information is included on the change in cash and/or bank balances, although some transactions in financial assets and liabilities are included.
- The Ministry of Finance established a new Intergovernmental Relations (IGR) unit in August 2018 to collect and compile quarterly data from local governments.
  - Templates for data collection were developed and assistance sought from GIZ to develop an application for planning and budgeting to allow automated data collection and compilation.
  - As of end 2018, the unit had already received 22 reports, for the first quarter of 2018.
  - The IGR is planning to produce a “Yellow Book” for local government starting in 2020; reporting from the application would provide a comprehensive and consistent data source for GFS compilation.

### Compilation of GFS using existing data
- The mission compiled a partial GFS statement of operations using data available at the CSO for the period covering 2015 to 2018, using the GFS Multiperiod Classification Assistant.
- The local government data include transactions in financial assets and liabilities labeled as revenue and expenditure:
  - Revenue-side items include repayments on loan assets and borrowing, reclassified respectively as transactions in financial assets and liabilities.
  - Expenditure-side items include loan disbursements and repayment of payables reclassified as transactions in financial assets and liabilities.

### Automation, validation, and dissemination
- Manual transcription into Excel is time consuming and error prone; recommended initial step is to have local governments fill a designated template and submit electronically to the CSO.
- Longer term: develop a data portal with the template embedded for direct collection from local governments.
- Data collected for 2015–2018 may need checking to fill gaps, validation, and then dissemination.
- While available data do not indicate changes in the stock of cash, it could be assumed that part (or all) of the difference between transactions in revenue and expenditure and financing other than the change in cash constitutes the change in cash and deposits at local government.

### Key statistics from preliminary local government GFS (2015–2018, aggregated sample)
- Total revenue averaged about 824 million Kwacha a year (about 0.36 percent of GDP).
- Expenditure averaged 705 million Kwacha a year (about 0.31 percent of GDP).
- Resulting in a net lending (surplus) on average a year of 118 million Kwacha (about 0.05 percent of GDP).
- These aggregates reflect the largest local government entities and indicate the limited significance of local governments in Zambia’s financial operations.

### Recommended Actions (Local Government)
- Check and validate the preliminary data on local governments for 2015–2018, collect and compile data for the missing local government entities and start disseminating the data upon validation.
- Establish mechanisms to automate data collection from local governments, starting initially with standardized data collection templates, and then in time, with data collection portals.

---

### E. GFS Compilation for Social Security Funds

### Source data and coverage
- The mission identified two social security funds: NAPSA and WCFB.
- The PFU at the CSO collected financial statements in print form and transcribed them into Excel for GFS compilation.
- Available data coverage:
  - WCFB data cover the period 2013–2016.
  - NAPSA data cover the period 2014–2017.
- Prior to closure of the Zambia National Provident Fund (ZNPF) and its merger with NAPSA, separate financial statements were available for both ZNPF and NAPSA.
  - Consolidated data are available only starting with fiscal year 2016 for the defunct ZNPF and NAPSA.
  - Data for 2014–2015 present aggregates for ZNPF and NAPSA separately; the mission aggregated data for ZNPF and NAPSA for the years 2014 and 2015.
- Data do not indicate transactions between the two schemes for that period that may warrant consolidation, but that remains to be confirmed.

### Compilation of GFS from financial statements
- Available data are quite comprehensive for compiling GFS for the social security funds subsector.
- Main challenges:
  - Use of manually entered data from paper files can produce transcription errors, especially signs of entries in financial statements and the statement of cashflows.
  - The three main statements transcribed into Excel were:
    - (i) the statement of comprehensive income
    - (ii) the statement of financial position
    - (iii) the statement of cashflows
  - Notes associated with the statements were not transcribed.
- A GFS classification assistant was used to compile GFS statements for the social security funds subsector.
  - Adjustments were made to signs in the statement of comprehensive income (claims and pensions paid recorded with a negative sign in financial statements were reverted to a positive sign to record an expense transaction).
  - The statement of cashflows relies primarily on the accuracy of transcription into Excel on the notation of increases/(decreases) to determine sign and recording as a revenue or expense item.
- General guidance on compilation of GFS from financial statements is provided in Appendix III.

*Source: Excerpt from IMF mission report chapter on Zambia (content unit 1zmbea2020001).*

### 27. GFS statements for the social security funds subsector presents a statistical

### 27. GFS statements for the social security funds subsector presents a statistical

### Statistical discrepancy in social security funds GFS
- Preliminary GFS statement of operations includes a statistical discrepancy possibly due to omissions, wrong sign assignment, or error in manual data transcription.
- The discrepancy is not large relative to net lending/net borrowing: 5.3 percent of net lending/net borrowing on average between 2015–2017.
- The integrated stocks and flows show a minor “residual” (the change in stock positions not explained by transactions and other economic flows).
- Required checks: conformity of data transcribed into Excel with the print statements, and inclusion of details in notes to the financial statements.

### Fiscal aggregates and balance-sheet positions for social security funds (2014–2017)
- Preliminary GFS indicate aggregate positive fiscal balances over the period 2014–2017.
- The WCFB had net borrowing (deficit), small relative to large net lending of NAPSA (NAPSA’s assets in 2016 were seven times those of the WCFB).
- Average surplus in the social security funds subsector over 2015–2017: 3.3 billion Kwacha (about 1.6 percent of GDP).
- Total assets of social security funds at end 2017: roughly 23 billion Kwacha (about 9.3 percent of GDP).
- Largest component of assets: financial assets about 8.2 percent of GDP (mostly represented by debt securities).
- Debt liabilities at end 2017: 535 million Kwacha (about 0.2 percent of GDP), entirely payables.
- No available details on estimate of net implicit obligations for future social security benefits from available data; such estimates would presumably exist from financial statements or actuarial studies.

### Recommended Action (social security funds)
- Investigate the source of the statistical discrepancy in the GFS statement of operations and the residual from the integrated balance sheet for the social security funds subsector and reduce or eliminate the discrepancy/residual.

### GFS compilation for Extrabudgetary Units (EBUs): data coverage and issues
- Financial statements collected for about 29 institutional units tentatively classified as EBUs in the preliminary institutional table.
- Possible total EBUs (upon confirmation): up to 224 EBUs.
- Coverage represented: 13 percent of all possible EBUs and accounts for 10.5 percent of all grants from BCG to EBUs (comparing total grants revenue of the 29 units with available data, and grant expense from the BCG).
- At least up to 8 of the 29 units could be reclassified as BCG entities based on funding structure.
- Data span: 2011 to 2017 but not consistently for all entities; only PSPF and LASF cover 2011–2016 continuously.
- Inclusion of one large entity for only one year (NRFA: data only for 2017) distorts subsector aggregates for that year.

### Preliminary GFS for EBUs: quality and 2017 aggregates
- Very preliminary GFS for EBUs have high statistical discrepancies and need further improvement.
- For 2017 (data available for only 10 entities):
  - Preliminary revenue: about 32.6 billion Kwacha (about 13.2 percent of GDP).
  - Expenditure: about 42.6 billion Kwacha (about 17.3 percent of GDP).
  - Net borrowing (deficit): about 10 billion Kwacha (about 4.1 percent of GDP).
- Issues identified: incomplete coverage, statistical discrepancies, possible inaccuracies due to manual data collection and processing.

### Recommended Actions (EBUs)
- Review and validate the preliminary data for EBUs for the 2011–2017 period and collect additional data for the missing most representative EBUs (based on relative size).
- Establish mechanisms to collect data from EBUs on a more regular basis.

### GFS compilation for General Government
- An experimental consolidation of available subsectors (BCG, EBU, LG, SSF) into general government GFS can be envisaged in the very short run, despite imperfect coverage.
- General guidance on consolidation provided in Appendix IV (not reproduced here).

### GFS compilation for Public Corporations: data collected and next steps
- Financial statements collected for 22 public nonfinancial corporations out of possible 50, and all known 6 public financial corporations for 2010–2017.
- Combined assets covered:
  - Public nonfinancial corporations: about 75.7 billion Kwacha (about 34.9 percent of GDP).
  - Public financial corporations: about 50.7 billion Kwacha (about 23.4 percent of GDP).
- Financial statements include income statements, statements of financial position, and cashflow statements, but mapping to GFS categories has not yet been completed.
- Priority suggested: begin GFS compilation for largest corporations (e.g., ZESCO) while collecting missing data for other corporations.

### Recommended Actions (public corporations)
- Review and validate data entry from available financial statements of public corporations.
- Collect financial statements for missing public nonfinancial corporations, and for missing years, for those available.
- Establish mechanism to collect data from public corporations on a more regular basis.

### Data dissemination
- Annual GFS series compiled during the mission were disseminated to the GFSY; coverage currently limited to budgetary central government with transactions covering budgetary central government and stock positions limited to debt liabilities.
- No high-frequency GFS are currently disseminated.
- National Summary Data Page (NSDP) dissemination has lapsed; last available data on government operations and debt posted on the NSDP dates back to 2016.
- As coverage improves and consolidation is performed, dissemination for broader government coverage should be considered.

### Recommended Actions (dissemination)
- Start disseminating GFS for broader government coverage (i.e., Central Government, and General Government) as consolidated data becomes available.
- Resume data dissemination on the NSDP.

### Follow-up and technical assistance
- A follow-up GFS TA mission is tentatively scheduled towards the end of 2019 to assist with improvements in GFS compilation and dissemination.
- Mission aims:
  - Improve data compiled for all subsectors of general government.
  - Initiate consolidation for BCG and EBUs to produce preliminary central government GFS, and explore consolidation with local governments.
  - Investigate and explain differences in BCG data between the CSO and the Ministry of Finance.
  - Validate and complete EBU data for 2011–2017 by collecting additional data for the most significant entities.

*Source: 1zmbea2020001 - 27. GFS statements for the social security funds subsector presents a statistical (PDF chapter).*

### Appendix I. Status of Implementation of Previous Technical Assistance Mission

### Appendix I. Status of Implementation of Previous Technical Assistance Mission

### Coordination of GFS Compilation Process and Methodology
- Recommendation 1: Establish the Coordination Committee by undersigning MoU of Coordinating Committee.
  - Target Date: 1,2 June 2018
  - Responsible: Coordinating Committee
  - Implementation Status as of November 2018: Partly completed. A draft MoU exists yet to be consented among parties. CSO wrote to stakeholder institutions to nominate members of the TWG.
- Recommendation 2: Submit Strategy of National Development of Statistic for approval.
  - Target Date: 2 December 2018
  - Responsible: CSO
  - Status: Partly completed. Preparation of the NSDS2 was launched in July 2018.
- Recommendation 3: Submit Statistics Act for approval.
  - Target Date: 2 December 2018
  - Responsible: CSO
  - Status: Completed. Bill approved by Cabinet. Still to go to Parliament.
- Recommendation 4: Seek Public Finance Act approval.
  - Target Date: 2 June 2018
  - Responsible: MOF
  - Status: Completed. Act is in place since June.
- Recommendation 5: Develop a Zambia’s GFS Manual.
  - Target Date: 2 March 2019
  - Responsible: CSO
  - Status: Ongoing. Currently editing the manual.
- Recommendation 6: Develop an annual publication to report Zambia's financial statistics.
  - Target Date: 2 December 2018
  - Responsible: CSO
  - Status: Ongoing. Currently compiling data for the sub-sectors of general government.
- Recommendation 7: Present quarterly Summary Report on progress with GFS implementation.
  - Target Date: At the end of each quarter
  - Responsible: CSO
  - Status: Not done. Not yet.

### Institutional Coverage of General Government and Public Sector
- Recommendation 8: Refine the preliminary list of extrabudgetary units.
  - Target Date: June 2018
  - Responsible: CSO
  - Status: Not done. Working on list. Require funds to hold meeting with GFS TWG and consult.
- Recommendation 9: Finalize the institutional classification of public sector units by considering main principles of GFS institutional coverage: (i) residency, (ii) type of institutional unit, (iii) government control and (iv) main activity (market or non-market producer).
  - Target Date: 2 June 2018
  - Responsible: CSO
  - Status: Not done. Working on list. To some extent require involvement of GFS TWG to further consult.
- Recommendation 10: Review list of SOE and autonomous funds under LG control, especially Water and Sewerage Companies.
  - Target Date: June 2018
  - Responsible: CSO
  - Status: Not done. Working on list. To some extent require involvement of GFS TWG to further consult. All utility Companies are operating as corporations supposed to generate profits for LG.
- Recommendation 11: Discuss and agree on the structure of the public sector in Zambia with proposed members of the Coordinating Committee; classification of public corporations to be discussed and agreed by all providers of macroeconomic statistics.
  - Target Date: 1 September 2018
  - Responsible: Coordinating Committee
  - Status: Not done. Require funds to hold GFS TWG meetings.
- Recommendation 12: Ensure that GFS coverage is consistent with national accounts, external sector statistics; and monetary and financial statistics are consistent.
  - Target Date: Ongoing
  - Responsible: Coordinating Committee
  - Status: Not done. Require funds to hold GFS TWG consultative meetings.

### Accounting Rules and Systems
- Recommendation 13: Incorporate GFS coding system into the IFMIS.
  - Target Date: 2 March 2019
  - Responsible: MOF and CSO
  - Status: Not done. TBC.
- Recommendation 14: Final reviews of the automated process of calculating COFOG in the IFMIS currently being done by the AG should get analytical support from CSO.
  - Target Date: 2 June 2018
  - Responsible: MOF and CSO
  - Status: Not done. TBC.
- Recommendation 15: Use the revised CoA from now on to assist in the compilation of GFS.
  - Target Date: Ongoing
  - Responsible: CSO
  - Status: Ongoing.
- Recommendation 16: Public entities should address issues regarding harmonization of the basis of recording, CoA system and the fiscal year.
  - Target Date: December 2020
  - Responsible: Coordinating Committee
  - Status: Not done. Require funds to hold GFS TWG consultative meetings.

### Source Data
- Recommendation 17: GFS compilers should get access to the IFMIS and FMIS systems: send AG a letter requesting access and indicate the type of profile needed.
  - Target Date: March 2018
  - Responsible: CSO
  - Status: Ongoing. Letter Drafted.
- Recommendation 18: Seek clarification from MOF on the recording of financing transactions (disbursement, amortization, interest and other financing costs) and liabilities stock positions.
  - Target Date: March 2018
  - Responsible: CSO
  - Status: Not done.
- Recommendation 19: Investigate further the recording process of revenues particularly regarding the recording of “budget” and “Project support grants” and loans (in-cash and in-kind) from foreign and international donors and expenses.
  - Target Date: March 2018
  - Responsible: CSO and AG
  - Status: Not done.
- Recommendation 20: Reconcile revenue and expenditure data presented in the GRZ Financial Reports versus administrative data extracted from the IFMIS; reconciliation of various source data to derive statistics should be standard practice.
  - Target Date: 1,2 Immediate and ongoing
  - Responsible: Coordinating Committee
  - Status: Not done. Require funds to hold GFS TWG consultative meetings.
- Recommendation 21: Reconciliation process of deposits data should be a priority and established as routine.
  - Target Date: 1 Immediate and ongoing
  - Responsible: MOF and BOZ
  - Status: Not done. No information.
- Recommendation 22: Validate templates left by the mission to collect data from public entities other than BCG and start immediate distribution.
  - Target Date: Immediate
  - Responsible: CSO
  - Status: Not done. Still to engage in bilateral consultative process with institutions.
- Recommendation 23: Keep on collecting financial statements data of public entities other than the BCG entities, including public corporations (state owned enterprises).
  - Target Date: 1,2 Ongoing
  - Responsible: CSO
  - Status: Ongoing.
- Recommendation 24: Expand coverage of debt compilation and reporting beyond central government and start compiling and reporting publicly guaranteed debt.
  - Target Date: Immediate and ongoing
  - Responsible: CSO and IDM
  - Status: Ongoing.
- Recommendation 25: Study the possibility to start reporting market and nominal value of debt instruments.
  - Target Date: December 2018
  - Responsible: IDM
  - Status: Not done. TBC.

### GFS Compilation and Reporting
- Recommendation 26: GFS time series for FY 2011-2016 completed during the mission needs to be verified and officially resubmitted to the AFR and STA team for surveillance and inclusion in the IMF’s annual GFSY.
  - Target Date: 1,2 Immediate
  - Responsible: CSO
  - Status: Completed.
- Recommendation 27: Compile and disseminate the debt data to AFR and upload it on the World Bank website.
  - Target Date: 2 June 2018
  - Responsible: CSO
  - Status: Not done. TBC.
- Recommendation 28: Compile and submit the Enhanced General Data Dissemination System (e-GDDS) in order to guarantee consistency and seek training on it.
  - Target Date: June 2018
  - Responsible: CSO
  - Status: Ongoing.
- Recommendation 29: Review GFS high-frequency data and compile it in the same manner as annual process.
  - Target Date: Immediate and ongoing
  - Responsible: CSO
  - Status: Not done. Not yet.

### Consistency of GFS and Other Macroeconomic Data
- Recommendation 30: Consistency of GFS and other macroeconomic data should be part of frequent discussions of the Coordinating Committee.
  - Target Date: Ongoing
  - Responsible: Coordinating Committee
  - Status: Not done. Still to hold 1st meeting.

### Resources and Technical Assistance
- Recommendation 31: Attend available GFS and debt statistics courses.
  - Target Date: 2 Ongoing
  - Responsible: CSO
  - Status: Ongoing.
- Recommendation 32: IMF department and training centers (MEFMI) should forward GFS course invitations to the CSO to nominate officials responsible for GFS compilation.
  - Target Date: Ongoing
  - Responsible: IMF and its Training Centers
  - Status: Ongoing.
- Recommendation 33: Improve work conditions for Zambia’s GFS compilers.
  - Target Date: Immediate
  - Responsible: CSO
  - Status: Ongoing. Been having discussions with management.
- Recommendation 34: Seek funds to support proper GFS compilation.
  - Target Date: Immediate
  - Responsible: CSO
  - Status: Ongoing. Been having discussions with management.

*Note: 1 Constitutes a priority recommendation; 2 Also included in the action plan for implementing the GFSM 2014.*

### Appendix II — Terms of Reference for the Government Finance Statistics Technical Working Group: Background and Objectives
- Background highlights:
  - CSO mandated by the Statistics Act (cap 127 of the laws of Zambia) to collect and compile statistics for all macroeconomic sectors including GFS.
  - CSO disseminates GFS to the IMF Statistics Department; GFS are inputs to other macroeconomic statistics, including national accounts (GDP).
  - Recent technical assistance missions identified deficiencies in coverage and comprehensiveness of GFS that undermine data quality and other macroeconomic statistics.
  - CSO is setting up a technical working group (TWG) comprising all data-providing entities to improve quality, comprehensiveness, and coverage of GFS.
- Current Challenges (summary):
  - Data coverage: Current GFS limited to budgetary entities; excludes autonomous/extra-budgetary central government entities, local governments, social security funds, and public corporations.
  - Data comprehensiveness: Lack of comprehensive data on assets and liabilities; no systematic collection of government stock positions.
  - Data classifications: Gaps remain in alignment with GFSM 2014 and PSDSG 2013; inconsistencies across government datasets (MOF vs CSO).
  - Data timeliness and frequency: Only annual GFS compiled and disseminated with lag > 11 months; no monthly/quarterly dissemination.
- Objectives of the GFS Technical Working Group:
  1. Foster and support initiatives aimed at promoting efficiency in the production of timely government finance statistics across public-sector units.
  2. Promote data sharing among institutions to enhance compilation of government finance statistics.
  3. Harmonise government finance statistics input data from different institutions to avoid inconsistencies.
  4. Facilitate institutional (interagency) coordination in the production of government finance statistics.
  5. Check for completeness of scope and coverage of government finance data in Zambia.
  6. Ensure comprehensive coverage of government finance statistics data for rebasing of national accounts for the year 2017.
  7. Help validate the institutions register (i.e. public sector institutional table) covered under government finance statistics.

### Appendix II — Institutions on the Technical Committee
- Central Statistical Office—Economic and Financial Statistics Division
- Ministry of Finance—Budget—Budget Execution
- Ministry of Finance—Budget—Intergovernmental Fiscal Relations
- Ministry of Finance—Economic Management and Finance
- Ministry of Finance—Investment and Debt Management
- Accountant General—Local Government Unit
- Accountant General—Revenue
- Accountant General—Expenditure
- Accountant General—Asset Management
- Office of the Auditor General—Public Debt and Investment
- Industrial Development Corporation
- Ministry of Local Government
- Lusaka City Council
- Bank of Zambia
- National Assembly
- Pensions and Insurance Authority
- Local Authorities Superannuation Fund
- NAPSA
- Public Service Pensions Fund

### Appendix III — Compilation of GFS from Financial Statements: Key Guidelines
- Primary sources:
  - Income statement, balance sheet, and cash flow statement for extra-budgetary units, social security funds, public corporations, and local governments where financial statements are main sources.
- General process:
  - Prepare bridge tables and derivation tables to convert national presentation to GFSM 2014 presentation.
  - Steps:
    a) Prepare bridge tables for income statement, balance sheet, and cash flow statement.
    b) Calculate changes between closing and opening balance sheet values for each asset and liability category.
    c) Prepare derivation tables for major GFS aggregates.
    d) Compile GFSM 2014 detail tables and summary statements.
    e) Verify the accuracy of the process.
- Classification and adjustments:
  - Classify all items in financial statements according to GFSM 2014.
  - Identify and exclude other economic flows (price changes/revaluations, other changes in volume of assets/liabilities) from revenue, expense, and transactions in assets/liabilities.
  - Adjustments needed where financial statements treat as revenue/expense items that GFSM 2014 classifies elsewhere (e.g., disposals/acquisitions of nonfinancial assets; loan disbursements/repayments; profits/losses on asset sales).
- Verification checks:
  - Major GFS aggregates should correspond to derivation table values; iterate between bridge, derivation, and detail/summary tables.
  - Net lending/borrowing should be consistent by alternative calculations; differences recorded as a statistical discrepancy.
  - Closing balance sheet category values should equal opening value plus transactions and other economic flows during the period.
  - Net change in financial asset Currency and deposits in Statement of Government Operations should equal net change in stock of cash in Statement of Sources and Uses of Cash (and institution’s cash flow statement).
- Specific treatment guidance:
  - Provisions: Not treated as liabilities in GFS; treat provisions as a proxy for a loss (another change in volume of assets).
  - Depreciation: GFS uses consumption of fixed capital; depreciation can be used as a proxy for consumption of fixed capital only if national accounts figures are unavailable and assets are valued at fair market prices.
  - Compilers should first attempt to obtain consumption of fixed capital figures from national accounts compilers; use depreciation as proxy only if those figures are not available.
- Caution on signs:
  - Pay close attention to signs when mapping financial statement items to GFS aggregates; incorrect sign conventions can create discrepancies (examples: negative revenue items treated as expenses; differing sign conventions for financing between GFSM 1986 and GFSM 2014).

*Italic: Source — Appendix I. Status of Implementation of Previous Technical Assistance Mission (implementation status as of November 2018).*

### Appendix IV. Consolidation of General Government GFS

### Appendix IV. Consolidation of General Government GFS

### Purpose of Consolidation
- Consolidation: method of presenting statistics for a set of units grouped together as if they constituted a single unit, involving elimination of all transactions and reciprocal stock positions between units in the subsector or sector being consolidated.
- Aggregation: simple addition of all data categories across subsectors; consolidation follows aggregation to eliminate internal interactions.
- Main analytical rationale:
  - Provides a more effective way of assessing the overall impact of government operations on the total economy and the sustainability of government operations.
  - Eliminates double counting of transactions or stocks among units being consolidated, producing aggregates not affected by internal interactions.
- Illustration (countries A and B):
  - At central government level both countries have identical values for revenue, expense and gross operating balance (GOB).
  - Country A: central government spends 110 directly and transfers 10 to local governments, which spend the whole amount.
  - Country B: central government spends 30 directly and transfers 90 to local governments, which spend the whole amount.
  - Unconsolidated aggregates:
    - Country A general government revenue 110; expense 130.
    - Country B general government revenue 190; expense 210.
  - After consolidation (eliminating internal transfers):
    - Both countries show general government revenue 100; expense 120.
    - The Gross Operating Balance (GOB) of the general government is unaffected by consolidation (GOB remains -20 for both countries).
- Consolidation scope:
  - Intra-governmental consolidation always takes place when compiling GFS for any given subsector.
  - Inter-governmental consolidation takes place only when compiling GFS for a sector that groups together more than one subsector (typical sectors: central government, general government, nonfinancial public sector, public sector).
  - Intra-governmental consolidation within a subsector is always done before inter-governmental consolidation.

### Key Items to Focus on Consolidation
- Practical objective: eliminate—in a consistent manner—transactions and positions that have a significant effect on final aggregates; perfection is not required.
- Priority flow items to identify and eliminate for central government and general government sectors:
  - transfers between government units;
  - transactions in financial assets and liabilities;
  - interest income/expense.
- Only eliminate intergovernmental taxes and purchases of goods and services when known to be large.
- Priority stock items:
  - loans;
  - debt securities.
- Transfers:
  - Grants between levels of government, and transfers to other institutional units (universities, hospitals, nonprofit institutions) are generally easy to identify and should be eliminated through consolidation.
  - For social security funds, general transfers between subsectors should be consolidated.
  - Direct payments by government as an employer to social security schemes are never consolidated; these are shown as if paid to employees and treated as compensation of employees.
- Financial transactions and holdings:
  - All direct loans between government units should be eliminated through consolidation.
  - All acquisitions of other governments’ securities in direct transactions should be eliminated.
  - Government debt instruments acquired on secondary markets should NOT be consolidated in transactions data (they should be consolidated in balance sheet or debt statistics).
  - Interest income from and expense to other government units should always be subject to consolidation.
  - Consolidated balance sheets should exclude reciprocal creditor-debtor relationships within the units being consolidated; priority given to loans and debt securities.
- Accounting practice:
  - Preferably identify transactions for consolidation in accounting transaction codes with the counterparty clearly identified.
  - If counterpart identification is missing, revising the chart of accounts to include it should be a priority to facilitate consolidation and automation.

### Rules of Thumb for Consolidation
- Discrepancies between counterparties may arise from timing, classification, accounting, or valuation differences; addressing them improves overall data quality.
- Practical rules of thumb when reconciling counterpart records:
  - "The 'one-side' rule of thumb." If there is convincing evidence from one party that a transaction took place, impute it to the other side even if counterpart records are absent; modify records for the unit where transactions cannot be directly identified.
  - "The 'top-down' rule of thumb." Rely on budgetary or central government accounts when they are more complete, timely, detailed, and accurate; impute transaction timing and value to counterparties (for example, local governments) when clear evidence exists that the central government made a transfer.
  - "The 'creditor is right' rule of thumb." For transactions in financial assets and liabilities, normally rely on creditor records as most reliable; for loans the creditor usually maintains the most complete records; for securities, especially bearer instruments, creditor records may be the only reliable source.
- Judgment and cost/benefit rule:
  - Effort and cost to identify an amount to be consolidated should be directly proportional to the expected amount and its impact on aggregates.
- Reconciliation tool:
  - Use a table presenting both parties to the flows or stocks and the amounts each party has recorded, then apply the rules of thumb to reach consistency while keeping balancing items unchanged.
  - Applying rules of thumb often leads to corrections to the accounts of one or both parties, modifying balancing items; these corrections are a benefit of the consolidation process, not a result of consolidation itself.
  - Consolidation is symmetric: once amounts are reconciled, eliminate the same amount for both parties.

### Impact on Aggregates and Balancing Items
- Main impact: consolidation affects the magnitude of aggregates by eliminating internal interactions.
- Consolidation adjustments do not impact balancing items in the GFS system when consolidation is applied symmetrically within a given table or statement.
- Example from table (unconsolidated vs consolidated figures preserved exactly):
  - Unconsolidated Country A:
    - Central Gov. revenue 100; Local Gov. revenue 10; General Gov. revenue 110.
    - Central Gov. expense 120 (110+10); Local Gov. expense 10; General Gov. expense 130.
    - GOB (R-E) central -200; local -20; general -200 -20.
  - Unconsolidated Country B:
    - Central Gov. revenue 100; Local Gov. revenue 90; General Gov. revenue 190.
    - Central Gov. expense 120 (30+90); Local Gov. expense 90; General Gov. expense 210.
    - GOB (R-E) central -200; local -20; general -20.
  - Consolidated Country A:
    - Central Gov. revenue 100; Local Gov. revenue 10; Consol. Column -10; General Gov. revenue 100.
    - Central Gov. expense 120; Local Gov. expense 10; Consol. Column -10; General Gov. expense 120.
    - GOB (R-E) central -200; local 0; general -20.
  - Consolidated Country B:
    - Central Gov. revenue 100; Local Gov. revenue 90; Consol. Column -90; General Gov. revenue 100.
    - Central Gov. expense 120; Local Gov. expense 90; Consol. Column -90; General Gov. expense 120.
    - GOB (R-E) central -200; local 0; general -20.

*Appendix IV. Consolidation of General Government GFS*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1zmbea2020001.pdf_