1.1.1. Coverage of Institutions

Advanced

11. In 2015, the public sector in Colombia comprised 3822 institutional entities that together accounted for public expenditure equivalent to about 50.8 percent of GDP (Table 3). CIEFP maintains a comprehensive list of public entities that is updated on a quarterly basis and published on the CGN’s website. The list, called CUIN (Clasificación entidades según código Único Institucional), categorizes entities according to Government Finance Statistics Manual (GFSM) 2014 guidelines. The CGN publishes another list that groups the same entities following the national standards that are applied to the presentation of the budget and consolidated financial statements. Following CUIN, in 2015, the Colombian public sector included:⁸

• Budgetary central government (BCG): comprising 102 executive, legislative and judicial bodies that include 23 ministries and departments, 7 agencies, 7 funds and 65 other public establishments (institutes, commissions, autonomous units, among others). The latter includes the UGPP, a pension fund covering public sector employees hired before Law 100 (1993) was enacted (see Box 1).

• Extrabudgetary central government: comprising 140 entities, of which 6 are professional councils; 37 corporations; 38 national funds (including the coffee fund, the national housing fund, and the fund for economic emergencies – FOGACOOP); 25 institutes; and 17 universities.⁹

• State and local governments: comprising 1,644 entities, of which 143 state-level budgetary and extrabudgetary entities (including 32 department administrations, 50 institutes, 16 universities and 45 other units) and 1501 local-level entities (including 1,103 municipalities, 36 municipal associations, 31 funds, 201 institutes and 130 other entities).

• Social security sector: comprising 29 pension funds (including the general PAYGO system administered by Colpensiones and other pension funds for the military, police, the legislative branch and teachers); the health insurance system; and 9 health centers.

• Public corporations:¹⁰ comprising 74 financial and 1822 non-financial corporations, including Ecopetrol - the Colombian oil company - and several municipal and departmental utility companies (electricity and water/sanitation) (See Section 3.3.2).¹¹

Table 3.

Public Sector Institutions and Finances, 2015

(in percent of GDP, unless otherwise stated)

Sources: Annual consolidated financial statements 2015; Annual GFS questionnaire 2015; Annual state of public finance report 2015; CHIP database; individual financial statements for public corporations; BanRep’s annual financial statements 2015 and staff estimates.

Table 3.

Public Sector Institutions and Finances, 2015

(in percent of GDP, unless otherwise stated)

				Number of entities	Revenue	Expenditure	Balance	Intra-PS expenditure	Net expenditure	percent net expenditure
Public Sector				3,822	44.3	50.8	-6.5	0.0	50.8	100.0
	General government			1,925	30.9	35.6	-4.7	0.0	35.6	70.1
		Central government		242	18.5	22.4	-3.9	5.8	16.6	32.6
			Budgetary central government	102	18.6	22.6	-4.0	9.7	13.0	25.5
			Extrabudgetary units and funds	140	3.8	3.7	0.2	0.1	3.6	7.1
		Social Security		39	8.2	8.2	0.0	1.9	6.3	12.4
		State governments		143	3.9	3.8	0.0	0.3	3.6	7.0
		Local governments		1,501	8.5	9.4	-0.9	0.2	9.2	18.0
	Central Bank			1	0.3	0.2	0.0	0.0	0.2	0.5
	Nonfinancial public corporations			1,822	12.5	15.3	-2.8	0.9	14.4	28.3
	Other financial public corporations			74	1.6	0.7	0.9	0.1	0.6	1.2

Sources: Annual consolidated financial statements 2015; Annual GFS questionnaire 2015; Annual state of public finance report 2015; CHIP database; individual financial statements for public corporations; BanRep’s annual financial statements 2015 and staff estimates.

View Table

Table 3.

Public Sector Institutions and Finances, 2015

(in percent of GDP, unless otherwise stated)

				Number of entities	Revenue	Expenditure	Balance	Intra-PS expenditure	Net expenditure	percent net expenditure
Public Sector				3,822	44.3	50.8	-6.5	0.0	50.8	100.0
	General government			1,925	30.9	35.6	-4.7	0.0	35.6	70.1
		Central government		242	18.5	22.4	-3.9	5.8	16.6	32.6
			Budgetary central government	102	18.6	22.6	-4.0	9.7	13.0	25.5
			Extrabudgetary units and funds	140	3.8	3.7	0.2	0.1	3.6	7.1
		Social Security		39	8.2	8.2	0.0	1.9	6.3	12.4
		State governments		143	3.9	3.8	0.0	0.3	3.6	7.0
		Local governments		1,501	8.5	9.4	-0.9	0.2	9.2	18.0
	Central Bank			1	0.3	0.2	0.0	0.0	0.2	0.5
	Nonfinancial public corporations			1,822	12.5	15.3	-2.8	0.9	14.4	28.3
	Other financial public corporations			74	1.6	0.7	0.9	0.1	0.6	1.2

Sources: Annual consolidated financial statements 2015; Annual GFS questionnaire 2015; Annual state of public finance report 2015; CHIP database; individual financial statements for public corporations; BanRep’s annual financial statements 2015 and staff estimates.

12. Colombia belongs to a select group of countries that publish financial statements covering the entire public sector. The annual financial statements prepared by the CGN consolidate the entire public sector, including the central bank, and are the basis for assessment of this indicator. The statements are presented in three volumes: (i) consolidated financial statements that include the public-sector balance sheet, a statement of operations, and notes; (ii) the national government financial statements consolidating all national-level entities; and (iii) financial statements at the subnational (territorial) level.

13. The 2015 financial statements covered 95 percent of public sector entities and 90 percent of public sector expenditures (Figure 4). The exclusions mainly related to a part of pension payments made by Colpensiones (Box 1) and investments undertaken by non-financial public corporations. While the budgetary transfers to the Colpensiones to bridge the gap between its income and expenditures are included, the subscriber contributions that Colpensiones received and payments it made to pensioners are not included. Investments are directly recognized in the balance sheet. These gaps can be filled with data available in other fiscal reports.

View Full Size

Coverage of Public Sector Institutions in Fiscal Reports

(in percent of expenditure)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: Staff estimates.Note: “Not Reported” refers to expenditures of units not consolidated in fiscal reports.

• Download Figure
• Download figure as PowerPoint slide

14. A weakness is the lack of harmonization of the definitions of subsectors, making it difficult to compare information across reports. Financial statements and budget execution reports follow national classification. Fiscal statistics follow old international standards (GFSM 1986) while the GFS questionnaire is aligned with GFSM 2014. Using different definitions of institutional subsectors undermines the utility of the reports for supporting the policy analysis and public debate. The absence of a cash flow statement is another weakness of the financial statements that needs to be overcome.

1.1.2. Coverage of Stocks

Good

15. The consolidated balance sheet and the annual GFS questionnaire provide detailed information on both financial and non-financial assets and liabilities, as well as net worth of all subsectors. Colombia’s balance sheet comprises most financial and non-financial assets (including natural resources) and liabilities. The 2015 balance sheet provides a detailed breakdown of the public sector assets and liabilities:

• Non-financial assets amount to 52 percent of GDP, including 5.7 percent of GDP in subsoil assets (oil, gas and minerals) and 33.2 percent of GDP in buildings, structures, and land.

• Financial assets are estimated at 89.6 percent of GDP, including 24.6 percent of GDP in debt securities, mainly BanRep holdings related to international reserves, 18.4 percent in shares and equity of corporations, 9 percent of GDP in currency and deposits, and 29.9 percent of GDP in accounts receivable.¹²

• Reported liabilities amount to 112 percent of GDP, including 38 percent of GDP in debt securities, 14 percent of GDP in loans, and 27.7 percent of GDP in accounts payable.

• Net worth and net financial worth are reported at 29.6 percent of GDP and - 22.4 percent of GDP, respectively.

16. However, the public sector balance sheet published by CGN does not reflect fully all government assets and liabilities, and their true value (Figure 5). The main differences arise because, in the financial statements:

View Full Size

Public Sector Balance Sheet Coverage in Fiscal Reports, 2015

(Percent of GDP)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

• Download Figure
• Download figure as PowerPoint slide

• Pension liabilities are not fully recorded (see Box 1). Following international standards, all future pension payments under employment-related pension schemes should be recorded as direct liabilities. The balance sheet reports only part of the pension liabilities related to public service pension schemes prior to the 1993 reform (the so called “Régimenes Empleadores and UGPP”). The CGN’s resolutions 633 determined the gradual incorporation of these liabilities to the balance sheet, in a process to be concluded by 2029. In 2015, almost half of the liability was recorded (5.9 percent of GDP). The other half (6.7 percent of GDP) was reported as a memorandum item.

• Financial liabilities related to PPPs are likely underreported. The balance sheet reports limited information on PPP liabilities since Colombia still does not fully comply with IPSAS 32 methodology.¹³ Agencia Nacional de Infraestructura (ANI), the main infrastructure agency that operates PPP contracts, is working to adopt international accounting standards for the fourth generation (4G) PPP contracts, awarded from 2015 onwards. The third-generation contracts (3G, from 2011 to 2015) will also have their accounting records reviewed in line with IPSAS 32; however, the accounting of the first and second (1G and 2G) generation of contracts, awarded up to 2011, will not change (Section 3.2.4). The size of this gap is estimated to be 1.4 percent of GDP.

• Buildings, structures and other fixed assets are recorded at historical cost. The valuation method recommended by the GFSM is market value, which can potentially alter the non-financial asset position. The MHCP is conducting an inventory of central government fixed assets to better evaluate the current condition of these assets and update their accounting records accordingly. Law 42/1993 (article 107) requires every public building to be insured against liability for loss or damage to property; such insurance contracts could be used to support a proper valuation of assets. The mission’s computation (Table 1) does not consider the effect of market valuation of these assets.

17. The CGR’s audit findings regularly indicate under- and/or over- estimation of assets and liabilities. These findings led to an adverse opinion on the 2013 accounts and to an abstention on the 2015 accounts. In 2015, it found net underestimation of assets (1.2 percent of GDP) and liabilities (1.6 percent of GDP). The inconsistencies (both underestimation and overestimation) affected several items in the balance sheet. Most noticeable were those related to tax receivables and litigation.

18. Taking a more comprehensive view of the public sector balance sheet would slightly worsen the reported fiscal position. Initial estimates of the fully consolidated public sector balance sheet suggest an estimated negative net financial worth of around 30.5 percent of GDP and a net worth in 2015 of about 21.5 percent of GDP, comparable to Peru (Figure 6). This change in the reported fiscal position is mainly due to the inclusion of pension and PPP-related liabilities in the balance sheet. A re-evaluation of some of the main assets and liabilities (e.g., fixed assets, pension liabilities) and assessment of PPP-related liabilities is needed to fully reflect Colombia’s fiscal position (Figure 7). For example, a probability-based analysis (building on historical recovery trends) could be performed to better assess the value of accounts receivables.

View Full Size

Public Sector Net Worth in Selected Countries

(percent of GDP)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: FTEs.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Public Sector Gross Liabilities in Selected Countries

(percent of GDP)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

• Download Figure
• Download figure as PowerPoint slide

Box 1.

The Colombian Pension System

The Colombian pension system is rooted in a complex set of legislations. The Constitution of Colombia (Art. 48) defines the broad contours of the system, making the pension benefits universally available subject to meeting stipulated conditions. Law 100/1993 introduced the general pension system (Sistema General de Pensiones), which was modified by subsequent legislations – principally, Law 797/2003 that removed the exception granted to the Ecopetrol employees; Law 812, also of 2003, that brought teachers under the general pension system; and Law 923/2004 that reformed the scheme applicable to the armed forces and aligned it with Law 100/1993. Further, the Legislative Act 01/2005 amended the Constitution, integrated the various pension schemes that were in operation then, and removed the possibility of any special regimes as well as collective bargaining. Act 01/2005 also stipulated an end to all transitional arrangements by 2014. The system cannot be modified, except by the Congress through a Legislative Act.

The law makes affiliation mandatory for formal workers who have the option to select one out of two alternative schemes, and to switch from one to the other under certain conditions. The system covers both private and public sector employees under equal terms. The two schemes are:

• Régimen de Prima Media – RPM: a public defined-benefit scheme operated almost entirely by Colpensiones, a public financial corporation. This scheme is funded by contributions calculated as a fixed percentage (16 percent) on the worker’s wage, of which three-fourth is borne by employers and one-fourth by employees. The scheme operates on a PAYGO basis. Contributions received by Colpensiones, and other administrators, are used to meet the payment obligations of pensioners. Shortfalls are met by the government with budgetary transfers. While the legislative Act 01 of 2005 brought all PAYGO pension funds under the overall umbrella of the RPM (within the general pension system), special entities continue to serve certain categories of subscribers (e.g. military and police).

• Ahorro Individual con Solidaridad: a defined contribution individual capitalization scheme operated by private financial corporations (AFPs). The contribution rate is the same as in the RPM. Pension is paid out of the stock of resources accumulated in the individual account of the beneficiary over the working life.

In addition, there are schemes covering specific categories of public servants that are mainly covered by the central government budget:

• Régimenes Empleadores: these schemes cover civil servants who entered the public service before 1993. These are non-contributory schemes, administered by the employing government agencies, which pay pensions partly from the general budget. Régimenes Empleadores also covers the pension liability component corresponding to collective bargaining agreements in force prior to the 2005 legislative act.

• The UGPP (Unidad de Gestión Pensional y Parafiscales) manages pensioners belonging to entities falling under the employer’s pension regime that were subsequently dissolved.

The total actuarial pension liability in Colombia, considering the general pension system, Régimenes Empleadores and UGPP, reached COP 839 trillion (104.7 percent of GDP) in 2015.

Pension Scheme		Actuarial Liability
		COP trillion	% of GDP
Sistema General de Pensiones		738	92.1
	Régimen de Prima Media - Colpensiones	555	69.4
	Teachers	102	12.8
	National Police	40	5.0
	Military personnel	37	4.7
	Congress’ employees	3	0.3
Régimenes Empleadores - before 1993		73	9.2
	Ecopetrol	10	1.3
	Parapat-Telecom	10	1.2
	Other	53	6.6
UGPP		28	3.5
TOTAL		839	104.7

View Table

Pension Scheme		Actuarial Liability
		COP trillion	% of GDP
Sistema General de Pensiones		738	92.1
	Régimen de Prima Media - Colpensiones	555	69.4
	Teachers	102	12.8
	National Police	40	5.0
	Military personnel	37	4.7
	Congress’ employees	3	0.3
Régimenes Empleadores - before 1993		73	9.2
	Ecopetrol	10	1.3
	Parapat-Telecom	10	1.2
	Other	53	6.6
UGPP		28	3.5
TOTAL		839	104.7

According to GFSM 2014, the statistical treatment of pension schemes depends on whether the scheme is contributory or not, whether it is a defined-benefit or defined-contribution, and whether it is a social security or employment-related scheme. For defined-benefit schemes, the statistical treatment depends on the type of beneficiaries: when the beneficiary is the general population, or a large segment of the general population, the scheme is considered a social security scheme; whereas if individuals, households, or a group of employees are eligible to receive social benefits, the scheme would be considered an employment-related social insurance scheme. Under social security schemes, the link between benefits and contributions is not considered sufficiently strong to give rise to a financial claim on the part of contributors. As a result, no liabilities are recorded, but an estimate equal to the net implicit obligations for future social security benefits should be presented as a memorandum item to the balance sheet. Employment-related pension schemes are, on the contrary, considered to involve a contractual liability towards employee and registered as liabilities.

According to this criterion, the actuarial liabilities of the Régimenes Empleadores and UGPP would be recognized in the financial statements of the public sector as firm liabilities. The treatment of entities serving armed forces, teachers, and Congress staff requires a complex interpretation of their legal status. Per the Constitutional requirements, viewing these as administrative arrangements within the general pension system-designed to service certain large group of beneficiaries more efficiently, would imply that the actuarial liabilities of the entire general pension system (92.1 percent of GDP) are treated as contingent liabilities and disclosed in the financial statements.

Source: Annual Consolidated Financial Statements, 2015, on GFSM 2014 basis.

1.1.3 Coverage of Flows

Basic

19. The incomplete accrual recording, and the absence of a cash flow statement to accompany the consolidated financial statements, implies that Colombia does not meet the most advanced practices under the FTC. Fiscal reports mainly used for policy decisions are prepared on a modified cash basis while financial statements, although on accrual basis, are not complete. Annual financial statements cover accrual flows, including accounts payable/receivable, depreciation of fixed assets, and other economic flows (revaluations, exchange rate variation), but major omissions are social contributions to the PAYGO pension scheme and pension payments made by Colpensiones.

20. While most of cash to accrual adjustments can be obtained from the annual statements, some relevant accrued flows are not reported. Cash to accrual adjustments for each subsector (central government, social security, regional governments, and corporations) are derived by comparing revenue and expenditure figures from the accounts and the fiscal statistics. Nevertheless, important accrued flows related to pension liabilities (Régimenes Empleadores and UGPP) and PPPs are partially reported in the statements. Initial estimates, based on limited data, suggest the accrual of pension liabilities amounts to 2.0 percent of GDP every year and the annual investment in PPP could reach 0.8 percent of GDP (Table 4).

Table 4.

Cash to Accrual Adjustments

(Percent of GDP, unless otherwise stated)

Sources: Fiscal statistics; Annual consolidated financial statements 2015 and staff estimates.

Table 4.

Cash to Accrual Adjustments

(Percent of GDP, unless otherwise stated)

NFPS Overall Balance (below-the-line, modified-cash, Cierre Fiscal)			-3.4
	Cash-accrual adjustments on recognized transactions		-4.0
		Central government	-0.8
		Social security	-0.4
		States and municipalities	-0.3
		Non-financial public corporations	-2.5
NFPS Net Lending/Borrowing			-7.4
	Coverage adjustments - financial corporations		0.9
Public Sector Net Lending/Borrowing			-6.5
	Additional accrual adjustments Accrual of pension entitlements		-2.0
		PPP investment	-0.8
Augmented Public Sector Net Lending/Borrowing FTE			-9.3

Sources: Fiscal statistics; Annual consolidated financial statements 2015 and staff estimates.

View Table

Table 4.

Cash to Accrual Adjustments

(Percent of GDP, unless otherwise stated)

NFPS Overall Balance (below-the-line, modified-cash, Cierre Fiscal)			-3.4
	Cash-accrual adjustments on recognized transactions		-4.0
		Central government	-0.8
		Social security	-0.4
		States and municipalities	-0.3
		Non-financial public corporations	-2.5
NFPS Net Lending/Borrowing			-7.4
	Coverage adjustments - financial corporations		0.9
Public Sector Net Lending/Borrowing			-6.5
	Additional accrual adjustments Accrual of pension entitlements		-2.0
		PPP investment	-0.8
Augmented Public Sector Net Lending/Borrowing FTE			-9.3

Sources: Fiscal statistics; Annual consolidated financial statements 2015 and staff estimates.

21. Recognizing these additional accrued expenses would increase Colombia’s public sector deficit. On an accrual basis, the NFPS deficit would reach 7.4 percent of GDP, 4 percentage points higher than currently reported in fiscal reports. Expanding the coverage to include the financial corporations would reduce the deficit by 0.9 percentage point. Finally, incorporating the two major unreported accrued flows (pension entitlements and PPP contracts) would turn the public sector net lending/borrowing into a deficit of around 9.3 percent of GDP.

22. Apart from cash to accrual adjustments, some important flows across subsectors are not identified in the current fiscal reports. Central government transfers to pension funds, such as Colpensiones, and to the health insurance system are not explicitly reported. These flows are consolidated into a recurrent expenditure line item in the quarterly fiscal bulletin (Cierre Fiscal), making it difficult to reconcile the figures reported by Colpensiones and the health system. Such transfers amounted to around COP 20 trillion in 2015 and are expected to grow in the coming years. Fiscal reports would improve in transparency if these transactions were fully disclosed in additional tables.

1.1.4. Coverage of Tax Expenditures

Good

23. Annual estimates of the main tax expenditures of the central government are published in the MFMP (Figure 8). Each tax expenditure is analyzed by category of expenditure (e.g., credit, deduction), by economic sector (e.g., education, construction) or by legal beneficiaries.¹⁴ In 2015, estimated revenue losses from tax expenditures amounted to about COP 62,392 billion (7.8 percent of GDP or 54 percent of tax revenue) higher than in most neighboring countries (Figure 9). The losses are concentrated in finance and insurance activities, energy and mining industries, and general tax expenditures that apply to public administration, defense and social security. The quantum of tax expenditures has been increasing in recent years. For example, according to MFMP, the fiscal costs of VAT exclusions increased by more than half during the five-year period between 2011 and 2015.

View Full Size

Tax Expenditures, 2015

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: MFMP 2016.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Tax Expenditures in Selected Countries

(Percent of GDP)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: IMF Fiscal Monitor April 2011 and UK FTE.Note: All estimates are for 2010, except when stated otherwise.

• Download Figure
• Download figure as PowerPoint slide

24. Introducing explicit controls and budgetary objectives for the size of tax expenditures would bring Colombia in line with advanced practices. Tax expenditure estimates cover the central government only and have no explicit budgetary limit. They do not cover taxes levied by subnational governments, which remain unquantified. In addition, most tax expenditures related to income tax and value-added tax (VAT) are not subject to any time limit. Moreover, time limits, where imposed, can be extended by law. The authorities have, thus, experienced difficulties capping the growth of tax expenditures.

25. The government’s estimates of tax expenditures could be improved to support cost-efficiency analyses. A recent study by the World Bank identified more than 200 separate tax expenditures in Colombia, and also assessed their value-for-money.¹⁵ The number of tax expenditures has grown substantially in recent years without a thorough ex-ante or ex-post analysis of their total fiscal cost, or an evaluation of their economic impact. In addition, the cost analysis conducted within the MFMP is a static one and does not account for behavioral changes that may result from the deductions or exemptions provided through the tax system. To facilitate decisions between using the tax system or direct subsidies to achieve policy objectives, the government could consider generalizing its use of cost-efficiency analysis. The comparative cost of using tax expenditures or direct subsidies should also be assessed.

1.2.1. Frequency of In-Year Reports

Not Met

26. Several in-year reports are produced by the various general directorates of the MHCP but none meets the coverage and timeliness required under the FTC. The quarterly fiscal reports produced by the DGPM are the only in-year reports providing a comprehensive view of the revenues, expenditures and financing. The DGPM produces these reports both for the central government and combined public sector operations. The reports are produced on a quarterly basis but with a lag of more than a quarter. The timeliness of these reports is affected by the time taken in the reconciliation of above-the-line deficit and below-the-line financing figures obtained from the Central Bank, which needs to be addressed by further strengthening the cooperation between the MHCP and Central Bank. Other reports, such as the monthly reports by the General Directorate of National Budget (DGPPN) on the execution of the expenditure budget, DIAN’s reports on tax collections, DGCPTN’s quarterly debt reports, DNP’s reports on the execution of investment projects, are brought out at a higher frequency and score better on timeliness, but provide only a partial view of the fiscal operations.

1.2.2. Timeliness of Annual Financial Statements

Advanced

27. The Constitution (Article 354) requires submission of the audited consolidated financial statements within six months of the end of the fiscal year. To meet this deadline, the CGN is required to submit the consolidated financial statements to the CGR latest by May 15 every year. These dates are met invariably. The CGR’s report is placed before Congress when it convenes in July. The financial statements are published simultaneously on the CGN’s website and the audit report on the CGR’s website.

1.3.1. Classification

Not Met

28. The lack of a harmonized classification system based on international standards remains a major weakness and impedes the government’s ability to produce quality fiscal information on a comparable and consistent basis during the annual fiscal cycle. There is no consistent application of economic, functional or program classifications that meets international standards. In compliance with the requirements of the Organic Budget Law (Law 38, 1989; Article 7 and 23) the budget is classified by institutions and economic nature of expenditure. For each institution, the budget is broken down into three broad components: operational, debt service and investments. The operational budget is further classified by economic types using a hierarchical structure with increasing granularity. The investment budget is classified by programs, sub-programs and projects, but an economic classification of programs/projects is missing. A functional view of expenditure—using 12 functions—is also provided in the budget documents. The functional classification broadly corresponds to the Classification of Functions of Governments (COFOG) at a higher level, but deviates at lower levels. There is no reporting of budget execution by functions. The monthly budget execution reports and the quarterly fiscal balance reports produced by the MHCP follow the budget classification. The financial statements follow a different classification structure, which is designed to meet the requirements of the accounting standards. A bridge table is used to map the accounting classification to the classification used in the GFSM.

29. Recognizing this weakness, the authorities have started a project to harmonize the classification systems used by different agencies. The MHCP-led CIEFP is tasked to propose policies and strategies that will enable the harmonization of fiscal statistics and the application of consistent, and homogeneous methodologies compatible with international standards on data dissemination and transparency. Based on a 2015 diagnostic study, an action plan has been developed. Technical working groups have begun working on six important areas: sectoral classification of entities; harmonization of classification systems; measurement of public debt; identifying the boundaries of the social protection sector; developing an integrated budget and accounting classification for the territorial entities; and consolidation of accounts.

30. While full harmonization would appear to be a medium-term objective, early achievements indicate good progress. A notable accomplishment has been the establishment of a Bureau, led by the CGN and supported by the National Statistics Department (DANE) and MHCP, for developing a classification of public sector institutions. A comprehensive list of public entities, categorized in accordance with the GFSM 2014, has been developed and coded using a unique identifier (CUIN). A new economic classification of budget, aligned with the GFSM 2014 and IPSAS, has been drafted, and will be embedded in the financial information system (SIIF) for a uniform application. Work is in progress on developing a unified classification structure that will harmonize the budget and the financial and statistical reporting. A policy note is being developed for submission to the National Council for Economic and Social Policy (CONPES) during the second half of the year 2017. Once approved, the note will serve as the framework for organizing and managing public financial management information.

1.3.2. Internal Consistency

Good

31. The fiscal reports present two of the three reconciliations demanded by this indicator. The quarterly fiscal reports present a reconciliation of the above-the-line deficit with below-the-line financing figures obtained from the central bank. Discrepancy between the two measures is reported as “statistical discrepancy” (Table 5), which has been significant in some years, and needs more rigorous reconciliation. The quarterly debt reports contain data on the debt issued during the period and the opening and closing stock positions. The reconciliation of debt issuance and stock can be derived from this information, although it is not explicitly presented. The presentation in this report could be enhanced by including a table that shows changes in debt stock due to transactions during the period and valuation changes, including those due to exchange rate variation. There is no report that reconciles financing with debt stock.

Table 5.

Reconciliation Between Deficit and Financing (COP Billions)

Source: MHCP.

Table 5.

Reconciliation Between Deficit and Financing (COP Billions)

Year	Fiscal Balance	Financing	Discrepancy
2016	-19,681	-19,037	644	3%
2015	-26,417	-27,163	746	3%
2014	-13,807	-13,724	83	1%
2013	-7,003	-6,247	756	11%
2012	3,581	1,862	1,719	48%
2011	13,604	12,585	1,019	7%

Source: MHCP.

View Table

Table 5.

Reconciliation Between Deficit and Financing (COP Billions)

Year	Fiscal Balance	Financing	Discrepancy
2016	-19,681	-19,037	644	3%
2015	-26,417	-27,163	746	3%
2014	-13,807	-13,724	83	1%
2013	-7,003	-6,247	756	11%
2012	3,581	1,862	1,719	48%
2011	13,604	12,585	1,019	7%

Source: MHCP.

1.3.3. Historical Revisions

Basic

32. Major revisions to fiscal data are disclosed, but no further explanations or bridge tables between old and new figures are provided. Revisions are reported in a short section—Fe de Erratas—in the quarterly fiscal bulletin (Cierre Fiscal), but without an explanation of the changes. Such revisions have, however, had a minor impact on the final fiscal balance. The MHCP updates retroactively its online databases containing fiscal aggregates. Fiscal statistics are also updated as they move from provisional to final status; however, in this case, the changes are not reported. To date, no revision has affected the compliance with the fiscal rule.

1.4.1. Statistical Integrity

Basic

33. Fiscal statistics have been disseminated in line with IMF’s Special Data Dissemination Standards (SDDS) since 2000. Colombia subscribed to the SDDS in 1996 but met all SDDS requirements only in 2000. Fiscal data reported under the SDDS framework are the same as those released in the quarterly fiscal bulletin (Cierre Fiscal).

34. More recently, Colombia has been submitting data to the IMF’s GFS Yearbook under the GFSM 2014 reporting format; however, there is room for improvement in compliance with the GFSM guidelines. The accuracy of the CGN’s compilation procedures is limited by the structure of the chart of accounts and current accounting practices. Furthermore, the CGN does not perform consistency checks in a systematic manner; such checks would lead to improvements in the classification of units and/or statistical treatments of operations, particularly the reciprocal transactions within subsectors. Checks should be undertaken against other sources of fiscal data, i.e., external sector and monetary statistics, national accounts and cash-based fiscal statistics compiled by the DGPM.

35. Entrusting the responsibility to produce fiscal statistics to a specific government agency, and if possible to a professionally independent body, would be in alignment with international good practice, and result in a higher score under the Code. An important element in statistical integrity is the professional independence of the agency responsible for producing fiscal statistics that will allow it to produce reliable statistics.¹⁶ Multiplicity of agencies compiling and publishing fiscal data using different data sources, methodologies and reporting practices, can create confusion, and be detrimental to transparency objectives.

36. CIEFP has been working to improve the inter-agency coordination. The committee should receive strong institutional support and must have access to all resources needed to carry out its activities. Regarding the medium-term action plan, priority should be given to integrating budget execution and accounting by: i) harmonizing the chart of accounts and codification of both systems; and ii) ensuring full interoperability between SIIF and CHIP.

1.4.2. External Audit

Basic

37. The CGR—the supreme audit institution—has not been in a position to certify financial statements in two recent years. In 2015, while the CGR certified the reasonableness of the treasury balance, it abstained from providing an opinion on the consolidated financial statements due to the lack of adequate evidence to judge the reasonableness of several accounts. A large proportion of entities received qualified opinion, pointing to control weaknesses in these entities. During the year, the CGR audited 95 entities representing 87 percent of the total public sector assets. Of these, over half (49)—representing 46 percent of the total assets and 27 percent of total liabilities of the 95 audited entities—received a qualified or adverse opinion. Another major issue was the lack of adequate explanation of the changes in the financial statements due to the adoption of new accounting standards (Resolution 743 of 2013) for the public listed companies. The CGR also commented on large accounts payable, including those pertaining to the government pension fund administrator (Colpensiones), weaknesses in the consolidation process that resulted in large unreconciled related-party balances (COP 43.2 trillion), and incorrect representation of assets (net underestimation by COP 9.7 trillion) and liabilities (net underestimation by COP 12.9 trillion). In 2013, the CGR gave a negative opinion on the accounts for similar reasons.

38. Aiming to modernize its accounting practices, Colombia is undertaking a gradual transition to IPSAS. The CGN published in 2013 an action plan to adopt International Public Sector Accounting Standards containing an implementation schedule up to 2019.¹⁷ The transition is gradual for the different types of public entities, incorporated corporations, non-incorporated corporations and general government units. The CGN issued three resolutions in the period 2013-2015, establishing the new accounting standards for these entities. The first, Resolution 743/2013, became operational in 2015 for the incorporated public corporations. Resolution 414/2014 came into force in 2016, and applied the international standards to the unincorporated corporations. The implementation for the general government units, originally scheduled to take effect in 2018, has been postponed to 2019 (Resolutions 533/2015 and 113/2016). It is expected that on successful completion of transition, the accounts will present a more consistent and credible position in accordance with international standards.

39. Strengthening of internal controls, particularly at the sub-national level, would be important to improve the quality of the financial statements and to avoid the recurrence of audit qualification. The nature and recurrence of audit observations point to possible control weaknesses. The large and increasing number of entities receiving such qualification should be a matter of concern. Strengthening controls would require improving procedural weaknesses, if any, addressing capacity constraints, providing better guidance and documentation, and stronger accountability arrangements. More specifically: interoperability between CHIP and information systems used by the subnational entities, including the public corporations at the subnational level, should be improved; universal introduction of financial information systems at the subnational levels should be pursued further; and data validation and controls in CHIP should be improved.

1.4.3. Comparability of Fiscal Data

Basic

40. Fiscal reports in Colombia score low on comparability. Reports are prepared by different agencies with divergent institutional coverage, accounting base and methodology, and classification. There are at least three distinct reporting plans:

• Reports, such as the MFMP, quarterly fiscal bulletin, and the annual compliance report, focus on fiscal aggregates broken down by sub-sectors. The definitions of subsectors and the economic classification follow old international standards (GFSM 1986).

• The annual budget, the DGPPN’s monthly budget execution reports, the DNP’s monthly report on the execution of investment projects, and the CGR’s annual and in-year reports on the budget execution follow the budget classification.

• The CGN’s annual financial statements are prepared on an accrual basis using the accounting classification with a view to meeting the requirements of the accounting standards, and used for compiling GFS data.

Due to the coverage and methodological differences, the deficit figures, as well as other main fiscal aggregates disclosed in these reports, differ from each other (Table 6).¹⁸ Likewise, the debt reports produced by the CGR and the MHCP differ in their coverage.

Table 6.

Comparison Between DGPM and CGR Fiscal Statistics

(Percent of GDP, unless otherwise stated)

Source: Quarterly fiscal bulletins and Annual State of Public Finance report 2015.

Table 6.

Comparison Between DGPM and CGR Fiscal Statistics

(Percent of GDP, unless otherwise stated)

Sector/Subsector		Fiscal Balance - 2015 (% of GDP)
		CGR	DGPM
Public Sector		-3.73	-3.40
Non Financial Public Sector			-3.40
General Government
	Budgteary Central Government	-296	-3.00
	CG EBUs - Public Establishments	0.32	-0.10
	Subnationals - Budgetary		-0.50
	Subnat. EBUs - Public Establishments	0.01	0.00
	Social Security	0.27	0.40
Non Financial Public Corporations
	Central level	0.02	-0.20
	Subnat. Level	-1.00	-0.10
Fogafin		0.00	0.10
Banco de la República		-0.05	0.00

Source: Quarterly fiscal bulletins and Annual State of Public Finance report 2015.

View Table

Table 6.

Comparison Between DGPM and CGR Fiscal Statistics

(Percent of GDP, unless otherwise stated)

Sector/Subsector		Fiscal Balance - 2015 (% of GDP)
		CGR	DGPM
Public Sector		-3.73	-3.40
Non Financial Public Sector			-3.40
General Government
	Budgteary Central Government	-296	-3.00
	CG EBUs - Public Establishments	0.32	-0.10
	Subnationals - Budgetary		-0.50
	Subnat. EBUs - Public Establishments	0.01	0.00
	Social Security	0.27	0.40
Non Financial Public Corporations
	Central level	0.02	-0.20
	Subnat. Level	-1.00	-0.10
Fogafin		0.00	0.10
Banco de la República		-0.05	0.00

Source: Quarterly fiscal bulletins and Annual State of Public Finance report 2015.

41. Improving comparability of reports would require presenting reconciliation tables for the main fiscal aggregates between the different reporting plans. The financial statements could include a reconciliation of main aggregates with the annual fiscal bulletin. Similarly, the GFS and the CGR’s budget execution reports should present a reconciliation with the outturns reported in the fiscal bulletin, and should reconcile with the annual financial statements.

2.1.1. Budget Unity

Basic

49. The budget process in Colombia is highly fragmented with four separate types of budgets:

• The central government budget (PGN) incorporates the gross revenues, expenditures, and financing of central government ministries and agencies; approved by Congress.

• The royalties budget (Sistema General de Regalias, SGR), covering the receipt and allocation of natural resource revenues, approved by Congress every two years;

• The budgets of subnational governments (comprising governorates, departments, and municipalities) approved by local legislatures; and

• The budgets of 29 public corporations that are approved by the Supreme Council for Fiscal Policy (CONFIS).

50. Although Colombia is a unitary state, a sizeable proportion of public spending is carried out by the regional and local administrations through two separate funds. Given their limited tax capacity, states and municipalities are heavily reliant on funding by the two national resource sharing systems, the Sistema General de Participation (SGP) and the SGR, which transfer approximately 10 percent of GDP in resources to subnational governments. The SGR draws its revenues from oil and other natural resources. State and local governments account respectively for around 10 percent and 20 percent of total general government expenditure (net of intra-government transfers). Non-financial public corporations are an important component of the Colombian public sector, with over 15 percent of the GDP in expenditure, half of which is by Ecopetrol, and about one-third of which pertains to hospitals and health service providers.

51. No information on the revenues or spending of extrabudgetary units is included in the budget documents. The PGN excludes some 111 extrabudgetary units (EBUs), which represent about 10 percent of total revenues and spending by the central government (Table 9).¹⁹ These entities include a wide range of regulatory bodies, research and training institutions, government commissions, and entities delivering public services.²⁰ The reporting of their financial conditions and performance is variable, and is not included with budget documentation: some publish quarterly and annual reports, but many only report on an annual basis. In addition, some ministries, agencies and EBUs receive revenues that are earmarked for specific purposes (e.g., road maintenance, hospital expenses, educational fees). Earmarked revenues account for some 7 percent of total central government revenues, and full information is disclosed in the budget documents.²¹ The social security system (healthcare, retirement pensions) is also outside the PGN, apart from annual transfers made by the government to meet any deficit in social security contributions. The MFMP includes a chapter on social security finances. The government also publishes a half-yearly social security bulletin that provide more in-depth analysis of the social security system.

Table 9.

Extrabudgetary Spending by Central Government Units

Source: MHCP.

Table 9.

Extrabudgetary Spending by Central Government Units

Number of extrabudgetary units (EBUs)		Annual spending by EBUs (COP billion)		Total spending by PGN (COP billion)		Spending by EBUs as a ratio of PGN spending (percent)
2014	2015	2014	2015	2014	2015	2014	2015
110	111	25,081	25,151	257,400	245,330	9.74	10.25

Source: MHCP.

View Table

Table 9.

Extrabudgetary Spending by Central Government Units

Number of extrabudgetary units (EBUs)		Annual spending by EBUs (COP billion)		Total spending by PGN (COP billion)		Spending by EBUs as a ratio of PGN spending (percent)
2014	2015	2014	2015	2014	2015	2014	2015
110	111	25,081	25,151	257,400	245,330	9.74	10.25

Source: MHCP.

2.1.2. Macroeconomic Forecasting

Advanced

52. Colombia’s fiscal and budget documentation provides a clear and comprehensive analysis of the key macroeconomic forecasts and assumptions underlying the budget. The MFMP includes forecasts of key macroeconomic aggregates (including GDP, inflation, current account balance, and commodity prices) ten years ahead. These forecasts are consistent with the projections and assumptions developed in the National Development Plan (PND), which has a four-year time horizon. The MHCP also presents the government’s Financial Plan for the next budget year. Along with the Annual Operating Investment Plan (POAI) and the national budget (PGN), the Financial Plan is one of the government’s main budgetary tools²² and covers the consolidated public sector (Article 6 of Decree 111 of 1996). The MFMP is generally updated in the second half of the year whereas the Financial Plan is first presented in December, and usually revised in March (Section 2.3.1).

53. Although relatively accurate, both Colombia’s one year-ahead and medium-term GDP growth forecasts tend to show an optimistic bias in recent years. The mean deviation of real GDP forecasts one year ahead was only 0.1 percentage points between 2009-2016 but 1.0 percentage points between 2012-2016.²³ The government’s one-year-ahead forecasts of real GDP growth have hence been relatively accurate, although more optimistic than the forecasts of the central bank, the international financial institutions, and independent agencies in recent years (Table 10). Projections of real GDP over the medium term also show a slight optimistic bias (Figure 10), partly explained by the unforeseen oil price shock in 2015.

View Full Size

Evaluation of Macroeconomic and Fiscal Forecasts, 2009-2016

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source. MHCP, FAD staff calculations.

• Download Figure
• Download figure as PowerPoint slide

Table 10.

One-year-ahead Forecasts of Real GDP Growth, 2013-2015 (percent)

Source: FAD staff.

Table 10.

One-year-ahead Forecasts of Real GDP Growth, 2013-2015 (percent)

	2013	2014	2015
Actual	4.9	4.4	3.1
Forecasts:
Authorities (MHCP)	4.8	4.7	4.8
IMF	4.4	4.2	4.5
World Bank	4.5	4.3	4.5
Consensus	4.3	4.5	4.6

Source: FAD staff.

View Table

Table 10.

One-year-ahead Forecasts of Real GDP Growth, 2013-2015 (percent)

	2013	2014	2015
Actual	4.9	4.4	3.1
Forecasts:
Authorities (MHCP)	4.8	4.7	4.8
IMF	4.4	4.2	4.5
World Bank	4.5	4.3	4.5
Consensus	4.3	4.5	4.6

Source: FAD staff.

54. Inflation forecasts rely mainly on the central bank’s mid-range inflation target over the medium term, and have tended to underestimate inflation in recent years. The average deviation of inflation forecasts one year ahead was minus 0.8 percentage points between 2009-2016 and minus 1.5 percentage points between 2013-2016. Revenue forecasts also exhibit a recent optimistic bias (averaging about 0.6 percent of GDP in the past three years), but forecasts of spending by the central government one year ahead have been reliable, with an average deviation of 0.1 percentage points between 2009-2016. The reconciliation of different vintages of macroeconomic and fiscal forecasts is discussed in Section 2.4.3.

2.1.3. Medium-Term Budget Framework (MTBF)

Basic

55. The MGMP published by the government annually has many positive characteristics. The MGMP for 2016 sets out: (i) the macroeconomic and fiscal context, based on the assumptions and projections described in the MFMP (see Section 2.3.1); (ii) an analysis of fiscal constraints in the medium term; (iii) an assessment of the objectives of government programs and projects, drawn largely from the PND; and (iv) projections of medium-term spending plans for the period 2017-2020, broken down by investment and operating expenses. The document also sets out indicative spending ceilings for 29 sectors (education, defense, health, transport, and so on) for the forthcoming budget year and three additional years to guide ministries and public institutions in preparing their budgets.

56. The MGMP, however, displays some important gaps when compared with good international practice.²⁴ The focus of the MGMP is on investment spending, and the document does not provide a full coverage of all expenditures. It adopts the classification of sectors used in the PND, which currently differs from the classification scheme used in the annual budget, though as noted in Section 1.3.1 of this report the government is discussing options for harmonizing the classification schemes used in budgeting, accounting and fiscal reporting. No breakdown is provided by economic category or program. The MGMP presents indicative ceilings for three out-years, but these ceilings have been substantially revised in recent years (Table 11), and do not provide a credible basis for planning spending over the medium term.²⁵ The processes of preparing the MGMP and the annual budget are largely separate. Finally, the government does not publish any reconciliation of the projections contained in the MGMP with those in the previous year’s document.

Table 11.

Revisions to Medium-Term Expenditure Ceilings, 2013 to 2016 (COP billion)

Source: MHCP and FAD staff calculations. Note: The figures show changes in the estimates of total spending made in successive vintages of the MGMP. The final row of figures measures the percent difference between the outturn spending and the initial estimate of spending made four years previously.

Table 11.

Revisions to Medium-Term Expenditure Ceilings, 2013 to 2016 (COP billion)

	2013	2014	2015	2016
Out-year 3 estimate	120.4	125.4	141.6	157.3
Out-year 2 estimate	117.9	134.1	148.6	166.2
Out-year 1 estimate	127.8	145.6	159.5	169.8
Budget year estimate	138.6	155.4	167.3	167.3
Outturn	144.4	156.0	163.4	167.3
Change (percent)	+ 19.9	+ 24.4	+ 15.4	+ 6.4

Source: MHCP and FAD staff calculations. Note: The figures show changes in the estimates of total spending made in successive vintages of the MGMP. The final row of figures measures the percent difference between the outturn spending and the initial estimate of spending made four years previously.

View Table

Table 11.

Revisions to Medium-Term Expenditure Ceilings, 2013 to 2016 (COP billion)

	2013	2014	2015	2016
Out-year 3 estimate	120.4	125.4	141.6	157.3
Out-year 2 estimate	117.9	134.1	148.6	166.2
Out-year 1 estimate	127.8	145.6	159.5	169.8
Budget year estimate	138.6	155.4	167.3	167.3
Outturn	144.4	156.0	163.4	167.3
Change (percent)	+ 19.9	+ 24.4	+ 15.4	+ 6.4

Source: MHCP and FAD staff calculations. Note: The figures show changes in the estimates of total spending made in successive vintages of the MGMP. The final row of figures measures the percent difference between the outturn spending and the initial estimate of spending made four years previously.

57. Bringing Colombia’s MGMP in line with best-practice MTBFs would require changes in several areas, and substantial capacity building in the MHCP and spending agencies. Differences between the MGMP and MTBFs produced in advanced countries are substantial, as shown in Table 12. Similar enhancements of medium-term budgeting in Colombia could be considered.

Table 12.

Differences Between Colombian MGMP and Advanced MTBF Model

Source: FAD staff.

¹Some advanced countries have adopted a fixed framework for the MTBF, i.e., one that is only changed at the end of the 3-4-year period.

²Some advanced countries have binding ceilings for the out-years, e.g., Sweden, U.K. (partially).

Table 12.

Differences Between Colombian MGMP and Advanced MTBF Model

MTBF Characteristics	Colombia	Advanced
4- or 5-year MTBF framework	Yes	Yes
Indicative ceilings for out-years	Yes	Yes1
Full alignment with government’s fiscal policy objectives and macroeconomic forecasts	Yes	Yes
Spending estimates for out-years rolled over from one MTBF to the next	No	Yes2
Budget preparation process for capital and recurrent spending fully integrated	No	Yes
Definition of capital and recurrent spending aligned with international standards	No	Yes
A single process for preparing MTBF and annual budget, and fully integrated documentation	No	Yes
Reliable forward estimates of spending in out-years	Partly	Yes
Planning margins or planning reserves	No	Yes
Carry forward of spending from one year to the next, with well-defined restrictions	No	Yes

Source: FAD staff.

¹Some advanced countries have adopted a fixed framework for the MTBF, i.e., one that is only changed at the end of the 3-4-year period.

²Some advanced countries have binding ceilings for the out-years, e.g., Sweden, U.K. (partially).

View Table

Table 12.

Differences Between Colombian MGMP and Advanced MTBF Model

MTBF Characteristics	Colombia	Advanced
4- or 5-year MTBF framework	Yes	Yes
Indicative ceilings for out-years	Yes	Yes1
Full alignment with government’s fiscal policy objectives and macroeconomic forecasts	Yes	Yes
Spending estimates for out-years rolled over from one MTBF to the next	No	Yes2
Budget preparation process for capital and recurrent spending fully integrated	No	Yes
Definition of capital and recurrent spending aligned with international standards	No	Yes
A single process for preparing MTBF and annual budget, and fully integrated documentation	No	Yes
Reliable forward estimates of spending in out-years	Partly	Yes
Planning margins or planning reserves	No	Yes
Carry forward of spending from one year to the next, with well-defined restrictions	No	Yes

Source: FAD staff.

¹Some advanced countries have adopted a fixed framework for the MTBF, i.e., one that is only changed at the end of the 3-4-year period.

²Some advanced countries have binding ceilings for the out-years, e.g., Sweden, U.K. (partially).

2.1.4. Investment Projects

Basic

58. A wide range of information on public investment projects is publicly available. The PND contains details of planned public investment projects over a four-year period, including a list of ongoing and pre-selected investment projects, with a projection of their overall costs, as well as their planned beginning and end dates, classified by program and sector. The POAI provides both an aggregate view and individual information on planned investment spending, though only amounts to be appropriated for the budget year are disclosed. The DNP maintains a project database (SPI) that contains summary information on individual projects financed from the national budget, and a database of PPPs (the RUAPP), but without details of the multiyear fiscal obligations associated with these projects. A database of all procurement contracts (SECOP) is available on the web site of the National Procurement Agency.²⁶

59. The MFMP (Chapter VI) provides information on the total spending of all public investment projects for the next 30 years, with data on year-by-year spending presented for the early years. MFMP presents information on authorized multiyear expenditure commitments (vigencia futuras) for individual major projects, but data on the amount of spending already incurred, and estimated expenditure over the remaining life cycle of projects, is not available. Information on all these elements, however, is published for individual PPPs (see Section 3.2.4).

60. There is a standard requirement for entities to carry out economic and financial appraisals of new investment projects, using a methodology approved by the DNP. There is no differentiation between projects of different sizes. The results of project appraisals are not made publicly available, however, either before or after the projects have been approved.

61. In some cases, projects are approved without a prior assessment of the merits of the PPP mode compared to direct government investment. Good practice in public investment management is to first appraise whether a proposed project is likely to generate net social benefits, and then to rank and prioritize projects. At a second stage, an assessment is made as to whether these projects should be implemented through direct government investment or the PPP mode, using a methodology referred to as the public sector comparator. This methodology has not been promulgated in Colombia, and it appears that the many “unsolicited” projects²⁷ which are categorized as not requiring any public funding—which can include public contributions up to 20 percent of total costs—are being approved with only a limited review of their merits. There is a comparatively large number of unsolicited projects in Colombia (Figure 11).

View Full Size

Unsolicited PPP Proposals in Colombia in Comparative Perspective

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: DNP Estimates.

• Download Figure
• Download figure as PowerPoint slide

62. There is a high percentage of direct contracting of public procurement contracts (Figure 12). The legal framework contains a presumption in favor of open and competitive tender, but there are numerous exceptions allowed by the law.²⁸ In practice, in 2016, 40 percent of contracts by value were categorized as being procured through direct award (there is no separate data on major projects). In addition, contracts within the “special regime” (for entities of mixed public and private ownership) may also be procured through direct awarding, but the National Public Procurement Agency does not collect data that distinguishes between procurement methods within the special regime. The 2015 PEFA assessment established, from an assessment of procurement methods used for large projects in five large ministries, that more than 50 percent of contracts in 2014 used the direct contracting method.

View Full Size

Distribution of Public Procurement by Modality, 2016

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: National Public Procurement Agency.

• Download Figure
• Download figure as PowerPoint slide

2.2.1 Fiscal Legislation

Advanced

63. The framework for budgeting and fiscal management in Colombia is comprehensive, clearly established in law, publicly available, but highly complex. The Constitution includes many articles on budgeting and finance, which are supplemented by organic laws²⁹ on budgeting (e.g., Law 38 of 1989 and Law 179 of 1994) and on national development planning (Law 152 of 1994), as well as lower-level decrees and resolutions. There are more than 20 principle laws and decrees which are summarized in Annex III. This framework covers all aspects of budgeting, development planning, fiscal transparency, and fiscal reporting, as well as many aspects of fiscal risk management, including PPPs for example.

64. The legal framework sets out inter alia a clear timetable for budget preparation and approval, which has been regularly complied with in recent years (see Section 2.2.2). It also defines the key content requirements of the budget, and the Congress’s powers of amendment which both in law and in practice are extremely limited. The Congress may not increase any spending proposals made by the executive, and may only decrease items of spending with the approval of the MHCP (further discussed in Section 2.4.2).

2.2.2 Timeliness of Budget Documents

Advanced

65. The legal framework includes specific deadlines on the preparation and approval of the budget that have been fully respected in recent years. As shown in Table 13, the budget is submitted to the Congress and made available to the public five months before the start of the fiscal year, and is approved and published two months before the start of the year, although the full details of the budget are not made available until the end of December. The debate in the Congress comprises two phases: first, a discussion of fiscal policy issues, tax reform issues and budget ceilings in June/July, built around the submission of the MFMP; and, second, in August/September, an examination of the expenditure proposals in the draft budget by sector, entity, programs and investment projects. Government ministers, officials, the central bank, think tanks, and other organizations participate in these discussions. In terms of the key fiscal information that was disclosed, the recent PEFA Report gives the 2016 budget law a high rating against international standards.³⁰

Table 13.

Calendar for Submission and Approval of the Budget

Source: MHCP, PEFA Reports for 2009 and 2015, FAD staff.

¹These Committees of the Senate and the Chamber of Representatives cover economic policy and revenue issues (Committee #3) and the budget (Committee #4). They deliberate in joint sessions, but vote separately.

²The budget approved by the Congress comprises broad expenditure categories for each spending entity. During the “liquidation” process these ceilings are converted into detailed line item appropriations.

Table 13.

Calendar for Submission and Approval of the Budget

Deadline	Action Required
April, first week	Submission of a consolidated preliminary budget proposal to Congress
June 15	Submission of the MFMP to Congress
July 29	Submission of a draft budget law to Congress
August 15	The two economic committees1 of Congress decide whether the proposed budget complies with the organic budget law
September 25	The economic committees approve the Budget proposals
October 20	Congress approves the budget law
October 31	Budget Law is signed by the President and published in the Official Gazette
December 31	Completion of the “liquidation” procedure2 by the executive

Source: MHCP, PEFA Reports for 2009 and 2015, FAD staff.

¹These Committees of the Senate and the Chamber of Representatives cover economic policy and revenue issues (Committee #3) and the budget (Committee #4). They deliberate in joint sessions, but vote separately.

²The budget approved by the Congress comprises broad expenditure categories for each spending entity. During the “liquidation” process these ceilings are converted into detailed line item appropriations.

View Table

Table 13.

Calendar for Submission and Approval of the Budget

Deadline	Action Required
April, first week	Submission of a consolidated preliminary budget proposal to Congress
June 15	Submission of the MFMP to Congress
July 29	Submission of a draft budget law to Congress
August 15	The two economic committees1 of Congress decide whether the proposed budget complies with the organic budget law
September 25	The economic committees approve the Budget proposals
October 20	Congress approves the budget law
October 31	Budget Law is signed by the President and published in the Official Gazette
December 31	Completion of the “liquidation” procedure2 by the executive

Source: MHCP, PEFA Reports for 2009 and 2015, FAD staff.

¹These Committees of the Senate and the Chamber of Representatives cover economic policy and revenue issues (Committee #3) and the budget (Committee #4). They deliberate in joint sessions, but vote separately.

²The budget approved by the Congress comprises broad expenditure categories for each spending entity. During the “liquidation” process these ceilings are converted into detailed line item appropriations.

66. Procedures and the disclosure of information related to other aspects of fiscal policy and national development planning are also subject to strictly enforced legal deadlines. These requirements apply, for example, to the operations of and disclosures made by the MHCP, the DNP, CONFIS and CONPES on fiscal- and planning-related issues, as well as the MGMP (Section 2.1.3), and the MFMP, fiscal policy objectives and the fiscal rule (Section 2.3.1). They also define detailed operational procedures such as the issuance of budget circulars to spending ministries, and the setting and notification of spending ceilings, and the release of data to the public.

2.3.1 Fiscal Policy Objectives

Advanced

67. In 2003, the Fiscal Transparency and Responsibility Law improved the presentation of fiscal information and increased coordination among the different levels of government.³¹ This law (Law 819) required the national government and subnational governments (SNGs) to present each year a MFMP that represents the government’s main instrument of macro-fiscal programming. The MFMP must present a consistent 10-year macroeconomic framework which, in its initial formulation, comprised multiannual primary balance targets that were intended to guarantee fiscal sustainability in the medium term and long term. Controls on the spending and borrowing of SNGs were also substantially tightened in this period (Section 3.3.1). Finally, the notion of “fiscal sustainability” was added as an amendment to the Constitution in 2011.

68. To further strengthen the fiscal responsibility legislation and operate a sustainable and countercyclical fiscal policy,³² the government introduced a new fiscal rule in 2012. This rule comprises a structural deficit target of 1 percent of GDP from 2022 onwards with a transition period between 2012 and 2021. Intermediate targets of 2.3 percent of GDP in 2014 and 1.9 percent of GDP in 2018 were established. Additionally, the law requires that in each intervening year the structural deficit must be lower than in the previous year. The law includes an escape clause that permits deviations from the rule in exceptional circumstances, namely if output falls below its potential level, or oil prices fall below their long-term trend level, but the government has not yet made use of this provision. Since its establishment, the rule has been adhered to.

69. The law established an independent review process in the form of a Consultative Council. The Council has an independent membership comprising representatives of universities, think tanks, and the heads of the Economic Committees of the Congress (see section 2.4.1). It is responsible both for preparing the methodology for calculating the structural deficit, and measuring the government’s compliance with the rule.³³ The Council has some of the characteristics of the fiscal councils³⁴ that have been established in other countries, especially in Europe; it presents a report to the government in April setting out its projections of potential output and oil prices, and the structural deficit over a 15-year period. In June, the MHCP publishes, at the same time as the MFMP, a report on compliance with the fiscal rule, using calculations that have been agreed with the Council.

70. In addition to the MFMP, the MHCP also prepares an annual Fiscal Plan, releasing its preliminary version in December and publishing a revised version in March. This plan presents fiscal projections one year ahead, as well as an assessment of fiscal policy developments, and progress in implementing the fiscal rule. In preparing these documents, the MHCP holds regular consultations with counterparts in government and the central bank, as well as four meetings per year with the Congress, to discuss fiscal developments. In addition, the CGR publishes in August an annual report on the fiscal situation which includes an assessment of, and opinions on, the macroeconomic outlook and the government’s macroeconomic forecasts, fiscal policy developments and the fiscal rule, debt sustainability, and the quality of fiscal data.

71. While the fiscal rule has been adhered to since its introduction in 2012, looking ahead to 2022, the government may face challenges in ensuring that it remains on track. Fedesarrollo, a leading economic think tank in Colombia, has reported that the uncertainties surrounding the fiscal impact of the Peace Agreement could increase the deficit by about 0.5 percent of GDP.³⁵ The institute estimates that, on current policies, there is likely to be a fiscal gap of some 2.0-2.5 percent of GDP by 2022 which the next government will need to offset if it is not to breach the fiscal rule. The government also faces challenges in ensuring that modifications to healthcare and the pension system are made in the next few years that are fiscally sustainable (see Section 3.1.2).

2.3.2 Performance Information

Basic

72. A formal system of performance-based budgeting has not been developed in Colombia, though policymakers draw on a wide range of information in decision-making and setting spending priorities. Information on performance is most widely used in relation to capital investment projects. The national budget, however, does not incorporate a comprehensive programmatic classification, and information on the outputs or outcomes of public service delivery is limited. Similarly, budget execution reports contain little information on the impact and performance of public expenditure policies and projects. This is confirmed by the recent PEFA report which found no evidence that information on resources received by front-line service delivery units, such as schools or health clinics, was systematically collected by the ministries responsible for these services.³⁶ The only documents containing performance data are those prepared by the DNP to report on progress made by public institutions, through their investment projects, in achieving the objectives of the PND.

73. Some work is underway that may lead to improvements in performance information in coming years. The DNP, for example, has been preparing with the World Bank a results-oriented framework³⁷ for planning, monitoring, and evaluating investment projects on a systematic basis, but this framework has yet to be completed, tested, and implemented. In addition, the DNP has published performance assessments on the efficiency and effectiveness of service delivery; and the CGR has released management audits and sector policy analyses that focus particularly on the non-financial performance of investment projects.³⁸

2.3.3. Public Participation

Basic

74. A Citizens’ Guide to the Budget (Prepuesto Ciudano) is published after the budget is approved by Congress each year. The Guide is in an accessible form, centered around the use of graphics.³⁹ The 2017 Citizens’ Guide contained information on total revenues and expenditures for the budget year, macroeconomic assumptions, the sectoral allocation of spending, transfers to subnational governments, and an overview of the annual budget calendar. It also contained some data on per capita subnational transfers from the Sistema General de Participation across different geographic departments, and social spending by sector on particular groups of citizens, such as children and older adults. The Guide is published only in Spanish (not in indigenous languages), and the website is the only form of dissemination. The government has not consulted the public on the content of the Citizens’ Budget.

75. The government does not engage with or consult the public during its preparation of the annual budget. There is some opportunity for public input during the Congress’s consideration of the budget presented by the government, but not prior to that stage. The Open Budget Survey 2015 rated public participation in Colombia as weak during the executive stage of the budget, limited during the legislative stage, and adequate with respect to the CGR.⁴⁰

76. There are other examples of public outreach and engagement with respect to fiscal policy. The DNP conducts public consultations every four years in the preparation of the multiannual PND, which is an important determinant of investment spending in annual budgets. On the revenue side of the budget, Colombia achieved candidate status for the Extractive Industries Transparency Initiative (EITI) in 2014,⁴¹ while MHCP conducts public consultations on draft decrees within its field of responsibility, including proposed changes to tax administration.⁴²

The Ministry also maintains a Transparency Portal (the site received 89,062 visits from January 2017,⁴³ and publishes some fiscal data in open data format.⁴⁴

2.4.1 Independent Evaluation

Advanced

77. Although an independent fiscal council does not exist, the legal framework in Colombia provides for evaluation by multiple independent entities at different stages. These include the following:

• The independent Consultative Council, set up under the fiscal responsibility legislation (Law 1473 of 2011), defines the methodology and sets parameters for the operation of the fiscal rule (Section 2.3.1). The Committee comprises nine independent members, including the chairs of the two Congressional committees on economic and fiscal matters. It authorizes the MHCP’s report on compliance with the fiscal rule before its submission to Congress. The law requires the government to consult the Committee in case the fiscal rule is to be suspended.

• In accordance with the Constitutional requirements, the BanRep is invited by Congress to present its opinion on the government’s macroeconomic and fiscal forecasts and the MFMP during the budget deliberations.

• In discharging its “fiscal oversight” function mandated by the Constitution, the CGR in its “State of Public Finance” report provides an ex-post assessment of the government’s macroeconomic and fiscal forecasts, comments on developments in the external and internal macroeconomic environment, and analyzes in detail the government’s fiscal performance.

2.4.2. Supplementary Budgets

Advanced

78. Amendments to the approved budget are carried out in accordance with legal provisions and published on the MHCP’s website. Law 38 of 1989 and Decree 111 of 1996 regulate changes to the approved budget. In case it becomes necessary to increase the amounts of initially authorized appropriations, or to include a new item of expenditure in the budget, the law allows the government to place before Congress bills on reallocations and additions to the approved budget. As a general principle, only Congress can approve changes—additions or transfers across budget lines—to the approved appropriations contained in the annual budget law. In certain cases, the powers are delegated to the government, which are exercised through decrees and resolutions.⁴⁵ Reallocations across ministries or other public institutions, and across programs within a ministry or institution, are thus not permissible without Congressional approval. The law further demands that supplementary appropriations are obtained only after clearly establishing additional sources of funding to meet those expenditures, and on a certification by the CGN, or the head of budget in case of a public institution, of the availability of such resources. Budget laws, decrees, and resolutions authorizing revisions to the original budget are published on the MHCP’s website, and the CGR’s budget execution report consolidates all changes, comparing the outturn with both the original and revised budgets.

79. The government has largely maintained the credibility of the original budget. As shown in Table 14, changes in total budget appropriations have been relatively small in recent years. In each of the past three years, the budget was reduced because of a revenue shortfall. The magnitude of these changes, however, has been small.

Table 14.

Changes to Approved Budgets

(COP Trillions)

Source: MHCP.

Table 14.

Changes to Approved Budgets

(COP Trillions)

Year	Original Budget	Final Budget	Change
2016	216	210	-3%
2015	216	208	-4%
2014	203	197	-3%
2013	186	189	2%
2012	165	166	1%

Source: MHCP.

View Table

Table 14.

Changes to Approved Budgets

(COP Trillions)

Year	Original Budget	Final Budget	Change
2016	216	210	-3%
2015	216	208	-4%
2014	203	197	-3%
2013	186	189	2%
2012	165	166	1%

Source: MHCP.

2.4.3. Forecast Reconciliation

Not Met

80. The MFMP contains very limited information and discussion about the revisions to successive vintages of medium-term forecasts of revenue, expenditure, and financing.⁴⁶ Similarly, while the government’s update to the Financial Plan published each March presents and discusses new fiscal estimates for the past and current fiscal years, it does not provide any reconciliation of these projections, and does not quantify the impact of macroeconomic or new policy developments. The annual report published by the MHCP on compliance with the fiscal rule (Section 2.3.1) likewise presents a backward-looking view of the previous year’s budget, and does not discuss medium-term fiscal forecasts. The MHCP (DGPM) produces some analysis of forecast reconciliations for the ministry’s internal use. The inclusion of reconciliation tables in the MFMP—in which difference in successive vintages of fiscal forecasts are broken down into the relative impact of individual policy changes, macroeconomic determinants, and other factors, such as technical and accounting adjustments—would help improve the credibility of the government’s macroeconomic and fiscal forecasts.

3.1.1 Macroeconomic risks

Basic

86. Colombia is exposed to macroeconomic volatility, which in turn poses sizeable risks to budget. Even if, over the last decade, Colombia’s nominal GDP growth volatility was lower than in most neighboring countries, Colombia has experienced wide variations in its economic growth over that period, in part due to oil price volatility (Figure 13). Moreover, between 2011 and 2014 oil revenues contributed approximately 2.5 percent of GDP to annual central government revenues, against 1.6 percent of GDP between 2004 and 2010.⁴⁸

View Full Size

GDP Volatility

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

• Download Figure
• Download figure as PowerPoint slide

87. The MFMP presents sensitivity analysis of fiscal forecasts to macroeconomic assumptions but very limited discussion of alternative macroeconomic and fiscal scenarios. For example, the 2016 MFMP presents a sensitivity table summarizing the net impact on public revenues and spending for the next fiscal year of a change in the main macroeconomic parameters (exchange rate, oil price and production, inflation).⁴⁹ The MFMP also displays a limited alternative scenario analysis of the evolution of the central government debt over the next ten years, estimating the impact of a change in GDP forecasts by +/- 1 percentage point.⁵⁰ This analysis does not present the fiscal implications of these changes at a more disaggregated level nor examines closely the channels through which each scenario affects public finances.

88. The MFMP analysis could be expanded to include a more detailed presentation of the short- and medium-term fiscal outcomes associated with macroeconomic stress scenarios and stochastic analysis. The MFMP could discuss the channels through which the macroeconomic scenarios are expected to impact the main spending and revenue items in the budget. Further, the analysis could be disaggregated by level of government, given the weight of local and regional governments. Given the adoption of the fiscal rule and the setting of the path for the fiscal deficit, the analysis could also include (i) the impact of random shocks to fiscal outcomes based on their historical distribution and (ii) the impact of combined macroeconomic and specific fiscal risk shocks (see section 3.1.2 below).

3.1.2 Specific fiscal risks

Not Met

89. Specific fiscal risks are those that arise from narrow and identified sources of uncertainties. Specific fiscal risks include both explicit and implicit risks, from sources ranging from public debt management, guarantees, natural disasters, the financial sector, or from other public entities (subnational governments, and public corporations). This Indicator assesses the overall level of transparency with respect to a broad range of specific fiscal risks. Sections 2 and 3 of Pillar III consider in detail Colombia’s transparency and risk management practices with respect to main sources of specific fiscal risks.

90. Colombia regularly publishes detailed information on explicit contingent liabilities in the MFMP. The Fiscal Transparency and Responsibility Law (2003) requires that the MFMP contain an assessment and valuation of the main contingent liabilities and non-explicit debts. Since 2004, the MFMP has contained a detailed chapter on contingent liabilities, with a section also on pension and civil service severance exposures (non-explicit debt). Contingent liabilities covered include those in PPP contracts, public credit guarantees, legal action against the state, and callable capital in international financial institutions.

91. One of the major specific fiscal risks in Colombia stems from legal action against the State, and is covered in detail in Chapter 8 of the MFMP. Between 2000-2015 payments for judgements and reconciliations amounted to a cumulative COP8.4 trillion (in 2012 pesos), and have been increasing (Figure 14). Because of its significance—it is by far the largest explicit specific fiscal risk in Colombia—it has been the subject of considerable attention in recent years. Efforts to mitigate and manage the risks have included the creation in 2012 of the National Agency for Legal Defense of the State, which has developed a detailed database of claims and payments and is undertaking activities to reduce the risks. It is supporting the MHCP Risk Directorate in the development of improved risk quantification methodologies.⁵¹

View Full Size

Payments Under Legal Claims

(in COP millions)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: Colombian authorities.

• Download Figure
• Download figure as PowerPoint slide

92. The way in which the government’s contingent liabilities from legal action are disclosed could potentially prejudice its legal defense. Appropriations for expected payments are included in the budgets of 160 individual entities in the national budget. There is a concern that disclosing the expected cost of litigation against individual state entities could prejudice the state’s legal defense by revealing information with respect to what the government expects individual cases to cost. Other countries that disclose legal action as a contingent liability do so on a gross basis across central government (rather than entity by entity), and accompany the disclosure with a disclaimer that reporting the risk does not indicate any government acknowledgement of liability.

93. Overall, the MFMP disclosure on risks falls short of a summary of the main specific risks to the fiscal position. Table 16 presents an illustrative overview of selected specific fiscal risks in Colombia. The table separates explicit risks from implicit risks. Explicit risks are those that are legal obligations, the subject of firm commitments, or unavoidable (e.g. damage from disasters to government-owned property). The table also separates short- to medium-term risks, on the one hand, from longer-term challenges to fiscal sustainability e.g. pensions and health care. Important risks not at present covered in the MFMP include those from disasters, debt management, public corporations, subnational governments, and the private financial sector. The MHCP intends to progressively expand the coverage of risks in the MFMP, starting in the 2017 MFMP with discussion of risks from natural disasters.

Table 16.

Selected Fiscal Risks of the Central Government

Source: Colombian authorities and FAD staff estimates.

¹Data source: Table 8.14, 2016 MFMP. The figure in the notes to the financial statements (Table 3.95) for the contingent liability from legal claims is COP1,300 trillion (162% of GDP). However, this figure includes some very large claims that are considered to be of very low probability, which MHCP excludes from its gross exposure and expected cost calculations. There are minimal balances in the CFEE to meet legal claims.

²Source: MHCP Risk Directorate. PPP contingent liabilities includes concessions, APP projects, and projects without public contribution (APPIP-SRP). The expected costs for concessions and APP projects are paid into the FCEE. By law, realized APPIP-SRP project risks cannot be funded from the Contingency Fund, they are to be met from increased tolls or project restructuring.

³Obligations to provide additional share capital in International Financial Institutions if requested. Data from section 8.1.5, 2016 MHCP.

⁴From Section 8.2.2.1, 2016 MFMP

⁵The total value of central government buildings and structures on the government’s balance sheet, exposed to risk of damage in disasters. Law 42/1993 (article 107) requires every public building to be insured against liability for loss or damage to property; to the extent this covers disaster risks and is complied with, the net exposure would be significantly lower. There are also balances in the Disaster Fund (Fondo de Calamidades) to defray costs from disasters, the balance at December 2014 was COP52 billion.

⁶The government’s potential exposure to provide assistance, e.g., with respect to low income housing.

⁷As at end 2016, the gap between Fogafin’s reserves and total guaranteed deposits.

⁸Gross liabilities of public corporations from the balance sheet.

⁹The gross debt of subnational governments, from Report on Subnational Debt.

¹⁰The actuarial liability for the defined benefit pensions, from the balance sheet and notes to the financial statements.

¹¹2016 MFMP forecasts a transfer from the general budget to the health system of 0.25% of GDP per year over the next decade.

Table 16.

Selected Fiscal Risks of the Central Government

	Risk	Gross exposure COP trillion	Percent of GDP (2015)	Estimated expected cost COP trillion	Percent of GDP (2015)
Selected explicit fiscal risks
Contingent liabilities:
	Legal claims against the state1	96	12	51.3	6.4
	Debt guarantees	6.95	0.9	1.83	0.2
	Contingent obligations in PPP contracts2			15.3	1.9
	Callable capital3	20.9	2.6	3	0.4
Other explicit risks:
	Civil servant retroactive severance4	-	-	3	0.3
	Natural disasters5	83.9	10.5	-	-
Total selected explicit fiscal risks		207.75	26	74.4	9.2
Selected implicit fiscal risks
Disasters and other environmental risks6
Fogafin’s unfunded liabilities7		33	4.12
Public corporations8		228	28.5
Subnational governments 9		39	4.9
Total selected implicit fiscal risks		300	37.5
Memorandum item: risks to medium- to long-term fiscal sustainability
Pensions10				839	104.7
Health11				?	?

Source: Colombian authorities and FAD staff estimates.

¹Data source: Table 8.14, 2016 MFMP. The figure in the notes to the financial statements (Table 3.95) for the contingent liability from legal claims is COP1,300 trillion (162% of GDP). However, this figure includes some very large claims that are considered to be of very low probability, which MHCP excludes from its gross exposure and expected cost calculations. There are minimal balances in the CFEE to meet legal claims.

²Source: MHCP Risk Directorate. PPP contingent liabilities includes concessions, APP projects, and projects without public contribution (APPIP-SRP). The expected costs for concessions and APP projects are paid into the FCEE. By law, realized APPIP-SRP project risks cannot be funded from the Contingency Fund, they are to be met from increased tolls or project restructuring.

³Obligations to provide additional share capital in International Financial Institutions if requested. Data from section 8.1.5, 2016 MHCP.

⁴From Section 8.2.2.1, 2016 MFMP

⁵The total value of central government buildings and structures on the government’s balance sheet, exposed to risk of damage in disasters. Law 42/1993 (article 107) requires every public building to be insured against liability for loss or damage to property; to the extent this covers disaster risks and is complied with, the net exposure would be significantly lower. There are also balances in the Disaster Fund (Fondo de Calamidades) to defray costs from disasters, the balance at December 2014 was COP52 billion.

⁶The government’s potential exposure to provide assistance, e.g., with respect to low income housing.

⁷As at end 2016, the gap between Fogafin’s reserves and total guaranteed deposits.

⁸Gross liabilities of public corporations from the balance sheet.

⁹The gross debt of subnational governments, from Report on Subnational Debt.

¹⁰The actuarial liability for the defined benefit pensions, from the balance sheet and notes to the financial statements.

¹¹2016 MFMP forecasts a transfer from the general budget to the health system of 0.25% of GDP per year over the next decade.

View Table

Table 16.

Selected Fiscal Risks of the Central Government

	Risk	Gross exposure COP trillion	Percent of GDP (2015)	Estimated expected cost COP trillion	Percent of GDP (2015)
Selected explicit fiscal risks
Contingent liabilities:
	Legal claims against the state1	96	12	51.3	6.4
	Debt guarantees	6.95	0.9	1.83	0.2
	Contingent obligations in PPP contracts2			15.3	1.9
	Callable capital3	20.9	2.6	3	0.4
Other explicit risks:
	Civil servant retroactive severance4	-	-	3	0.3
	Natural disasters5	83.9	10.5	-	-
Total selected explicit fiscal risks		207.75	26	74.4	9.2
Selected implicit fiscal risks
Disasters and other environmental risks6
Fogafin’s unfunded liabilities7		33	4.12
Public corporations8		228	28.5
Subnational governments 9		39	4.9
Total selected implicit fiscal risks		300	37.5
Memorandum item: risks to medium- to long-term fiscal sustainability
Pensions10				839	104.7
Health11				?	?

Source: Colombian authorities and FAD staff estimates.

¹Data source: Table 8.14, 2016 MFMP. The figure in the notes to the financial statements (Table 3.95) for the contingent liability from legal claims is COP1,300 trillion (162% of GDP). However, this figure includes some very large claims that are considered to be of very low probability, which MHCP excludes from its gross exposure and expected cost calculations. There are minimal balances in the CFEE to meet legal claims.

²Source: MHCP Risk Directorate. PPP contingent liabilities includes concessions, APP projects, and projects without public contribution (APPIP-SRP). The expected costs for concessions and APP projects are paid into the FCEE. By law, realized APPIP-SRP project risks cannot be funded from the Contingency Fund, they are to be met from increased tolls or project restructuring.

³Obligations to provide additional share capital in International Financial Institutions if requested. Data from section 8.1.5, 2016 MHCP.

⁴From Section 8.2.2.1, 2016 MFMP

⁵The total value of central government buildings and structures on the government’s balance sheet, exposed to risk of damage in disasters. Law 42/1993 (article 107) requires every public building to be insured against liability for loss or damage to property; to the extent this covers disaster risks and is complied with, the net exposure would be significantly lower. There are also balances in the Disaster Fund (Fondo de Calamidades) to defray costs from disasters, the balance at December 2014 was COP52 billion.

⁶The government’s potential exposure to provide assistance, e.g., with respect to low income housing.

⁷As at end 2016, the gap between Fogafin’s reserves and total guaranteed deposits.

⁸Gross liabilities of public corporations from the balance sheet.

⁹The gross debt of subnational governments, from Report on Subnational Debt.

¹⁰The actuarial liability for the defined benefit pensions, from the balance sheet and notes to the financial statements.

¹¹2016 MFMP forecasts a transfer from the general budget to the health system of 0.25% of GDP per year over the next decade.

94. There are other significant short- to medium-term explicit risks to the fiscal position. These are areas where there is a greater than usual degree of uncertainty over revenue or expenditure forecasts. They include the peace process, the fiscal impacts of recent tax reforms, regional instability, and uncertainty over the amount of the annual budget transfer required to Colpensiones, given the difficulty of forecasting the numbers switching from private schemes to the public scheme.

95. In view of the range and magnitude of specific fiscal risks in Colombia, and the narrow fiscal space available, summary reporting of fiscal risks should be a priority for fiscal oversight and management. In the last 10-15 years, several countries have started publishing annual Fiscal Risk Statements, generally alongside their annual budgets presented to the legislature. Examples from comparable countries include Brazil, Indonesia, and the Philippines. Box 3 presents a suggested form of a Fiscal Risk Statement for Colombia. The statement would usefully begin with discussion of the government’s fiscal risk management strategy in the context of the medium-term fiscal strategy, recent progress in mitigating risks, and priority areas for further risk mitigation. To avoid any negative reactions when disclosing information on fiscal risks for the first time, care should be taken to include a clear statement of how the risk is being managed. The MHCP Risk Directorate could be tasked to compile the statement.

96. To further improve risk disclosure and management, the institutional arrangements could be further strengthened. Two specific actions could be considered: (i) the internal risk reporting within the MHCP could be strengthened by introducing an in-year (e.g. quarterly) risk report to be submitted to MHCP management; and (ii) a high-level inter-directorate coordination mechanism could be established, with the Risk Directorate providing the anchor. Risk management typically involves multiple units within (and outside) a ministry of finance, and a coordinated approach is required to improve its effectiveness.

3.1.3 Long-term sustainability of public finances

Basic

97. The MFMP presents an extensive study of the long-term sustainability of fiscal policy over a ten-year timespan. Along with the underlying macroeconomic assumptions, notably the oil price projections, this ten-year analysis encompasses forecasts of the main fiscal aggregates as well as a debt sustainability analysis. The discussion on public debt includes an analysis of sustainability under alternative growth scenarios.⁵² However, the analysis does not consider the possibility of a macroeconomic shock combined with the realization of specific fiscal risks, such as those stemming from natural disasters, litigation losses, or realization of other explicit or implicit contingent liabilities.

98. The 2016 MFMP includes pension and health spending projections over the next ten years. Regarding the pension system, recognized as a non-explicit liability, the MFMP presents the impact of past measures on the net present value of future liabilities of the pension scheme in Colombia.⁵³ It also discusses the evolution over ten years of the future costs of the pension scheme to the government budget; and annual budget transfers required to ensure the equilibrium of the pension system. In 2017, this transfer is expected to amount to 4 percent of GDP, and is forecast to decrease gradually to 3.5 percent of GDP in 2027 (Figure 15). Regarding the health system (SGSSS), the 2016 MFMP discusses the negative fiscal impact of Law 1571 / 2015 (which changed the coverage of reimbursed drugs and medical practices and increased the coverage of the population) and presents a probabilistic approach to its future equilibrium, depending on better efficiency and the ability to generate additional resources.

View Full Size

Forecast Transfers from the Budget to the Pension Scheme

(percent of GDP)

Citation: IMF Staff Country Reports 2018, 250; 10.5089/9781484372159.002.A001

Source: MFMP 2016.

• Download Figure
• Download figure as PowerPoint slide