1.1.1. Coverage of Institutions (Basic)

7. Armenia’s public sector comprises at least 3,464 units of various legal forms. The precise number is not known since a comprehensive list of public units doesn’t exist. As shown in Table 1.2 the units are distributed in the following subsectors:

• Central government, which comprises 1,961 units in total, including 66 central government budgetary units¹, and 1,895 SNCOs² which are classified as extra-budgetary funds (EBFs). Social security revenues and expenditure are integrated in the state budget.

• Local government, which comprises at least 1,201 units, including 502 municipalities and several thousands³ of EBFS – MNCOs, of which 699 units conduct some limited commercial activities in addition to non-market activities;

• Public nonfinancial corporations, which comprise around 288 corporations controlled by the central government, of which 241 are joint-stock companies. The number of corporations controlled by local government is not known; and

• Public financial corporations, which comprise 14 units, including the National Central Bank (NCB) and its 10 subsidiaries, and three state corporations controlled by the state (Armenia’s Development and Investment Corporation, Deposit Guarantee Fund of Armenia, and Export Insurance Agency of Armenia).

Table 1.2.

Armenia: Public Sector Institutions and Finances, 2016

(Percent of GDP, unless otherwise stated)

Source: GFSM 2014 report, Armenian authorities. Financial statements of public corporations (data for Central Bank include its 10 subsidiaries), and IMF staff estimates. Note: Numbers for “Revenue” and “Expenditure” don’t add up due to consolidation of inter-sectoral transactions.

Table 1.2.

Armenia: Public Sector Institutions and Finances, 2016

(Percent of GDP, unless otherwise stated)

				Number of entities	Revenue	Expenditure	Balance	Intra-PS expenditure	Net expenditure
Public Sector				3,464	28.1	34.8	-6.8	0.0	34.8
	General government			3,162	24.9	30.3	-5.4	-1.1	29.2
		Central government (consolidated)		1,961	23.5	28.9	-5.5	-2.3	26.7
			State Budget	66	23.1	28.5	-5.5	-5.2	23.4
			State Non-Commercial Organizations	1,895	3.3	3.3	0.0	0.0	3.3
		Local government (consolidated)		1,201	2.6	2.5	0.1	0.0	2.5
			Marzes, towns, and villages	502	2.4	2.3	0.1	0.0	2.3
			Municipal Non-Commercial Organizati	699	0.2	0.2	0.0	0.0	0.2
	Nonfinancial public corporations			288	3.8	5.0	-1.2	-0.1	4.9
	Central Bank			1	0.6	0.7	-0.1	0.0	0.7
	Other financial public corporations			13	0.0	0.0	0.0	0.0	0.0

Source: GFSM 2014 report, Armenian authorities. Financial statements of public corporations (data for Central Bank include its 10 subsidiaries), and IMF staff estimates. Note: Numbers for “Revenue” and “Expenditure” don’t add up due to consolidation of inter-sectoral transactions.

View Table

Table 1.2.

Armenia: Public Sector Institutions and Finances, 2016

(Percent of GDP, unless otherwise stated)

				Number of entities	Revenue	Expenditure	Balance	Intra-PS expenditure	Net expenditure
Public Sector				3,464	28.1	34.8	-6.8	0.0	34.8
	General government			3,162	24.9	30.3	-5.4	-1.1	29.2
		Central government (consolidated)		1,961	23.5	28.9	-5.5	-2.3	26.7
			State Budget	66	23.1	28.5	-5.5	-5.2	23.4
			State Non-Commercial Organizations	1,895	3.3	3.3	0.0	0.0	3.3
		Local government (consolidated)		1,201	2.6	2.5	0.1	0.0	2.5
			Marzes, towns, and villages	502	2.4	2.3	0.1	0.0	2.3
			Municipal Non-Commercial Organizati	699	0.2	0.2	0.0	0.0	0.2
	Nonfinancial public corporations			288	3.8	5.0	-1.2	-0.1	4.9
	Central Bank			1	0.6	0.7	-0.1	0.0	0.7
	Other financial public corporations			13	0.0	0.0	0.0	0.0	0.0

Source: GFSM 2014 report, Armenian authorities. Financial statements of public corporations (data for Central Bank include its 10 subsidiaries), and IMF staff estimates. Note: Numbers for “Revenue” and “Expenditure” don’t add up due to consolidation of inter-sectoral transactions.

8. Armenia’s public sector expenditures amount to 34.8 percent of GDP in 2016. Table 1.2 summarizes the distribution of public resources across the different subsectors of the public sector in 2016 and shows that:

• General government net expenditure accounted for 29.2 percent of GDP, of which around 91 percent was spent by the central government and 1 percent through local governments. Expenditures of non-commercial organizations of around 3.5 percent of GDP are almost entirely financed through government transfers;

• Public corporation net expenditure accounted for additional 5.6 percent of GDP, the vast majority of which was spent by nonfinancial corporations.

9. There is no single report that provides complete coverage of either general government or public sector activity. The budget execution reports consolidate budgetary central and local governments, however the activities of EBFs (SNCOs and MNCOs) are not covered. While the annual GFSM 2014 statistical reports present a more comprehensive sector coverage because they also include SNCOs, they do no capture MNCOs whose reports are not published. The NSS compiles and publishes annual SNA 2008 based on national accounts for general government, though the approach applied for sector definition is not fully compliant with international statistical standards.

10. Expanding the institutional coverage of Armenia’s fiscal reports to the entire public sector would increase the deficit by 1.3 percent of GDP and would have a material impact on revenue and expenditure. Including SNCOs and MNCOs will add 3.5 percent of GDP to both revenue and expenditure but will not change the deficit. Further expanding the coverage to the public sector by including public corporations, will add 4.4 percent of GDP to revenue and 5.7 percent of GDP to expenditure on an accrual basis, resulting in an overall public sector deficit of 6.8 percent of GDP.

11. A number of loss-making public corporations in Armenia would likely be classified in the general government sector. Out of 199 public corporations monitored by the government, 49 units have cumulated losses (negative retained earnings) of 2.8 percent of GDP. These are mainly corporations in the energy and water sectors and to a lesser extent in transport, defense, sport, and agriculture sectors. Their expenditures amounted to 3.7 percent of GDP (including net investment in non-financial assets of 1.8 percent of GDP) and their deficit to

1.9 percent of GDP in 2016 on an accrual basis. Figure 1.1 illustrates the coverage of government and public sector expenditure based on the current sectorization, and indicates the magnitude of unreported government expenditure if the loss-making corporations were consolidated within the government.⁴

View Full Size

Coverage of Public Sector Institutions in Fiscal Reports

(Percent of expenditure at each level)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Armenian authorities and IMF staff estimates.Note: “Not Reported” refers to expenditures of units not consolidated in summary fiscal (budget execution) reports.

• Download Figure
• Download figure as PowerPoint slide

1.1.2. Coverage of Stocks (Basic)

12. Armenia publishes data on stocks of the state budget deposits and debt. The annual State Debt Report provides comprehensive data on the state debt by residency, currency, maturity, and type of interest rate. It also provides data on the state budget cash deposits as reported in the Treasury State Account (TSA).

13. There is no single report subsector that provides an overview on all types of government assets and liabilities for any public subsector. As explained above, data on most state financial assets and liabilities of 25.2 and 51.8 percent of GDP respectively, are spread across various government agencies and their departments, but no comprehensive statement exists. Adding liabilities related to PPPs and pensions (as described below) will increase state liabilities to 68.1 percent of GDP. Single SNCOs report balance sheet data, but individual data on assets and debt of 3.8 and 2.9 percent of GDP respectively are not aggregated in any report. According to the authorities, the scope of local governments financial assets and liabilities is limited. They do not provide loans and have not incurred any debt due to strict borrowing rules.

14. Significant gaps are observed in data availability for non-financial assets and liabilities and their valuation:

• Only partial data on the state non-financial assets subject to privatization (2.8 percent of GDP) are available from the State Property Management Department (SPMD);

• Another portion of the state non-financial assets (17.4 percent of GDP) is reported by individual line ministries in their balance sheets,⁵ mostly assets used for operational activities;

• Data on the main government non-financial assets, such as infrastructure assets, natural resources, historical buildings, land, etc., are not available. A noticeable portion of these assets are deemed to be owned by local governments and operated by MNCOs (land, roads, water walls, pipelines, bridges, and agriculture assets). Ministry of Territorial Administration plans to introduce an inventory of non-financial assets for municipalities.

• Non-financial assets held by SNCOs (0.8 percent of GDP) and public corporations (11.8 percent of GDP) are reported in the financial statements of individual units but an aggregated report is not available.

• Non-financial assets and liabilities related to PPPs contracts (10.2 percent of GDP), explicit liabilities of the military pension scheme (6.1 percent of GDP) and implicit liabilities of the pay-as-you go social security pension scheme (51.5 percent of GDP)⁶ are not reported by the government.

15. Addressing these gaps would provide a more comprehensive view of the public fiscal stance. As shown in Table 0.1 and Figure 1.2, consolidated public sector asset holdings and liabilities, excluding implicit pension liabilities, are estimated to have been at least 102.3 percent of GDP and 101.9 percent of GDP, respectively, in 2016. Public sector net worth and net financial worth are estimated to have been 0.3 percent of GDP and -56.5 percent of GDP, respectively. Considering the estimated implicit pension liabilities,⁷ public sector net financial worth will account for –108 percent of GDP. These figures, however, do not reflect the government owned infrastructure assets, land, subsoil assets, and historical buildings as not even indicative data on their value are available to the Armenian government. In addition, the valuation of non-financial assets most probably does not reflect market prices. A proper valuation and adding the missing non-financial assets would likely have a noticeable positive impact on the government and public net worth.

View Full Size

Public Sector Balance Sheet Coverage in Fiscal Reports, 2016

(percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: GFSM 2014 report, Armenian authorities. Financial statements of public corporations, and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

16. Armenia’s net worth sits in the mid-range of other countries and gross liabilities are relatively low (Figure 1.3 and 1.4). This in part reflects the fact that Armenia does not have any state-owned banks and does not have a government controlled defined contribution pension scheme.

View Full Size

Public Sector Gross Liabilities in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: IMF staff estimates, Fiscal Transparency Evaluations (Portugal 2012, Ireland 2012, UK 2013, Finland 2015, Brazil 2016, Tunisia 2015, Romania 2014, Albania 2015, Peru 2015, Philippines 2014, Tanzania 2015, Russia 2014, Turkey 2013, Kenya 2014, Uganda 2016, Georgia 2016, Mozambique 2015, and Guatemala 2016).

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Public Sector Net Worth in Selected Countries

(percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: IMF staff estimates, Fiscal Transparency Evaluations

• Download Figure
• Download figure as PowerPoint slide

1.1.3. Coverage of Flows (Basic)

17. Fiscal budget execution reports and GFSM 2014 based statistical reports cover cash revenues, expenditures, and financing, but do not include accrued transactions and other economic flows. The most comprehensive picture of the general government flows is presented in the annual statistical reports for consolidated general government sector. While the GFSM 2014 reports are on a cash basis, the national accounts include a few accrual elements such as consumption of fixed capital.

18. Some government operations are not recorded in line with international standards resulting in underestimating the government deficit. The current national rules do not reflect the economic substance of some transactions as required by international statistical standards. For example, capital injections to loss-making public corporations should be treated as an expenditure and not as an acquisition of shares and equities. Dividends are treated as a revenue in the budget execution reports, but it should only be the case if they are paid out of the profit coming from the operational activities of the corporation and not from revaluation gains or reserves.

19. Recognizing of non-cash transactions not covered in the reports may have a noticeable impact on the general government deficit on an accrual basis. While delayed payments in trade credits and other payables seems to be insignificant, other non-monetary operations that would need to be included will affect government’s fiscal results. For example, inclusion of the capital investments performed through PPPs where the government is bearing most of the associated risks, need to be counted as a government investment in non-financial assets. This will significantly increase the government deficit at the time the construction and other related works are performed. Also, capturing of the net accrual pension entitlements of public military employees will impact the deficit as well as financing.

1.1.4 Coverage of Tax Expenditures (Good)

20. Armenia publishes regular estimates of the revenue loss from tax in the annual budget documentation. Since 2015, estimated tax expenditures are published in the explanatory note of the annual budget documentation including estimates for the three main taxes, VAT, corporate income tax and personal income tax (70 percent of total taxes). VAT expenditures with around 40 lines of exemptions in the tax code are presented by sectors (e.g. health, education, financial and insurance sectors) and by other broad categories of exemptions, but the breakdown of PIT and CIT expenditures includes only privileges on agriculture, dividends, and VAT thresholds. The State Revenue Committee plans to expand its reporting on individual tax exemptions in 2019.

21. There is no control on, or budgetary objectives for, the size of tax expenditures -such controls or objectives could help strengthen the revenue base. Legislation does not include a requirement to report tax expenditures nor to assess the quantitative impact of new legislation on tax exemptions. Tax expenditure accounts for almost 7 percent of GDP (Figure 1.5) and one third of total collected taxes and are among the highest compared with other countries. The government plans to reduce the size of tax expenditure. This could be done through a budgetary objective or limit on its size, and reporting on its implementation.

View Full Size

Tax Expenditures in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: IMF staff estimates, Fiscal Transparency Evaluations.

• Download Figure
• Download figure as PowerPoint slide

1.2.1. Frequency of In-Year Fiscal Reporting (Advanced)

22. In-year state budget and local government reports are published monthly, within a month. These budget execution reports present revenues, expenditure, and financing data and are published within 30 days of the end of month. Also, monthly data on the state budget debt are published within a month. Quarterly budget execution reports are published within one month at the end of each quarter. These reports provide more detailed revenue, expenditure, and financing data by economic categories, and expenditure data according to the functional classification. Armenia is among the half of the countries that publish monthly reports within a month (Figure 1.6).

View Full Size

Periodicity and Timeliness of Budgetary Reports by SDDS Subscribers

(Percent of Countries)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Special Data Dissemination Standards (SDDS) Database; General Data Dissemination Standards (GDDS) Database (2017)

• Download Figure
• Download figure as PowerPoint slide

1.2.2 Timeliness of Annual Financial Statements (Advanced)

23. Annual budget execution reports are published within 4 months of the following year. They are subject to an audit opinion of the Chamber of Audit (Indicator 1.4.2). The annual report provides data for the state and local government and some indicators on the performance of SNCOs based on different classifications (Indicator 1.3.1). The report also presents further details on grants and subsides, comparative tables by sectors, specific projects, detailed information of the government debt and loans, and other details. As explained above, annual financial statements are not yet fully developed. Armenia is among the group of countries that publish their annual fiscal report within 6 months (Figure 1.7). The sectoral national accounts for the general government compiled by the NSS are published with a delay of 18 months.

View Full Size

Timeliness of Annual Fiscal Reports in Selected Countries

(Number of countries)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: PEFA Scores Data Set (2005–16).

• Download Figure
• Download figure as PowerPoint slide

1.3.1. Classification (Advanced)

24. Fiscal reports provide information on revenue and expenditure by administrative, economic, functional, and program classification in line with international standards. The MoF’s quarterly budget execution reports, covering State Budget entities and extra-budgetary accounts, includes administrative, functional, and economic classification of spending. The annual execution report, which consolidates State Budget entities, extra-budgetary accounts and local governments, extends reporting by programs, defined at the sublevel of functions. Functional classification follows the United Nations’ Classification of Functions of Government (COFOG), while economic classification is in accordance with GFSM 2014. Revenues are aggregated over three types: tax revenues and state duties, official grants, and other income. The same revenue and spending data are published in the NSS fiscal statistics and presented by economic and functional categories.

25. Some current practices limit the usefulness of these reports:

• Actual expenditure from the Government Reserve Fund is classified into an eleventh category alongside the ten COFOG categories, instead of being classified by purpose. This constitutes a non-trivial proportion of spending (2.5 percent of total spending in the past three years) that is not properly classified.

• With the exclusion of SNCOs, the composition of spending by economic type for central government is distorted. Wages and salaries—the SNCOs’ main expenditure items—are paid through subsidies, goods and services, and grants from the state budget, and classified as such. As a result, central government compensation of employees is largely underestimated in these reports (by 47 percent in 2016—see Table 1.3).

Table 1.3.

Armenia: Spending by Economic Type, including and excluding SNCOs

(AMD, billion in 2016)

Source: GFS data

Table 1.3.

Armenia: Spending by Economic Type, including and excluding SNCOs

(AMD, billion in 2016)

	State Budget	Central Government
Compensation of employees	142	270
Use of goods and services	162	136
Interest	98	98
Subsidies	121	61
Grants	158	133
Social benefits	405	407
Other expense	195	198
Net/gross investment in nonfinancial assets	167	167
Total	1,449	1,470

Source: GFS data

View Table

Table 1.3.

Armenia: Spending by Economic Type, including and excluding SNCOs

(AMD, billion in 2016)

	State Budget	Central Government
Compensation of employees	142	270
Use of goods and services	162	136
Interest	98	98
Subsidies	121	61
Grants	158	133
Social benefits	405	407
Other expense	195	198
Net/gross investment in nonfinancial assets	167	167
Total	1,449	1,470

Source: GFS data

1.3.2. Internal Consistency (Basic)

26. Armenia reliably publishes one of the three internal consistency checks of fiscal data required under the Code. It publishes reconciled above-the-line fiscal deficits with below-the-line financing. While data on holdings of state debt by creditors are reported in the monthly debt bulletin and annual debt report, a reconciliation of the debt issued with the stock of debt by counterparties based on a security-by-security database or on sectoral national accounts, as envisaged by the FTC, is not available. Importantly, there is no reconciliation available between the net financing and the change in the stock of debt.

27. The absence of a balance sheet makes it difficult to present a comprehensive stock-flow reconciliation. Data on other economic flows, such as revaluations (e.g. due to exchange rates or market prices) and other changes in the volume of assets and liabilities (e.g. changes in the sector coverage) are to be integrated with balance sheets. Figure 1.8. presents the reconciliation of the change in the state debt and the individual elements contributing to the increase in the debt – the deficit and net acquisition of financial assets. While the major part of the debt increase in 2016 can be explained by the financing of the deficit and net acquisition of financial assets, in particular loans, in 2014 and 2015 0.7 and 0.2 percent of GDP, respectively, remain unexplained. The unexplained residuals could result from currency deterioration but may also include other unidentified factors or possible errors.

View Full Size

Stock-flow Adjustment, 2016

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: GFS and debt data.

• Download Figure
• Download figure as PowerPoint slide

1.3.3. Historical Revisions (Not met)

28. Fiscal statistics are not systematically revised to reflect new information. Data based on the GFSM 2001/2014 submitted for the IMF’s Government Finance Statistics Yearbook have never been revised. A major one-off methodological revision was done by the NSS in 2015 to the national accounts when moving from SNA 1993 to SNA 2008 and updated numbers were published in the National Accounts of Armenia Statistical handbook. However, revisions to fiscal statistics to reflect updated source data, and corrections of errors and omissions are not regularly published and do not indicate whether the numbers are provisional, revised, updated, or final. The implementation of some recommendations of this report would result in meaningful revisions to GFS and SNA data, which should be disclosed. Box 1 gives examples for required historical revisions.

1.4.1. Statistical Integrity (Good)

29. Fiscal statistics are compiled by a specific government agency and disseminated in accordance with international standards. Official fiscal statistics are compiled and disseminated by the Statistical Committee on the SNA 2008 basis. The Statistical Committee is a state body under the Government of the Republic of Armenia. The President of the Statistical Committee is appointed by the government of the Republic of Armenia upon nomination by the Prime Minister of the Republic of Armenia for a term of six years. The Law stipulates that statistics compiled by the NSS should follow best statistical practice and meet the principle of objectivity, reliability, accuracy, confidentially with respect to individual data, comparability with international norms and standards, and accessibility of statistical information to public. The website of the NSS provides user-friendly access to the statistical information. The MoF compiles statistical reports based on GFSM 2014 which are not published or reconciled with the national accounts. Capacity constraints, at both the NSS and MoF, may constrain full implementation of international statistical standards.

1.4.2. External Audit (Basic)

30. The Chamber of Control (CoC) has published an opinion on the annual budget execution report of the state budget. The opinion is a review of the compliance of the report with financial legislation and includes an assessment of the reliability of some of the financial data. However, it is not a financial audit per se, as it does not include: (i) a complete assessment of the report’s compliance with the relevant cash accounting standards; and (ii) a statement that the report presents a true and fair view of central government’s financial position.

31. Pursuant to a new Audit Law, the Chamber of Audit (CoA) was established in April 2018 to supersede the CoC with a view to strengthening the integrity of Armenia’s public finances. The new Law follows more closely the legal framework for supreme audit institutions recommended by the International Organization of State Audit Institutions (INTOSAI). The CoA will be responsible for conducting financial audits of central government financial statements according to international standards of auditing (ISAs). The independence of the CoA is enshrined in the Constitution as well as in the respective laws. Its independence has been strengthened in the new law, with its annual workplan being determined at its own discretion.

32. The CoA plans to start conducting financial audits according to ISAs for the central government budget execution report in 2018, gradually increasing the coverage. For this purpose, the CoA should review whether the cash accounting standards are sufficiently well defined to enable financial audits according to ISAs, and to the extent they are not, advise the MoF on improvements. It is also piloting financial audits of the separate financial statements of budget institutions in terms of the new IPSAS-based Armenian Public Sector Accounting Standards (APSAS).

1.4.3 Comparability of Fiscal Data (Basic)

33. Budget execution reports are prepared on the same economic basis as the budget and forecasts. The high-level as well as detailed levels of fiscal aggregates are based on the same classification and methodology. Transactions of extra-budgetary accounts are not included in the approved budget, but they are incorporated ex-post in the budget revisions to ensure comparable data with the actuals.

34. The differences between statistical reports based on the GFSM 2014 and SNA 2008 are significant, and no reconciliation is available. In principle, both methodologies should provide consistent data. However, the difference between the general government deficit (net lending/borrowing) figures are on average 0.6 percent of GDP (Figure 1.9). Much of these differences are likely to be explained by differences in the sector coverage, which in the national accounts is based on the type of activity (distinguishing commercial and non-commercial activities), rather than on the institutional units. Further elements impacting the differences are different source data used to compile statistics, and possible accrual elements in the SNA based accounts.

View Full Size

Net Lending/Borrowing for the General Government as reported in GFSM2001 and SNA 2008

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: GFSM 2014 and SNA 2008 data.

• Download Figure
• Download figure as PowerPoint slide

Recommendations

Issue: The current fiscal reports provide an incomplete picture on the fiscal performance of general government as not all institutional units are covered in the budget execution and statistical reports. Including all non-market bodies in the reports will provide a more realistic picture on the government performance.

Recommendation 1.1: Expand the institutional coverage of the budget execution and statistical reports.

• Ensure proper sectorization of public entities into the general government sector or public sector by applying the “market/non-market test” (as defined in the GFSM 2014 and SNA 2008);

• Prepare and maintain a comprehensive and up-to-date list of general government units and public corporations;

• Expand consolidated general government statistical reports by including municipal NCOs, and non-market public corporations according to the GFSM 2014 and SNA 2008 methodology; and

• Include state SNCOs in the consolidated central government budget execution reports.

Issue: Fiscal reports do not provide a comprehensive view of government net worth due to the absence of integrated balance sheet data. Fiscal policy can only be done thoroughly with knowledge of government overall wealth, the value and composition of its assets including their quality, and all explicit and implicit obligations.

Recommendation 1.2: Publish balance sheet for central government and expand coverage to general government.

• Collect and consolidate fragmented data on the general government assets and liabilities from the SPMD, ministerial balance sheet; Treasury, Public Debt Management, Department of Shares, Department of State Obligations, SNCOs, and other government bodies as relevant;

• Publish a consolidated central government balance sheet (including SNCOs) based on the GFSM 2014 methodology, initially with available information and gradually expand coverage of stocks and institutions to general government; and

• Present a state financial balance sheet covering all financial assets and liabilities and publish in a relevant report or bulletin on a monthly and annual basis.

Issue: The credibility of data on fiscal performance is undermined by unexplained differences in fiscal reports. To provide policy makers and the public with reliable data on public finances, performing regular quality checks via reconciliation of different indicators and datasets, and stocks and flows, is inevitable.

Recommendation 1.3: Enhance consistency, comparability, and integrity of fiscal reporting.

• Compile and publish the following tables which reconcile (i) the change in stocks and related flows by individual categories of financial assets and liabilities; (ii) the deficit/surplus and the change in in debt; and (iii) the differences between the main aggregates based on the GFSM 2014 and SNA 2008 methodology (including an explanatory note).

• Ensure that material government operations, as for example capital injections to loss-making corporations, are properly classified as above the line or, when relevant as below the line transactions, to reflect the economic substance rather than the nationally based legal form.

• Conduct systematic review of fiscal statistics with respect to compliance with international standards, incorporate new information and methodology (e.g. on sectorization) when relevant, and revise time series to ensure their consistency. Publish an accompanying note explaining revisions.

• Assign clear responsibilities on GFS in the Ministry of Finance and ensure adequate capacity.

Issue: Comprehensiveness and quality of fiscal reports will benefit from the new accounting system based on international standards.

Recommendation 1.4: Ensure that implementation of the new accounting system improves comprehensiveness, quality, and compatibility of fiscal reports.

• Enforce IPSAS-based national accounting standards and publish financial statements;

• Integrate cash-based budget execution reports and accrual-based accounting, and ensure reconcilability between the budget execution data, fiscal statistics, and financial statements;

• Provide comprehensive balance sheet data covering all general government assets and liabilities integrated with the flow data; and

• Ensure that the accounting information is designed to deliver information according to international statistical standards (GFSM 2014 and SNA 2008), including information on intra and inter public sector flows and stocks and counterpart sectors.

Table 1.4.

Armenia: Summary Evaluation: Fiscal Reporting

Table 1.4.

Armenia: Summary Evaluation: Fiscal Reporting

	Principle	Assessment	Importance	Recs
1.1.1	Coverage of Institutions	Basic: Fiscal reports consolidate most of the central government units and budgetary local government but exclude municipal non-commercial organizations and public corporations.	High: Public corporations with net expenditures of 5.6 % of GDP and non-equity liabilities of 40.9 % of GDP are outside fiscal statistics. A number of loss-making SOEs are likely to fall in the general government sector.	1.1
1.1.2	Coverage of Stocks	Basic: Fiscal reports cover state cash and deposits and debt data. No balance sheet is published.	High: Unreported public sector assets of 82.6 percent of GDP and liabilities of 34.1 percent of GDP.	1.2, 1.4
1.1.3	Coverage of Flows	Basic: Fiscal reports cover cash revenue and expenditure, no accrual information is published.	Medium: Misclassification of capital injections to loss-making SOEs and unreported non-monetary expenditure undermines reliability of the fiscal balances.	1.4
1.1.4	Coverage of Tax Expenditures	Good: Estimated revenue loss from tax expenditures is published annually by broad sectors and exemption categories, but no limits or objectives	Medium: Tax expenditures of 7 percent of GDP are relatively high, but the MoF has started to reduce them.	2.1
1.2.1	Frequency of In-Year Reporting	Advanced: In-year budget execution reports are published monthly, within a month.	Low: The government updates the budget at least monthly.
1.2.2	Timeliness of Annual Financial Statements	Advanced: Annual state budget execution reports are published within 4 months after the reporting period.	Low: Introduction of the new accounting system may delay availability of annual results.
1.3.1	Classification	Advanced: Information on revenue and expenditure by administrative, economic, functional, and program classification provided, in line with international standards.	Medium: 3 percent of spending not classified by purpose and exclusion of SNCOs distorts composition of spending by economic type	1.1
1.3.2	Internal Consistency	Basic: Fiscal reports provide only reconciled above-the-line fiscal deficit with below-the-line financing.	High: Unexplained residuals in stock-flow adjustments up to 0.7 percent of GDP.	1.3
1.3.3	Historical Revisions	Notmet: Fiscal statistics are not systematically revised to reflect new information.	Medium: Fiscal statistics do not reflect updated information that may potentially impact the fiscal results.
1.4.1	Statistical Integrity	Good: Fiscal statistics are compiled by the NSS and disseminated in accordance with SDDS.	Medium: Limited capacity may affect quality of statistics.
1.4.2	External Audit	Basic: The CoC has published an opinion on the annual budget execution report, but does not provide an overall assessment regarding true and fair view of the financial position.	Medium: CoA established in April 2018 with strengthened independence and updated methodology.
1.4.3	Comparability of Fiscal Data	Basic: Budget execution reports are prepared on the same economic basis as the budget, but no reconciliation between different reports.	High: Unexplained differences between the fiscal balance based on national accounts and GFSM 2014 vary up to 2.5 percent of GDP.	1.3.

View Table

Table 1.4.

Armenia: Summary Evaluation: Fiscal Reporting

	Principle	Assessment	Importance	Recs
1.1.1	Coverage of Institutions	Basic: Fiscal reports consolidate most of the central government units and budgetary local government but exclude municipal non-commercial organizations and public corporations.	High: Public corporations with net expenditures of 5.6 % of GDP and non-equity liabilities of 40.9 % of GDP are outside fiscal statistics. A number of loss-making SOEs are likely to fall in the general government sector.	1.1
1.1.2	Coverage of Stocks	Basic: Fiscal reports cover state cash and deposits and debt data. No balance sheet is published.	High: Unreported public sector assets of 82.6 percent of GDP and liabilities of 34.1 percent of GDP.	1.2, 1.4
1.1.3	Coverage of Flows	Basic: Fiscal reports cover cash revenue and expenditure, no accrual information is published.	Medium: Misclassification of capital injections to loss-making SOEs and unreported non-monetary expenditure undermines reliability of the fiscal balances.	1.4
1.1.4	Coverage of Tax Expenditures	Good: Estimated revenue loss from tax expenditures is published annually by broad sectors and exemption categories, but no limits or objectives	Medium: Tax expenditures of 7 percent of GDP are relatively high, but the MoF has started to reduce them.	2.1
1.2.1	Frequency of In-Year Reporting	Advanced: In-year budget execution reports are published monthly, within a month.	Low: The government updates the budget at least monthly.
1.2.2	Timeliness of Annual Financial Statements	Advanced: Annual state budget execution reports are published within 4 months after the reporting period.	Low: Introduction of the new accounting system may delay availability of annual results.
1.3.1	Classification	Advanced: Information on revenue and expenditure by administrative, economic, functional, and program classification provided, in line with international standards.	Medium: 3 percent of spending not classified by purpose and exclusion of SNCOs distorts composition of spending by economic type	1.1
1.3.2	Internal Consistency	Basic: Fiscal reports provide only reconciled above-the-line fiscal deficit with below-the-line financing.	High: Unexplained residuals in stock-flow adjustments up to 0.7 percent of GDP.	1.3
1.3.3	Historical Revisions	Notmet: Fiscal statistics are not systematically revised to reflect new information.	Medium: Fiscal statistics do not reflect updated information that may potentially impact the fiscal results.
1.4.1	Statistical Integrity	Good: Fiscal statistics are compiled by the NSS and disseminated in accordance with SDDS.	Medium: Limited capacity may affect quality of statistics.
1.4.2	External Audit	Basic: The CoC has published an opinion on the annual budget execution report, but does not provide an overall assessment regarding true and fair view of the financial position.	Medium: CoA established in April 2018 with strengthened independence and updated methodology.
1.4.3	Comparability of Fiscal Data	Basic: Budget execution reports are prepared on the same economic basis as the budget, but no reconciliation between different reports.	High: Unexplained differences between the fiscal balance based on national accounts and GFSM 2014 vary up to 2.5 percent of GDP.	1.3.

2.1.1. Budget Unity (Basic)

38. The annual budget documentation provides information on revenue, spending, and financing activities of budgetary central government and SNCOs. The budget documents cover projected revenue, and planned expenditure of state budget institutions, by administrative, functional, economic, and program classification. An annex to the budget presents projected gross revenue (i.e., budget transfers, and own source revenue), and planned spending of SNCOs, aggregated by line ministries.

39. Revenues from extra-budgetary accounts of budgetary entities are not included in the budget. Unlike SNCOs, which are off budget, extra-budgetary accounts are sub-accounts of budget institutions, such as the Police, and the State Revenue Committee, for earmarked non-tax revenues (e.g. fines), external grants, and sales income. Spending from extra-budgetary accounts are not part of the budget but authorized by government decree during the year. In 2018, extra -budgetary accounts amounted to AMD 33.8bn in 2016 – 2.8 percent of total revenue (Figure 2.1). These, together with the SNCOs own-source revenue, represented 4.7 percent of total state revenue (Figure 2.2), which is comparable to peers.

View Full Size

Composition of Revenue of EBAs

(percent of total EBA revenue, 2016)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Own-Source Revenues in Selected Countries

(percent of total CG revenue)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

• Download Figure
• Download figure as PowerPoint slide

40. The exclusion of extra-budgetary accounts will complicate the implementation of the current fiscal rules. As part of the revamp of the system of fiscal rules, the Budget System Law has been recently revised and it requires the government to commit to a cap on current expenditure when debt exceeds 50 percent of GDP (indicator 2.3.1). Ex ante expenditure does not include extra-budgetary accounts, while ex-post it does. An ex ante cap on expenditure may be violated ex post, as extra-budgetary revenue and spending materialize. Such violation may only be visible during budget execution. While currently extra-budgetary accounts remain small, this issue may gain more importance if reliance on extra-budgetary accounts were to increase to circumvent the cap on current expenditure.

2.1.2. Macroeconomic Forecasts (Advanced)

41. Budget documentation contains detailed macroeconomic forecasts that underlie budget planning. Three year-ahead forecasts of key macroeconomic variables (such as GDP, and inflation) are provided, their demand and supply determinants, as well as the assumptions on the external and domestic environment underpinning them. The recent performance of these variables, and their impact on fiscal aggregates are also comprehensively discussed. In a first round, forecasts are provided in the Medium-Term Expenditure Framework (MTEF) submitted to Parliament in June. These are updated and presented in the annual budget document.

42. Medium-term forecasts have been optimistic in recent years, but have not substantially differed from those made by external agencies. Between 2001–08, outturns for real GDP have exceeded forecast, while between 2011–2016, real GDP has underperformed (Figure 2.3). The same is true for inflation (Figure 2.4). Over the period 2011–16, real GDP has underperformed compared to the authorities’ forecasts, but this has been the case for forecasts made by other agencies. (Figure 2.5).

View Full Size

Real GDP Forecasts

(year on year, percentage change)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Ministry of Finance

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Inflation Forecasts

(year on year, percentage change)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Forecast Errors for Real GDP by Forecasters

(Actual minus projected, average 2011–2016)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Ministry of Finance; IMF data are from the April vintages of the WEO.

• Download Figure
• Download figure as PowerPoint slide

2.1.3. Medium-Term Budget Framework (Good)

43. The medium-term budget framework provides medium-term projections by ministries, functions, and programs, but is not fully transparent on the economic classification. Armenia introduced a medium-term expenditure framework (MTEF) in 2003. The annual MTEF document is published in July, three months before the draft budget is presented to Parliament in October. It includes:

• Forecasts for state budget revenues for the next three years, disaggregated by the first level of the economic classification;

• Outturns for state budget revenues, actual for the previous year and estimated for the current year, and in some parts for earlier years, but not disaggregated by the economic classification;

• Forecasts for state budget expenditures for the next three years, disaggregated by the functional, administrative, and program classification, as well as the capital/current distinction of the economic classification; and

• Outturns for state budget expenditures, actual for the previous year, and estimate for the current year, and in some parts for earlier years, disaggregated by the functional classification and in some parts by the capital/current distinction of the economic classification.

44. The deviations across MTEFs and with budget allocations highlight their loose relationship. The annual budget, which is approved less than six months after the MTEF, has differed markedly from the latter in recent years (Figure 2.6). The change in macro-fiscal forecasts during this period may explain part of this deviation. For instance, in 2017 the annual budget had to be tabled on a substantially lower deficit target than the MTEF as debt numbers gradually revealed that the debt brake would have to be applied. However, it is also the result of two weakly related processes, whereby ministries and agencies provide separate bids for the MTEF and the annual budget. Referring the MTEF to the Debt Sustainability Analysis (DSA) would improve consistency of fiscal planning.

View Full Size

Budget Ceiling vs. Ceiling in Same Year MTEF

(budget minus MTEF, in percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Budget documents

• Download Figure
• Download figure as PowerPoint slide

45. Fiscal outturns have deviated significantly from the forecasts in the MTEFs over the past decade, although less than for many comparable countries. Expenditure outturns have exceeded what was envisaged in the MTEF, particularly for the second outer year, when it represented around 1 percent of GDP (Figure 2.7a). This has been balanced with the conservative revenue projections (Figure 2.7b). As a result, the MTEF has provided more accurate estimates of the fiscal balance.

View Full Size

Cross-Country Comparison: Average Medium-Term Fiscal Forecast Error in Selected Countries, 2007–16

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: MoF, IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

2.1.4. Investment Projects (Basic)

46. The government’s total multi-annual obligations under investment projects are not disclosed in a comprehensive way, although substantial information is provided, especially in relation to external loans that finance them. All major projects are at present financed by loans from donors. Information on these projects is contained in several publications:

• The MTEF document includes forecasts of externally-financed capital expenditure projects for the following three years, by individual project and ministry, together with donor financing and government’s co-financing;

• The annual budget documentation further details this information;

• The annual and quarterly budget execution reports compare the budgeted amounts with actuals; and

• Two data sheets on the MoF’s website include information on the total value of projects as per loan agreements, actual loan disbursements and undisbursed amounts for individual projects, their deadlines, and their terms; one of the data sheets is on financial indicators of the government’s external loans and credits as of the year end, and the other is on the terms of the government’s external loans and credits as of the year end.

However, none of this information discloses the total obligations from the projects, and undisbursed amounts pursuant to loan agreements are not generally equal to commitments arising from construction contracts.

47. Cost-benefit analyses for major investment projects are not required prior to their approval according to a uniform methodology and are not always published. For donor-financed investment projects, cost-benefit analyses are usually undertaken according to the individual donors’ approved methodologies, but typically not published. There is no requirement for cost-benefit analyses for projects financed from general budget resources.

48. Major investment projects are generally required to be subject to open and competitive tender. Projects financed from general budget resources are subject to the procurement law, which requires competitive and open procurement, but projects financed by donors are exempt from the procurement law because of international agreements. The donors who are financing major projects in Armenia generally require open and competitive tendering.

49. General government investment in Armenia is not particularly high in comparison with other countries, but increased pressure on public investment could arise. General government public investment amounts to 3.3 percent of GDP (Figure 2.8). The authorities are planning to almost double nominal general government capital expenditure over the next four years. Given Armenia’s relatively high state debt of 53.6 percent of GDP in 2017, there is a need to ensure efficiency in public investment as fiscal space is limited.

View Full Size

Public Investment in Selected Countries

(General Government, percent of GDP in 2016)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Armenian authorities for Armenia, OECD for other countries.

• Download Figure
• Download figure as PowerPoint slide

50. Efforts are underway to further strengthen public investment management practices in Armenia. A new PPP Law is being drafted, and the government is expected to develop a reform agenda for public investment management in the near future. A Public Investment Management Assessment (PIMA) will be undertaken by the IMF later this year, which will provide recommendations to support this agenda.

2.2.1. Fiscal Legislation (Advanced)

51. The Budget System law (BSL) regulates the contents of the budget proposal, the budget preparation calendar, and a constitutional law regulates the legislature’s budget amendment rights.⁸ Article 15 of the BSL contains detailed provisions for the contents of the MTEF document, and article 16 provides similar requirements for the annual budget document. Article 21 gives detailed directions for how the MTEF and budget preparation processes shall be structured. This includes the deadlines for MTEF submission to Parliament (by July 20 each year) and for budget submission (90 days before the budget year). The BSL also regulates the preparation, approval and execution of municipal budgets (articles 27 – 36). The law on parliamentary procedures stipulates that Parliament can propose changes to the draft budget, and that the government must provide a revised budget proposal containing the changes that they accept. Parliament must accept or reject the revised budget in its totality, it is not possible to vote on individual items.

52. The legal and regulatory framework provides a transparent framework for budget preparation. The BSL was adopted in 1997 and has been amended regularly, most recently in 2017, to reflect innovations and improvements in the budget process. The BSL is supported by a series of annual decrees and regulations, which provide further guidance on the budget process. Each year, the Government issues a decree providing a detailed program and timetable for the MTEF and budget preparation process.⁹ Following budget approval, another decree is issued to regulate the budget execution process.¹⁰ This decree provides a quarterly breakdown of budget allocations and gives further directions for budgetary amendments, including for transactions through the extrabudgetary accounts of ministries and agencies.

2.2.2. Timeliness of Budget Documents (Good)

53. The budget proposal is submitted to Parliament 90 days before the budget year and approved by early December. The deadline for budget submission is prescribed in the BSL and is always complied with, according to the Ministry of Finance. Parliament has ample time for budget deliberations and budget approval is generally well before the budget year. Table 2.3 demonstrates that over the last five years the annual budgets have been submitted in time and approved before the end of the financial year.

Table 2.3.

Armenia: Dates for Draft Budget Submission and Approval 2014 –18

Source: Ministry of Finance

Table 2.3.

Armenia: Dates for Draft Budget Submission and Approval 2014 –18

Budget year	Date of budget submission	Date of budget approval
Draft State budget 2014	October 2, 2013	December 12, 2013
Draft State budget 2015	October 2, 2014	December 4, 2014
Draft State budget 2016	October 2, 2015	December 9, 2015
Draft State budget 2017	September 30, 2016	December 8, 2016
Draft State budget 2018	October 2, 2017	December 8, 2017

Source: Ministry of Finance

View Table

Table 2.3.

Armenia: Dates for Draft Budget Submission and Approval 2014 –18

Budget year	Date of budget submission	Date of budget approval
Draft State budget 2014	October 2, 2013	December 12, 2013
Draft State budget 2015	October 2, 2014	December 4, 2014
Draft State budget 2016	October 2, 2015	December 9, 2015
Draft State budget 2017	September 30, 2016	December 8, 2016
Draft State budget 2018	October 2, 2017	December 8, 2017

Source: Ministry of Finance

2.3.1. Fiscal Policy Objectives (Good)

54. Fiscal policy is currently guided by an upgraded set of fiscal rules, as set in the Budget System Law:

• The state deficit proposed for the budget year (t+1) cannot exceed 7.5 percent of GDP envisaged in the coming year budget (t+1) ;

• If state debt exceeds 40 percent of GDP, capital expenditures (t+1) should not be below the level of planned state budget deficit;

• If state debt is between 50 and 60 percent of GDP, the above still applies, and, in addition, the growth rate of primary current expenditure is capped by the historical (average of the last 7 years) growth rate of GDP;

• If state debt exceeds 60 percent of GDP, the government must, in addition to the above (but the growth rate of primary current expenditure will be capped at 0.5 percent lower of the average growth rate of GDP), total current expenditure cannot exceed the total amount of tax revenues for the upcoming year;

• When state debt exceeds 50 percent of GDP, the government is also required to present a corrective action plan in the MTEF to bring debt below 50 percent of GDP within 5 years; and

• When state debt exceeds 60 percent of GDP, the government must submit a program of measures to the National Assembly to bring debt below 60 percent within 5 years.

An escape clause in the Budget System Law allows the government to suspend the set of measures prescribed above in exceptional cases such as disasters, war, and negative economic shocks. This does not apply to the deficit limit of 7.5 percent of GDP.

55. These rules have evolved over time. The timeline in Figure 2.9 summarizes the changes. A debt rule was first introduced in the Law on Public Debt (LPD) in 2008 limiting public debt to 50 percent of GDP – no further debt could be issued once this limit is reached. At the same time, a deficit limit of 5 percent of GDP was introduced in the BSL. In 2009, these laws were amended to increase the deficit limit to 7.5 percent of GDP, and the debt ceiling to 60 percent of GDP, and introduce a debt brake of 50 percent, that requires the deficit not to exceed 3 percent of GDP. In 2015, the definition of debt in the debt rules was changed from state debt (government and the central bank debt) to government debt. In 2018, the debt and deficit rules were changed as described in the previous paragraph, and an escape clause was added.

View Full Size

The Evolution of Fiscal Rules

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Relevant legislation.

• Download Figure
• Download figure as PowerPoint slide

56. The deficit rule has not been binding in recent years, but the debt brake was applied in 2017 (Figures 2.10 and 2.11). Since the deficit ceiling was increased from 5 to 7.5 percent of GDP in 2009 in response to the Global Financial Crisis, it has since not been binding. The lower ceilings set within consecutive IMF programs has instead provided guidance to fiscal policy. Following the economic shock in 2014, debt rose above 50 percent of GDP, and the government was required to comply with the 3 percent deficit correction mechanism in preparing the 2017 budget. Debt is still above 50 percent of GDP, but the new fiscal rules framework adopted at the end of 2017 no longer requires the 3 percent deficit correction. The government is however expected to present a corrective action plan in the 2019–2021 MTEF, and set a cap on current spending by decree.

View Full Size

Deficits and the Deficit Rules

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: World Economic Outlook database.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Debt and the Debt Rules

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

• Download Figure
• Download figure as PowerPoint slide

57. The recent amendments aim at making the rules more flexible, while maintaining their credibility. The debt rules now contain an escape clause to allow the government to suspend their application in exceptional circumstances. In an effort to maintain the credibility of the rules-based framework, and following IMF advice, the authorities decided to keep the numerical limits unchanged, and complement them with an additional constraint at 40 percent. However, the 50 percent debt brake no longer requires an automatic adjustment and the 60 percent debt ceiling does not make new debt null and void.

58. Increased flexibility requires enhanced transparency. To be credible, the corrective action plan in the MTEF needs to be tightly linked with the budget, and consistent over time (see indicator 2.1.3). Systematically reconciling the MTEF with the budget, and the successive MTEFs would enhance credibility (see indicator 2.4.3). A comparison of government’s forecasts with other forecasters’ will also show a commitment to transparency, and can pave the way for better and more credible forecasts (see indicator 2.4.1).

59. Implementation of the rules will require strong external scrutiny. ¹¹ Ex-ante scrutiny on fiscal projections and compliance to fiscal rules can be provided by an independent fiscal institution, such as the Parliamentary Budget Office, once it is granted with relevant functions and builds sufficient capacity. Ex-post verification of compliance is equally important. This can be provided by the Chamber of Audit, which has, as per the 2015 Constitution, the mandate to perform external audit on the use of state funds covered in the Annual Budget Law. Inclusion of the government fiscal objectives, including the corrective action plan, in the Annual Budget Law will give it the required mandate to undertake ex-post compliance.

2.3.2. Performance Information (Advanced)

60. The state budget documentation includes ex ante and ex post performance information for all programs and their sub-programs. A program classification has been in use since 2008 for budget presentation and reporting. An annex of the budget message has detailed information on objectives, outputs, and outcomes for all programs and sub-programs. All programs have outcome descriptions and about half have some outcome indicators. Quantitative indicators are used comprehensively for outputs. The annual budget execution report includes a detailed discussion on performance against the targets and give reasons for the deviations. Box 2.1 shows an example of a current program of the Ministry of Labor and Social Affairs, which has 16 sub-programs, and the description in the 2018 budget of one of its subprograms.

61. Reforms to improve the quality and presentation of performance information will be implemented for the 2019 state budget.¹² These will be associated with the introduction of appropriation by programs, replacing the current appropriations of a combination of administrative unit, function, and economic item. The 2019 budget documentation will also include quantitative outcome and output indicators in the annex to the budget message. The introduction of appropriation by programs will give additional stimulus to improve the performance budgeting system in the near future, including potentially a more homogenous presentation of programs and subprograms with improved, aligned performance indicators, more detailed information by adding a third level to the current program classification, and a more transparent link of outcome information to appropriations.

2.3.3. Public participation (Basic)

62. Armenia has recently published a short document entitled the budget guide. It includes a description of the budget system and the budget process, as well as a summary of the main budget aggregates and underlying economic indicators. The guide has the merit of being comprehensive, but its readability can be improved. The budget guide is detailed, with useful information on major government initiatives, but does not present detailed information on its practical implication for the lives of typical citizens.

63. The public’s opportunity to participate in budget deliberations is limited, but some new initiatives are taking place at the level of municipalities. The public is able to comment on the draft budget after it is submitted to Parliament through a website.¹³ However, neither citizens, nor NGOs, and business groups can formally attend parliamentary committee’s hearings, or submit written comments to the various committees on the draft budget. At the local government levels, some initiatives within large municipalities have taken place, such as live broadcasting of local budget discussions, and the allocation of one percent of the local budget directly by citizens, but these remain small in relation to overall spending. International evidence suggests that participatory budgeting can enhance the quality of spending and improve tax compliance,¹⁴ which is an issue of particular importance for Armenia.

2.4.1. Independent Evaluation (Not Met)

64. The government does not publish a comparison of its forecasts with those of independent forecasters and they are not subject to independent evaluation. Neither the MTEF nor the annual budget documentation compare the government’s economic and fiscal forecasts with those of independent institutions. An official statement of the Central Bank of Armenia to Parliament validates the government’s proposed budget, but it does not contain an evaluation of its forecasts.

65. The Budget Office of the Parliament does not evaluate the government’s forecasts and policies, including compliance with the fiscal rule. It has been granted functional independence in 2016, with a mandate to support the Parliament on budgeting issues. It is currently lightly staffed (three technical staff in total), and its main role is to help Parliamentarians understand budget propositions and their implications. Going forward, for the Budget Office to play a role as an independent fiscal watchdog, it will need a stronger mandate and will require capacity building.¹⁵ The example of Sweden illustrates how an institution with a limited budget can succeed (Box 2.2).

Box 2.2.

The Swedish Fiscal Policy Council – Effective Despite a low Budget

The role of the Swedish Fiscal Policy Council (SFPC) is to focus on providing independent evaluation of the Swedish government’s fiscal policy. This role complements that of three other independent institutions responsible for producing forecasts, and scrutinizing budget execution. The council’s resources are limited (approximately US$ 1 million in 2011) and is staffed by a handful of part-time staff and consultants.

One of the main tasks of the SFPC is to review the Government’s plan against its fiscal rule, designed in Sweden as a surplus objective, and an expenditure ceiling. The SFPC has offered both favorable and unfavorable assessments in that regard. While in 2012, it endorsed the government’s policy, stating that “fiscal policy pursued in 2011 and 2012 has complied with the current fiscal framework…and there is little risk of expenditures exceeding the expenditure ceiling in the next few years”, in 2013, it declared that “the indicators reported by the government suggest that the surplus target will not be met” and that “the government should present a clear plan for meeting the surplus target during the relevant business cycle.”

The government’s forecast error of the real sector has declined (Table 2.4). While this could be due to a host of other factors, it is in line with the experience of other countries, where the creation of an independent fiscal council has helped reduce some of the bias in projections.

Sweden: Forecast Errors for Real GDP

(in percent)

Sweden: Forecast Errors for Real GDP

(in percent)

	2000–2005	2006–2010	2011–2015
Root mean square error	0.58	1.87	0.47

View Table

Sweden: Forecast Errors for Real GDP

(in percent)

	2000–2005	2006–2010	2011–2015
Root mean square error	0.58	1.87	0.47

Source: Case Studies of Fiscal Councils – Functions and Impact, Curristine et al, IMF Policy Paper 2013; Forecast errors were computed from data in the Sweden Convergence Plans 2000–2015 and IMF WEO.

Table 2.4.

Armenia: Summary Evaluation: Fiscal Forecasting and Budgeting

Table 2.4.

Armenia: Summary Evaluation: Fiscal Forecasting and Budgeting

	Principle	Assessment	Importance	Rec
2.1.1	Budget Unity	Basic: Own revenues (from non-tax and external grants) of extra budgetary accounts and spending are not presented in approved budget.	High: Extra budgetary accounts are 2.8 percent of revenues, and contribute to high in-year revisions.	2.3
2.1.2	Macroeconomic Forecasts	Advanced: Budget documentation includes forecasts of main macro variables, their components and underlying assumptions.	Medium: Forecast have tended to be optimistic in recent years but bias is comparable with other forecasters.
2.1.3	Medium-term Budget Framework	Good: The MTEF includes outturns for two previous years and revenue and spending projections for next three years by ministry, function and program, but only by high level economic category.	High: There have been large variations between MTEF forecasts and outturns, with an optimistic bias.	2.1
2.1.4	Investment Projects	Basic: Major projects are required to be contracted via open and competitive tender. Government does not disclose total multi-year obligations. Cost benefit analyses not systematically undertaken or published.	High: Plan is to double capital expenditure in next 4 years, under tight fiscal space. A forthcoming PIMA will provide specific recommendations to improve management of public investment.
2.2.1	Fiscal Legislation	Advanced: The Budget System law regulates budget calendar and budget contents. Legislature’s amendment rights are defined in a Constitutional law.	Low: The legal framework provides a well-defined framework for budget preparation.
2.2.2	Timeliness of Budget Documents	Good: The budget proposal is submitted to Parliament 90 days before the budget year and approved by early December.	Low: Budget submission and approval deadlines have been met in the last 5 years.
2.3.1	Fiscal Policy Objectives	Good: Fiscal rules are precise and time-bound, but have been revised several times (change in 2008: to deficit rule; 2015: of debt definition; 2017: new rules)	High: 50 percent debt brake is being applied since 2017 and a corrective action plan is required.	2.1–2.3
2.3.2	Performance	Advanced: Budget documentation reports targets for, and performance against, the outcomes to be achieved in each major government policy area.	Medium: There is scope to improve performance information and its links to appropriations to improve decision making.
2.3.3	Public Participation	Basic: Budget guide with summary of budget aggregates. Participatory initiatives taking place at the level of municipalities.	Low: Budget documents are very comprehensive and include also non-financial program indicators.
2.4.1	Independent Evaluation	Notmet: No comparison with independent forecasters is provided. Budget Office does not evaluate government’s forecasts.	High: Fiscal rule revision brings more flexibility, but requires stronger monitoring.	2.2
2.4.2	Supplementary Budget	Notmet: Legislation gives government wide mandate to change budgetary allocations. As such, no supplementary budget required.	High: In-year budget revisions without Parliamentary approval of 8.6 percent of spending.	2.3
2.4.3	Forecast Reconciliation	Notmet: Differences between fiscal forecasts from one year to the next is not shown and discussed.	Medium: Absolute differences between budget and same year MTEF, and between outturns (excluding EBAs) and budget averaged 0.6 and 0.96 percent of GDP respectively in last five years.	2.1

View Table

Table 2.4.

Armenia: Summary Evaluation: Fiscal Forecasting and Budgeting

	Principle	Assessment	Importance	Rec
2.1.1	Budget Unity	Basic: Own revenues (from non-tax and external grants) of extra budgetary accounts and spending are not presented in approved budget.	High: Extra budgetary accounts are 2.8 percent of revenues, and contribute to high in-year revisions.	2.3
2.1.2	Macroeconomic Forecasts	Advanced: Budget documentation includes forecasts of main macro variables, their components and underlying assumptions.	Medium: Forecast have tended to be optimistic in recent years but bias is comparable with other forecasters.
2.1.3	Medium-term Budget Framework	Good: The MTEF includes outturns for two previous years and revenue and spending projections for next three years by ministry, function and program, but only by high level economic category.	High: There have been large variations between MTEF forecasts and outturns, with an optimistic bias.	2.1
2.1.4	Investment Projects	Basic: Major projects are required to be contracted via open and competitive tender. Government does not disclose total multi-year obligations. Cost benefit analyses not systematically undertaken or published.	High: Plan is to double capital expenditure in next 4 years, under tight fiscal space. A forthcoming PIMA will provide specific recommendations to improve management of public investment.
2.2.1	Fiscal Legislation	Advanced: The Budget System law regulates budget calendar and budget contents. Legislature’s amendment rights are defined in a Constitutional law.	Low: The legal framework provides a well-defined framework for budget preparation.
2.2.2	Timeliness of Budget Documents	Good: The budget proposal is submitted to Parliament 90 days before the budget year and approved by early December.	Low: Budget submission and approval deadlines have been met in the last 5 years.
2.3.1	Fiscal Policy Objectives	Good: Fiscal rules are precise and time-bound, but have been revised several times (change in 2008: to deficit rule; 2015: of debt definition; 2017: new rules)	High: 50 percent debt brake is being applied since 2017 and a corrective action plan is required.	2.1–2.3
2.3.2	Performance	Advanced: Budget documentation reports targets for, and performance against, the outcomes to be achieved in each major government policy area.	Medium: There is scope to improve performance information and its links to appropriations to improve decision making.
2.3.3	Public Participation	Basic: Budget guide with summary of budget aggregates. Participatory initiatives taking place at the level of municipalities.	Low: Budget documents are very comprehensive and include also non-financial program indicators.
2.4.1	Independent Evaluation	Notmet: No comparison with independent forecasters is provided. Budget Office does not evaluate government’s forecasts.	High: Fiscal rule revision brings more flexibility, but requires stronger monitoring.	2.2
2.4.2	Supplementary Budget	Notmet: Legislation gives government wide mandate to change budgetary allocations. As such, no supplementary budget required.	High: In-year budget revisions without Parliamentary approval of 8.6 percent of spending.	2.3
2.4.3	Forecast Reconciliation	Notmet: Differences between fiscal forecasts from one year to the next is not shown and discussed.	Medium: Absolute differences between budget and same year MTEF, and between outturns (excluding EBAs) and budget averaged 0.6 and 0.96 percent of GDP respectively in last five years.	2.1

66. In the absence of a fiscal council, a comparison with existing forecasts could be included in budget documentation. Several institutions produce regular macroeconomic forecasts for Armenia, including the Central Bank, international organizations, such as the IMF, and rating agencies, such as Fitch. An example is shown in Figure 2.12, which compares the annual forecast for real GDP for the budget year contained in the MTEF, with Fitch’s forecasts produced at about the same time, and the quarterly range forecasts of the Central Bank. Publishing such comparison can help reduce forecasting bias, by encouraging internal discussion during their preparation, and encouraging the adoption of more robust forecasting techniques.

View Full Size

Forecast Comparison for Real GDP

(2013–2018)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: Ministry of Finance.

• Download Figure
• Download figure as PowerPoint slide

2.4.2. Supplementary Budget (Not Met)

67. The government is granted high flexibility to change budget appropriations. The Budget System Law and the Annual Budget Law allow the government to:

• increase spending to the extent that state budget receipts exceed planned (Article 11 -Annual Budget Law);

• proportionally reduce (sequester) spending when there is a risk that there will be a shortfall in budgeted receipts of up to 10 percent (Article 23 – Budget System Law); and

• make virements across programs of up to three percent of the approved budget (Article 23 -Budget System Law).

68. As a result, there has been no need to request a supplementary budget in the past. Even in 2009, when revenue fell significantly as a result of the Global Financial Crisis, the government could still sequester spending under Article 23. In recent years, this flexibility has allowed the government to revise the budget upwards in year, typically to accommodate extra capital spending, which have averaged 3.3 percent of the approved budget over the period 2012–16 (Figure 2.13).¹⁶

View Full Size

Revisions to Budget (excluding EBAs)

(2012–2016, percentage change of approved budget)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: MoF

• Download Figure
• Download figure as PowerPoint slide

69. In Armenia, additional spending is regularized through regular, often weekly decrees and does not require Parliamentary approval. In addition to changes in budget appropriations, two other types of spending are approved in-year: (i) spending from extra budgetary accounts; and (ii) spending from the Government Reserve Fund (respectively 2.8 and 2.5 percent of the approved budget in 2014–16). In total, an average of 8.6 percent of total spending was undertaken without direct ex ante Parliamentary approval during 2014–16 (Figure 2.14), although it had ex ante parliamentary authority and was validated ex post in Parliament’s approval of the annual budget execution report. In many countries, particularly in the OECD, material changes to the budget are made through supplementary budgets, with Parliamentary approval. This occurs on average one to three times a year (Figure 2.15). These supplementary budgets are made necessary when the economic situation requires a change in the fiscal stance, or when new policy initiatives are proposed.

View Full Size

Spending Authorized Outside of the Budget Law

(in percent of approved budget, 2014–16)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: MoF

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Average Number of Supplementary Budgets in Selected Countries

(between 2009–11)

Citation: IMF Staff Country Reports 2019, 134; 10.5089/9781498315098.002.A001

Source: OECD

• Download Figure
• Download figure as PowerPoint slide

2.4.3. Forecast Reconciliation (Not Met)

70. The budget documentation does not provide a comparison of current fiscal forecasts with previous forecasts. Fiscal forecasts have changed significantly from the budget-year ceiling in MTEF to the annual budget—which is approved six months later—averaging 0.61 percent of GDP in absolute terms between 2012–2017 and have deviated from outturns (net of extrabudgetary accounts) by an average 0.96 percent of GDP over the same period. Neither the MTEF nor the annual budget document forecasts refer to previous macroeconomic and fiscal forecasts, and explain how and why these differences arise.

71. The new fiscal rules require stronger transparency on fiscal forecasts. The corrective mechanism embedded in the debt rules gives flexibility to the government to propose a credible medium-term fiscal path to bring debt down when it exceeds 50 percent of GDP (see section 2.3.1). When it subsequently updates the path, it will need to explain the underlying reasons for doing so. These could be related to changes in economic factors, new policies required since the last update, and other one-off factors that may justify the adjustment. This could be usefully represented in the form of a reconciliation table. Box 2.3 provides an illustration of a reconciliation table that can be included in the next MTEF.¹⁷

Box 2.3.

Forecast Reconciliation

Table 2.5 shows how a reconciliation table that can be presented in the forthcoming MTEF for 2019–2021:

• I. The starting point is the MTEF 2017–2019: the outturn for 2017 can be reconciled with (i) the plan in the MTEF 2017–2019; (ii) the proposed budget for 2017; and (iii) the revised budget for 2017, and the ceilings for 2018 and 2019 can be reconciled with the ceilings for the same year in the MTEF 2018–2020. The differences can be decomposed into policy, economic factors, one-offs, and accounting changes. There will be a difference between actual and plan, but attempts should to be made to minimize this residual.

• II. The same process can be applied to reconcile the MTEF 2018–2020 with (i) the approved and revised budget for 2018; (ii) the proposed ceilings for 2019 and 2020 in the MTEF 2019–2021

• III. The table can be extended to reconcile the MTEF 2019–2021 with the proposed budget for 2019.

A more granular (bottom-up) decomposition of outturn vs. approved budget can also be presented (see table 2.6).

Indicative Reconciliation Table for Aggregate Expenditure across MTEFs and Annual Budgets

(AMD billion)

Indicative Reconciliation Table for Aggregate Expenditure across MTEFs and Annual Budgets

(AMD billion)

View Table

Indicative Reconciliation Table for Aggregate Expenditure across MTEFs and Annual Budgets

(AMD billion)

View Expanded

Breakdown of Forecast Errors for 2018

(billion AMD)

Source: IMF

Breakdown of Forecast Errors for 2018

(billion AMD)

		2017 forecast	Outturn	Difference	of which:
					Policy changes	Economic factors	Residual
Taxes and duties		x	x	x	x	x	x
	value added tax	x	x	x	x	x	x
	profit tax	x	x	x	x	x	x
	income tax	x	x	x	x	x	x
	excise tax	x	x	x	x	x	x
	customs duty	x	x	x	x	x	x
	income tax	x	x	x	x	x	x
	other taxes and duties	x	x	x	x	x	x
Other income		x	x	x
Official transfers		x	x	x
Expense		x	x	x	x	x	x
	wages	x	x	x	x	x	x
	goods and services	x	x	x	x	x	x
	interest payment	x	x	x	x	x	x
	subsidies	x	x	x	x	x	x
	transfers	x	x	x	x	x	x
	pensions and allowances	x	x	x	x	x	x
	other	x	x	x	x	x	x
Acquisition of non-
financial assets		x	x	x

Source: IMF

View Table

Breakdown of Forecast Errors for 2018

(billion AMD)

		2017 forecast	Outturn	Difference	of which:
					Policy changes	Economic factors	Residual
Taxes and duties		x	x	x	x	x	x
	value added tax	x	x	x	x	x	x
	profit tax	x	x	x	x	x	x
	income tax	x	x	x	x	x	x
	excise tax	x	x	x	x	x	x
	customs duty	x	x	x	x	x	x
	income tax	x	x	x	x	x	x
	other taxes and duties	x	x	x	x	x	x
Other income		x	x	x
Official transfers		x	x	x
Expense		x	x	x	x	x	x
	wages	x	x	x	x	x	x
	goods and services	x	x	x	x	x	x
	interest payment	x	x	x	x	x	x
	subsidies	x	x	x	x	x	x
	transfers	x	x	x	x	x	x
	pensions and allowances	x	x	x	x	x	x
	other	x	x	x	x	x	x
Acquisition of non-
financial assets		x	x	x

Source: IMF