This paper on Bosnia and Herzegovina presents the report on the government finance statistics technical assistance mission in Bosnia and Herzegovina. Further on developing reconciliation processes, the mission provisionally finalized research to establish reconciliation procedures, without fully eliminating statistical discrepancies. On the compiling of nonbudgetary public sector units, the mission continued the development of compilation processes as started during the May 2018 mission. Considering the differences in outcomes on balance sheet transactions between the ‘old’ and the ‘new’ compilation process, further research is required to test the plausibility of these compilation processes and outcomes. The mission will liaise with IMF’s European Department on an appropriate implementation procedure in coordination with other reporting units in Bosnia and Herzegovina that are also revising fiscal surveillance to the Government Finance Statistics Manual 2014 framework. The mission succeeded in resolving statistical discrepancies—at least from accounting technical point of view.

Abstract

This paper on Bosnia and Herzegovina presents the report on the government finance statistics technical assistance mission in Bosnia and Herzegovina. Further on developing reconciliation processes, the mission provisionally finalized research to establish reconciliation procedures, without fully eliminating statistical discrepancies. On the compiling of nonbudgetary public sector units, the mission continued the development of compilation processes as started during the May 2018 mission. Considering the differences in outcomes on balance sheet transactions between the ‘old’ and the ‘new’ compilation process, further research is required to test the plausibility of these compilation processes and outcomes. The mission will liaise with IMF’s European Department on an appropriate implementation procedure in coordination with other reporting units in Bosnia and Herzegovina that are also revising fiscal surveillance to the Government Finance Statistics Manual 2014 framework. The mission succeeded in resolving statistical discrepancies—at least from accounting technical point of view.

Summary of Mission Outcomes and Priority Recommendations

1. A technical assistance (TA) mission was conducted by Mr. Deon Tanzer, the government finance statistics (GFS) advisor1 for South East Europe during the period February 18–22, 2019, to support the Bosnia and Herzegovina authorities, with a specific focus on the Republika Srpska (RS), in improving GFS for decision making. This mission was conducted within the context of the second phase of the Swiss State Secretariat for Economic Affairs (SECO) GFS capacity building project. The mission met with officials from the Ministry of Finance of the Republika Srpska (MOF RS), and the Republika Srpska Institute of Statistics (RZS).

2. The mission would like to thank the staff of the Republika Srpska authorities for their courtesy and willingness to share their knowledge with the mission. It is especially grateful to the staff of the MOF RS for their invaluable assistance in organizing the mission (see Section E for the list of officials met during the mission).

3. The mission was a follow-up to the October 2018 GFS TA mission, conducted by the GFS advisor. This mission’s focus was to finalize the reconciliation processes, to update the compilation process Highways, Roads, and the Development Bank of the Republika Srpska (IDBRS) for fiscal surveillance, to continue developing a compilation process for other nonbudgetary public sector units, to initiate the development of consolidation procedures, and to discuss classification questions in context of GFS compilation per the European System of National and Regional Accounts 2010 (ESA 2010) and the Government Finance Statistics Manual 2014 (GFSM 2014). In the Bosnia and Herzegovina context, the mission aims to support integrated Bosnia and Herzegovina GFS compilation to Eurostat and the IMF as sent by the Central Bank of Bosnia and Herzegovina (CBBH)2 to Eurostat. This includes the ESA 2010 transmission programme (ETP) and the GFS Yearbook tables.

4. On developing reconciliation processes, work was continued on reconciling the financial result with relevant items in the balance sheet [A], and on reconciling stock and flows of groups of balance sheet items [B]. These both serve as fundamental steps in the compilation of GFS. [A] is required to ensure that fiscal balance reconciles with financing, whereby the statistical discrepancy3 can be eliminated or minimized. [B] aims to clearly explain changes in the value of assets and liabilities through (recorded) transactions and other economic flows (without any differences), further ensuring consistency in the GFS framework. Since the October 2018 mission, the MOF RS tied some loose ends to resolve differences between the concept GFS data and fiscal surveillance data sent to the IMF’s European Department (EUR); the MOF also liaised intermittently with the GFS Advisor on resolving issues. This led to some additional adjustments in the recording of revenue, expenditure and financing transactions, whereby these data can now be fairly aligned.

5. Further on developing reconciliation processes, the mission provisionally finalized research to establish reconciliation procedures, without fully eliminating statistical discrepancies. A key challenge is the complex composition of the underlying accounting which has proven arduous to unravel. Nonetheless, the compilation and reconciliation procedures can be usefully implemented for quarterly and annual GFS reporting to Eurostat and the IMF’s Statistics Department as a step towards sound and consistent GFS.4 As an extension on this, these processes can be incorporated into the compilation of quarterly fiscal surveillance files sent to EUR. However, ongoing work is needed to rationalize the underlying accounting whereby the understanding of underlying economic data, composition of accounts and accounting consistency can be simplified. GFS compilation procedures can be simplified if a pure unit-by-unit compilation is feasible. See Section A. ‘Reconciliation Processes’.

6. On incorporating, separately, the compilation process of Roads, Highways, PE Forests RS and the IDBRS, the mission compiled 2015 and 2016 data per GFSM 2014 for these units using the same approach for nonbudgetary public sector units developed in previous missions. The compiled data were largely reconciled in line with reconciliation processes [A] and [B]. These data were also compared with data used in compiling the budget balance sheet as well as for fiscal surveillance transactions. The balance sheet data only show some correspondence, while the transactions data are more closely aligned, whereby NLB data in these files indicate somewhat different deficits than in fiscal surveillance. See Section B. ‘Compilation of Roads, Highways, PE Forests RS, and IDBRS’.

7. On the compiling of nonbudgetary public sector units, the mission continued the development of compilation processes as started during the May 2018 mission. These processes are being developed by translating financial statements of these entities into the GFSM 2014 and ESA 2010 frameworks conform a standardized process. The RZS, with the support of the MOF, compiled 15 additional units since the October 2018 mission. This mission reviewed the outcomes and completed the reconciliation process for some of these units. As next steps, the mission developed and documented detailed process steps to reconcile data. See section C. ‘Compilation of Extrabudgetary Units’.

8. On consolidation, the mission initiated the development of a GFS consolidation process and discussed key principals that are followed in this process. The GFS Advisor reviewed current consolidation in the fiscal surveillance file and recommended improvements in the approach. As an extension of this, the mission discussed the concept of attributing collected tax to the receiving unit. See Section I. D. ‘Consolidation’.

9. On other issues, since the October 2018 mission, an official from the MOF Accounting Department and an official from the MOF Treasury Department have left their respective departments and thereby are no longer active in the project. The MOF Accounting Department has employed a new official who is relatively new to GFS. The Head of Treasury department has replaced the Treasury Department official for this project.

Table 1.

Bosnia and Herzegovina, Republika Srpska: Priority Recommendations

article image

Detailed Technical Assessment and Recommendations

A. Reconciliation Processes

10. Reconciliation is a fundamental step in the compilation of GFS. The process entails translating the logic of accounting frameworks into the logic of the GFS framework. First, by verifying the link between revenue and expense flows with reciprocate balance sheet flows [A]. Second, by separately identifying, from accounting records, transactions and other economic flows, and thereafter verifying the link of these flows with the changes in the stock of the related balance sheet account [B].

Reconciliation Process [A]

11. The goal of reconciliation process [A] is to ensure that net lending (+) / net borrowing (-), also known as the fiscal balance, links with financing with the minimum difference possible, and thereby minimizing the so-called statistical difference. The Republika Srpska uses accrual accounting, whereby the financial result from the income statement is reconciled against a related account in the balance sheet the ‘Own Sources and Funds’ account.

Reconciliation Step [B]

12. The goal of reconciliation process [B] is to ensure that recorded transactions can be linked up with recorded stock values for distinct balance sheet items. In the context of accrual accounting used in the Republika Srpska, recording per the double accounting principle, this link should be discerned directly from the accounting records. This is because all transactions, albeit revenue, expense, acquisitions, repayments, revaluations, provisions, etcetera are recorded with an exact counterpart in the balance sheet. GFS would require revaluations to be recorded as other economic flows; unwind transactions like provisions and reserve accounting; and adjust valuations, where the accounting standards differ with GFS in this respect.

Continued Development

13. The advisor and the RS authorities evaluated the GFS reconciliation process [A], and [B] as developed in the concept GFS compilation file during and after the May 2018 mission. The file translates integrated financial statement data of the Republika Srpska budget and, step-by-step, compiles GFS data per the GFSM 2014 framework. The process is explained in the October 2018 report and was found to resolve reconciliation process [B], but still led to statistical discrepancies associated with reconciliation process [A].

14. In parallel, the file compares outcomes with the fiscal data sent to EUR using the Harmonized Reporting Template (HRT) per the GFSM 2001 framework. A key objective is to ensure GFSM 2014 outcomes are as sound as possible and, as a spin-off, to update the HRT compilation process with any new insights. The MOF, item-by-item, updated the HRT compilation of revenue and expenditure per the insights gained in this research, whereby the differences visible in Table 2 are nearly eliminated. After the HRT updates, revenue and expense data are now closely aligned, while net investment in nonfinancial assets data show a difference of between -4.4 and +7.0 million KM each year.

Table 2.

Republika Srpska Budget: Comparison Revenue and Expenditure, Revised ‘New’ GFSM 2014 Framework Data and ‘Old’ HRT data

(Million KM, 2014–2016) [CONCEPT DATA]

article image

15. This means that if HRT is revised per these insights the NLB will increase by 46.1, 0.9, and 23.4 million KM over 2014, 2015, and 2016 respectively. See Table 2. A significant source of changes stems from including 47 and 77 chart of accounts (COA) items listed in Table 2, but also from recording nonfinancial transactions using an indirect approach.

16. However, differences in financing data are not resolved. Past research as indicated that alignment is cumbersome to achieve, in part because of the differing approaches used— HRT uses data from an auxiliary table ‘Periodic Report on Expenditure by Accounting Funds’, while the concept file uses an indirect approach to determine transactions (See Box 1).

Calculating Balance Sheet Transactions Using the Indirect Approach

The indirect approach calculates financial transactions (TR) by subtracting other price changes (OPCs) and other volume changes (OVCs) from the net flow between the opening balance (OB) to the closing balance (CB) for each asset and liability account—See formula [1].

TR = CB – OB – OPCs – OVCs [1]

OPCs are identified by requisite revenue and expenditure COA items.

OVCs are determined for each balance sheet item where a government unit is newly consolidated into the Republika Srpska budget (or redefined outside the Republika Srpska budget).

In principle, the TR recorded in the direct approach should match with the indirect approach, but past research proved that the accounting data are complex to unravel and match neatly.

17. The differences in ‘new’ (per the GFSM 2014 framework) and ‘old’ (per the existing HRT) data can be explained by some key changes in the compilation process:

  • a. ‘New’ data are compiled using integrated financial statements. Hereby, transactions related to Roads, Highways, and PE Forests RS (RH&F) and the Development Bank of the Republika Srpska (IDBRS) are unraveled from the Republika Srpska budget data and thereby not yet included in this presentation (and will be included at a later compilation stage).

  • b. Balance sheet transactions are calculated indirectly per formula [1], instead of directly using transactions recorded under COA items 5, 6, 8, and 9. This impacts the calculation of transactions of financial (COA 6 and 9) and nonfinancial (COA 5 and 8) assets. The significant changes in outcomes (for example, consumption of fixed capital) require additional research into the underlying accounting.

  • c. The recording of revenue, expenditure, and financing transactions changed with the update of the bridge from the COA and to GFSM 2014 conducted during this project, whereby some items have changed classification. This includes COA line-items 47 and 77 (see also paragraph 18)

18. The update of the bridge from the COA to GFSM 2014 and ESA 2010 has adjusted the recording of COA line-items 47 and 77 COA. In the current HRT compilation process these items are excluded. Previous insights noted that these relate to specific accrual recordings and were recommended to be excluded in compiling HRT. However, the research conducted over the course of the project indicates that some items are revenue and expense items aligned with the GFSM 2014 framework, others relate to OPCs and should be recorded as part of balance sheet flows (see also paragraph 17. b), while items related to provisions should be unwound rather than excluded. Table 3 provides an overview of the classifications per ESA 2010 and GFSM 2014 frameworks.

Table 3.

Republika Srpska: Recording Revenue and Expense per GFSM 2014 and ESA 2010

article image
LegendOPC Other price changes TR Transactions# Provisions require adjustments to reflect time of recording* Record as revenue / expense transactions per GFSM 2014

19. A further outcome of this process is that balance sheet data can be compiled in line with the ESA 2010 and GFSM 2014 frameworks. An important note here is that valuation, recognition, and instrument coverage criteria used to compile Republika Srpska financial accounts have not been tested on the extent that these overlap with ESA 2010 and GFSM 2014. See Table 3 for 2014–2016 data. Importantly, these data currently exclude the balance sheets of the RH&F, IDBRS, local government, some social security funds, and extrabudgetary units. See Section B on compiling concept data that can be used to include and consolidate some of these entities to the general government presentation as a subsequent step in the compilation process.

Table 4.

Republika Srpska Budget, Balance Sheet per GFSM 2014,

(Million KM, 2014–2016) [CONCEPT DATA]

article image

20. As described in paragraph 16, a comprehensive stocks-and-flows reconciliation and presentation is possible for the comprehensive balance sheet. Hereby, it is possible to get insights into the position of the budgetary government, assets and liabilities. It is also possible to gain insights into additional underlying factors leading to the changes in the value of debt besides the contribution from NLB. See also Table 5.6

  • a. During 2015, the central debt increased by almost 200 million KM. This is largely explained by a net incurrence of debt of approximately 100 million KM and revaluation of loans debt— probably through exchange rate devaluations.

  • b. In 2016 a similar net incurrence of debt occurred and the inclusion of the pension fund significant contributed to the increase in the value of debt (recorded as OVC’s).

  • c. Considering the low level of NLB in 2015 and 2016, the net incurrence of debt is mostly explained by largely equal transactions in assets.

Table 5.

Republika Srpska Balance Sheet Stocks and Flows

(Million KM, 2015 and 2016) [CONCEPT DATA]

article image

By using the indirect method, it is notable that negative revaluations (as recorded in the income statement) are the key explainer for transactions in assets.

21. Considering the differences in outcomes on balance sheet transactions between the ‘old’ and the ‘new’ compilation process, further research is required to test the plausibility of these compilation processes and outcomes. Although this ‘new’ compilation process provides sound outcomes for nonbudgetary units, this process still has uncertain outcomes due to the complex nature of the accounting data from which these are compiled. Previous mission reports detailed steps to unravel the underlying balance sheet data, adjust specific line items, account for OVCs, record COA account 47 and 77, etcetera to take steps toward compiling GFS with plausible outcomes. Notwithstanding, further research will still be required.

22. The difficulty in the compilation process also highlights the need to evaluate the underlying accounting data. During the development process, the mission noted anomalies in the accounting practice when compiling the Republika Srpska budget financial statement:

  • a. The value of assets, legally owned by the Republic, are derived from special supplemental balance sheet data maintained by the RH&F are aggregated into the consolidated balance sheet. Considering that the RH&F are the economic owners, standard practice would be to consolidate fully the comprehensive financial statement of these units as opposed to the partial approach used here. After unravelling the balance sheet, current accounting practice does not indicate a one-on-one relationship between the Financial Result (AOP 330) and the closing balance of Financial Result of the current year (AOP 155). This is standard practice in accounting—by exception this is not the case. By not having this one-to-one relationship GFS compilation starts with a statistical discrepancy.

  • b. Further, balance sheet items Permanent Source of Funds (AOP 147) and Reserves (AOP 149) record changes each year. Standard accounting practice would see a limited set of identifiable recordings in these accounts, resulting from, among others, funds withdrawals or injections, or technical changes arising from accounting regulations. The GFS compiler needs insights into counterpart balance sheet flows to record adjustments elsewhere. Further research is needed to review current recording procedures in these accounts. A well-documented list of recordings, explaining the basis for recording and the counterpart balance sheet recording will inform GFS compilation greatly.

  • c. The RS Tax authority is developing and recording accrual estimates for RS taxes. Due to constant changes in laws and regulations, as well as ongoing development work, the estimation techniques are also evolving. The outcomes recorded in account 7719 are thereby uncertain and are currently unwound from accrued revenue. Linked to this, tax receivable recorded on the balance sheet (e.g. AOP 018, 019, and 055) are significantly high. Fluctuating between 50 percent (2014) and 100 percent (2016) of the Republika Srpska taxes. In comparison, VAT receivable are 1 percent of VAT revenue. Annual transactions in other accounts receivable are thereby unusually high. Given the uncertainty of tax data it is recommended to use a cash only recording for GFS compilation. This will mean that not only will revenue be recorded on a cash basis, but all associated other accounts receivable will need to be unwound from the balance sheet. As an alternate, an often-used approach in compiling GFS is the so-called ‘time-adjusted’ cash method. Hereby accrual estimates are based on modifying the month a tax is collected with a specified time delay from the period it is earned.

23. The concept GFS compilation file uses four-digit COA data from the financial statement to compile GFS tables. Further refinements to the process will be available once six-digit COA data are used:

  • a. Provide a clearer breakdown of transactions and balance sheet classifications. The current compilation file includes a limited set of manual adjustments to breakdown some recordings.

  • b. More information on counterpart transactions and positions (relevant for consolidation).

  • c. A better combination between OVCs & OPCs from the income statement with the requisite balance sheet items. The current process uses assumptions to combine codes. This does not impact total transactions of either financial assets or liabilities, rather on each individual balance sheet item.

24. The mission will liaise with IMF EUR on an appropriate implementation procedure in coordination with other reporting units in Bosnia and Herzegovina that are also revising fiscal surveillance to the GFSM 2014 framework.

Table 6.

Additional Recommendations

article image

B. Compilation of Roads, Highways, PE Forests RS, and IDBRS

25. The mission compiled concept GFS data per GFS 2014 of RH&F and IDBRS using the same approach as extrabudgetary units (see mission October 2018 report). The data covers 2015 and 2016. The financial statement data of these units are available per the same Chart of Accounts (COA) template as used for the extrabudgetary units. Thereby the bridge from COA to GFSM 2014 is the same.

26. Having data in template form allows following a largely standardized procedure to compile these data unit-by-unit, year-by-year. Other recent years can thereby be compiled by copy-pasting data from the respective templates into the compilation file, whereby outcomes are calculated automatically. A review of unit-by-unit outcomes is still required to research and address any special recordings of economics events or (one-off) anomalies in the accounting.

27. The GFS Advisor developed an Excel-based compilation procedure for each of these units, whereby data are compiled in separate annual sheets following a layout akin to the GFSM 2014 framework. Data are recorded in the worksheet using a stacked layout in the following top-to-bottom order: (i) nonfinancial transactions (TR); (ii) closing balance (CB); (iii) balance sheet transactions (BS_TR); (iv) other economic flows (OEF); (v) and adjustments (Adj). Any adjustments are recorded using a double accounting principle—equal debit and credit transactions.

Table 7.

Sample—Stacked Presentation GFSM 2014 Framework in Compilation File, Descriptions and AOP Codes are for Illustration Purposes

article image

28. A dashboard is visible at the top of the table to view high-level aggregates and assist in compiling, especially to eliminate the statistical discrepancy between NLB and financing. The outcomes in the concept files indicate that the statistical discrepancy can be resolved for almost each unit, although a number require more in-depth research into accounting practices to resolve the statistical discrepancy.

Table 8.

Dashboard Presentation IDBRS

(Thousands KM, 2015 and 2016) [CONCEPT DATA]

article image

29. The data compiled directly from financial statements are discernibly more accurate than those used to consolidate Republika Srpska budget data, not only because these follow (to a large extent) standard accounting practice. The financial statements offer detailed revenue and expense data, as well as more detailed balance sheet data. The link between profit or loss for the year is directly discernable with the requisite item in the Capital Account. Hereby, it is possible to derive a statistical discrepancy of 0 as is evident in Table 9. However, more research is required in translating the financial statement of PE Forests RS into GFS since statistical discrepancies are visible. From a cursory view, the Capital Account reflect changes from year-to-year that are not easily explainable.

Table 9.

Dashboard Presentation (Dis)Aggregated RH&F

(Thousands of KM, 2016) [CONCEPT DATA]

article image

30. Financial statement balance sheet data only compare to a degree with the data used to consolidate into the Republika Srpska budget balance sheet. The value of assets, legally owned by the Republic, are derived from special supplemental balance sheet data maintained by the RH&F are aggregated into the consolidated balance sheet whereby only a subset of the balance sheet items—in name only—is incorporated into the core budget balance sheet and no one-on-one match exists between these line-items. Further, off balance sheet items of Forests are recorded as ‘Other natural assets’ (AOP 073) in the core budget balance sheet. The own source funds used for the core budget account are derived as the increase/decrease incorporated assets. Last, the financial profit of the year does not compare directly with the requisite balance sheets item.

Table 10.

Comparison RH&F Source Data and Data Consolidated into the Republika Srpska Budget Balance Sheet

(Thousands of KM, 2016) [CONCEPT DATA]

article image

31. Previous missions recommended the need to unwind the Republika Srpska budget financial statement of the balance sheet data from RH&F used there. As evidenced in this comparison the data consolidated into the Republika Srpska balance sheet do not reflect the true economic value (rather the construed value from a legal ownership perspective) of the combined RH&F balance sheets and thus warrant the previous missions’ recommendation. The next step would be to use the unwound core budget data and the add in the comprehensive RH&F as a second step, as presented in Table 2 thereby reflecting the economic value of all assets and liabilities. As a next step, these data would also require consolidation on items, such as on-lending, grants provided, and interest. GFS compilation would be supported even further if the Republika Srpska budget data can be compiled using comprehensive financial statements, unit-by-unit (at the most disaggregated level feasible).

32. The issue of economic ownership was raised concerning Highways. Economic ownership differs from legal ownership in that the owner of an asset is the unit that holds the risks and rewards of an assets. This means that economic owner is the unit that is entitled to revenue an asset may generate and is responsible for maintaining an asset, but also will incur any gains or losses on the assets due to price changes, accidents or disasters (keeping aside factors such as insurance). The oral information provided by MOF officials indicated that Highways RS holds the risks and rewards of the land and structures that the highways are built on. Hereby all physical assets (land, drainage, roadbed, road shoulder, bridges, tunnels, asphalt, etcetera) are inextricably connected and operated, maintained and owned by Highways RS. Further, past accounting practice recorded the part of value of physical road structures as ‘other intangible assets’ (AOP 006) which seems implausible give the nature of the assets. These should be recorded under the requisite nonfinancial asset.

33. A further issue of economic ownership covers the core assets of PE Forest RS since these are recorded as off-balance sheet items, again reflecting a separated legal and economic ownership. The data compiled in Table 11 thus does not contain the associated data, whereby the net investment in nonfinancial assets excludes these transactions for PE Forests RS.

Table 11.

Aggregated Republika Srpska Budget Data with RH&F—Nonconsolidated

(Thousands of KM, 2016) [CONCEPT DATA]

article image

C. Compilation of Extrabudgetary Units

34. Several public sector units form part of general government, but not the budget, per ESA 2010 and GFSM 2014. These extrabudgetary units therefore need to be compiled as part of GFS from financial statements that are reported to the Agency for Intermediation, Informatics and Finance-APIF. Likewise, public sector units outside general government report financial statements to the APIF.

35. To compile GFS, a procedure is required to translate accounting financial statements into the ESA 2010 and GFSM 2014 frameworks. During the May 2018 mission, the GFS Advisor, together with the RZS and MOF RS started developing a concept GFS compilation file. This in line with a compilation procedure discussed during March 2018 SECO-funded GFS workshop held in Ljubljana, Slovenia. The procedure follows a systematic approach to unravel a(n) (audited) financial statement and translate line-by-line income statement and balance sheet items into the ESA 2010 and GFSM 2014 frameworks. It aligns fundamental accounting principles with the frameworks in a step-by-step approach. It, for instance, adjusts for provisions, exchange rate gains, and impairments that are treated differently in accounting and in GFS. See May 2018 mission report Section B. ‘Compilation of Nonbudgetary Public Sector Units’ and Appendix II ‘Proposed Compilation Process for Extrabudgetary Units’ of this report for a wider explanation.

36. With this backdrop, the mission continued work on developing compilation processes for extrabudgetary units. RZS RS compiled 15 more units since last mission with mixed outcomes. The mission thus reviewed some of these units, discussing issues one-by-one to strengthen understanding of the procedure. A key goal in reviewing each unit was to strengthen the understanding of the largely standardized reconciliation procedures for the Republika Srpska-specific reporting templates, as developed in the October 2018 mission.

37. Over-and-above the procedures described in the May 2018 mission report, the mission articulated specific procedures used to reconcile statistical discrepancies of extrabudgetary units. See Box 2.

Additional Procedure to Resolve the Statistical Discrepancy of Extrabudgetary Units

article image
article image

38. By following this procedure, the mission succeeded in resolving statistical discrepancies—at least from accounting technical point of view. However, these files are still at a conceptual phase that have been used to illustrate basic compilation and resolving the statistical discrepancy. The next steps will be to convert these into compilation files while also reviewing outcomes, gaining insights into the economic meaning of these data and analyzing time series developments. To move towards a compilation file the following steps are envisaged:

  • a. Improve the layout—for example, the stacked layout used for Roads, Highways and Forest.

  • b. With a stacked layout, adjustments can be recorded at the bottom.

  • c. Develop an output file whereby year-by-year, unit-by-unit data can be aggregated in a standardized manner.

D. Consolidation

Methodology

39. Consolidation is a method of presenting statistics for a set of units (or entities) as if they constituted a single unit. In the GFS framework, the data presented for a group of units are consolidated. Two forms of consolidation are required. Intrasectoral consolidation is consolidation within a particular subsector to produce consolidated statistics for that particular subsector—for example, within the central government subsector. Intersectoral consolidation is consolidation between subsectors of the general government to produce consolidated statistics for a particular grouping of government units—for example, consolidation between central, state, and local governments, to present the consolidated general government (GFSM 2014 paragraphs 3.153–3.136).

40. In the case of the Republika Srpska, 2 key themes require consolidation, grants and on-lending. In the case of grants, government units provide resources for either current or capital expenditure to other government units (generally in the form of cash transfers and, exceptionally, in the form of in-kind transfers). These transfers flow from the Republika Srpska budget to municipalities, funds, and extrabudgetary units, from municipalities to extrabudgetary units, etcetera. In the case of on-lending, the Republika Srpska borrows funds from foreign governments or agencies and international organizations and lend these funds to devolved government units, such as Roads, Highways or municipalities. The devolved government unit is thus responsible for repaying the loan and interest to the foreign entity through the Republika Srpska budget.7

41. When consolidating, transactions and positions between units are cancelled out against each other, whereby the combined units are presented as a single unit. See g for a stylized example of consolidation:

  • a. In this example, units A, B, C, & D make up subsector I. Additional units E, F, and G make up the rest of the sector II.

  • b. Within subsector I, unit D pays grants of 100 to each of units A, B, and C, while unit C pays 40 grants to unit A, unit A pays 10 grants to unit B, etcetera.

  • c. Units A, B, and C also receive grants from units E, F, & G, while unit D receives 10 grants from unit E, and unit B pays 40 to unit F.

  • d. Units E, F, and G pay and receive grants among themselves.

  • e. In total units A, B, C, and D pay and receive grants of 370 within subsector I, and all units pay and receive grant of 880 within the wider sector II.

  • f. When consolidating among units, receipts for subsector I are reduced from a gross value of 710 to a consolidated value of 340, payments by subsector I are reduced from a gross value of 420 to 50.

  • g. When consolidation among units, receipts for and payments by sector II are reduced from a gross value of 880 to 0.

Table 12.

Stylized Example Consolidation Between Units within Subsector I and Sector II

article image

42. During the mission the advisor provided a conceptual description on consolidation and procedures to consolidate in the context of GFSM 2014. The MOF currently consolidates various entities as part of its presentation of the Republika Srpska budget, the Pension Fund, and special supplemental balance sheet data maintained by the RH&F. As discussed in Sections A and B, it is more appropriate for GFSM 2014 compiling purposes to compile Republika Srpska government data unit-by-unit and consolidate as a last step. Crucially this will allow the GFS compiler to internally reconcile each unit, before consolidating with other units. This will entail compiling the Republika Srpska budget (in its most stripped-down version), the Pension Fund, Roads, Highways, Forests, IDBRS (and the six funds it manages), municipalities, agencies, social security funds, extrabudgetary units, etcetera one-by-one.

43. Once all relevant units have been compiled and reconciled, these units will then be consolidated covering the main themes grants, loans and the associated interest. Other themes such as purchases of goods and services (among government units) and payments of (property) taxes to other Republika Srpska government units can be regarded as sufficiently immaterial not to consolidate.

article image

Tax attribution

44. Related to consolidation and grants, the mission also discussed the importance of attributing taxes received to the collecting unit. Research on this is relevant in Bosnia and Herzegovina across all governments because a limited set of tax collecting agencies, such as the Tax Administration of Republika Srpska (TA RS) and the Indirect Tax Authority of Bosnia and Herzegovina (ITA) collect taxes on behalf of other government bodies. For instance, the TA RS collects personal income and corporate tax in the Republika Srpska, that is subsequently divided by a fixed key among the Republika Srpska budget and municipalities. It is important to analyze the underlying information on whether the tax received can be attributed directly to municipalities or are collected by the TA RS and subsequently transferred as grants to the municipalities.

45. GFSM 2014 paragraphs 3.33 and 3.34 explain the issue of attribution. “In some cases, one government unit collects taxes and then transfers some or all of them to another government unit. Depending on the arrangement, the taxes passed on to the second government unit may be reassigned as tax revenue of that unit or they can be recorded as tax revenue of the collecting unit and a grant from that unit to the second government unit. A tax is attributed to the government unit that: (i) exercises the authority to impose the tax

(either as a principal or through the delegated authority of the principal); and (ii) has final discretion to set and vary the rate of the tax.” See also GFSM 2014 paragraphs 3.35–3.40 for further explanations.

46. The mission did not analyze the underlying information but will address this question further during a follow-up mission.

E. Officials Met During the Mission

article image

Appendix I. The Recording of Provisions

1. In financial accounting, a provision is a present liability of an entity arising from past events, the settlement of which is expected to result in an outflow, but with an uncertain amount or timing. For instance, if a policy decision is made to retrench staff in stages during future years, the accountant records a provision of the total estimated retrenchment costs in the period the decision is made. See Figure 1 for an illustration.

  • a. The amount recorded is an estimate of the future cost (per accounting regulations).

  • b. This amount is recorded as an accrued expense ① and a provision liability ② (on the balance sheet) in the period the decision is made.

  • c. In the period the retrenchment costs are actually paid, an outflow is recorded— decreasing the provision liability ③.

  • d. Periodically, the estimate is reevaluated. If the estimate is evaluated to be higher, an accrued expense and provision liability is recorded again per the adjustment amount— see (b) above. If the estimate is evaluated to be lower, a provision release ④ is recorded as part of income, decreasing the provision liability ⑤.

Figure 1.
Figure 1.

Illustration Adjusting and Unwinding of Provisions from Financial Statements Recording to GFS Recording.

Citation: IMF Staff Country Reports 2020, 075; 10.5089/9781513537252.002.A001

2. To comply with ESA 2010 and GFSM 2014, an adjustment of provisions is accordingly recorded. The recording of provisions in the financial accounts differs with the recording in ESA 2010 and GFSM 2014 mainly due to the time of recording. Whereas provisions are future uncertain expenses recorded now in financial accounts, ESA 2010 and GFSM 2014 only records transactions during the period or “at the time economic value is created, transformed, exchanged, transferred, or extinguished” (see ESA 2010 paragraph 1.101 and GFSM 2014 paragraph 3.62). See Figure 1 for an illustration of the process to unwind and provision recording into GFS.

  • a. This requires ‘unwinding adjustments’ of accrued provision expense and releases of provisions recorded as income. At the same time, outflows recorded in the balance sheet should be recorded as expense accrued in the current period.

  • b. The resulting adjustment in revenue leads this line-item to be reduced to zero ⑥.

  • c. The resulting adjustment in expense leads this line-item first to be reduced to zero, and subsequently to be equal to the outflow recorded in the provision liability ⑦.

  • d. The provision liability disappears from the balance sheet.

1

The GFS Advisor for South East Europe is hosted by the Center for Excellence in Finance, Ljubljana, Slovenia.

2

The CBBH currently compiles consolidated Bosnia and Herzegovina GFS data sent to the IMF and sent to Eurostat on an experimental basis. The project is geared toward including and integrating sound GFS compilation processes of the Brčko District, the Federation of Bosnia and Herzegovina, the Republika Srpska, and the national institutions into a unified process. Currently no agreement is in place among these compiling institutions on a future division of labor, integration of processes, and the associated quality checks.

3

In the sequence of accounts as described by the System of National Accounts 2008 (SNA 2008), ESA 2010, and the GFSM 2014 framework, the government deficit or surplus is equivalent to the concept ‘net lending (+)/net borrowing (–)’ (NLB). In fiscal analysis, NLB is also referred to as the fiscal balance. Conceptually, the deficit or surplus should equal the concept —the net acquisition of financial assets minus the net incurrence of liabilities. Any difference between these two concepts is called the statistical discrepancy.

4

See footnote 2 on current and potential future compilation responsibilities.

5

This is a recommendation still pending completion of Phase 1 of the SECO-funded GFS implementation project.

6

These data are subject to qualification because OPCs for assets need to be assessed further and will probably reallocated to other assets with additional research.

7

On-lending per the RS Law on Borrowing, Debt and Guaranties means “Indirect debt of the Republika Srpska is the debt incurred on behalf and for the account of the debtors (local government units, social security funds, the Investment Development Bank and the funds managed by it, public enterprises and public sector institutions) to be repaid by the Republika Srpska from payments made by the debtor for its relevant debt.

Bosnia and Herzegovina: Technical Assistance Report-Government Finance Statistics Mission
Author: International Monetary Fund. Statistics Dept.