Colombia: Public Finances – any cause for concern?

Colombia’s public finances have come under the spotlight in recent years amidst recurrent adverse external shocks, rising social spending pressures, ongoing challenges in raising revenues, persistent (optimistic) biases in fiscal planning and, as of late, the back loading of fiscal consolidation plans following the Covid-19 shock. The rapid progression of the public debt ratio and the capacity for future policy adjustment have, in particular, become points of concern and have, since the summer 2021, materialized in Colombia losing its investment grade status. However, overly focusing one’s attention on debt levels, debt dynamics or the speed of fiscal adjustment to assess fiscal sustainability in Colombia can lead to overlook important risk-mitigating aspects of the sovereign’s credit profile.

Despite facing a challenging scenario of its underlying debt drivers, the sovereign maintains a solid capacity to support debt backstopped by a favorable interest-to-growth differential, low contingent liabilities, a manageable debt-servicing burden and a sound institutional framework. Looking forward, engaging the broadest swath of society in shaping fiscal policy represents a significant challenge that could – if done inclusively – pay important dividends in terms of both economic and fiscal outcomes.

In the summer of 2021, Colombia de facto lost its “investment grade” status – which it acquired back in 2011.[1] Colombia had faced recurrent pressures on its sovereign rating (chart 1), in the wake of the reversal in the commodity super-cycle in 2014-15. The latter spurred a rapid decline in oil prices – a staple which accounts for close to 40% of exports, 1/3 of FDI and some 10% of total fiscal revenues. The shock led to a sizeable deceleration of real GDP growth (chart 2), a cumulative loss of 5.5% of GDP in oil revenues over the period 2015-18 [2] and as a result heightened pressure to increase non-oil revenues through tax reform.

A combination of factors precipitated Colombia’s fall out of investment grade, according to the rating agencies: a deterioration in key fiscal metrics aggravated by the oil and pandemic shocks, an unfavourable socio-political environment to pass fiscal reforms, limited visibility of the post-pandemic adjustment process, in particular concerns over ability to cut down deficits and stabilize debt (the general government debt ratio has close to double since 2012 reaching 65% of GDP in 2020 (chart 3).

The depreciation trend of the peso since the oil price shock (chart 4) has also made debt stabilization hard to achieve as a little over 1/3 of public debt is denominated in foreign-currency, one of the highest levels amongst Latin America’s largest economies (chart 5). The government’s decision to maintain headline fiscal deficits in the range of [7-8.6%] over period 2020-2022 contributed to further put public finances in the spotlight as Colombia was only one of the few emerging markets that decided to backload its fiscal adjustment plans to 2023. Perhaps more importantly, the country faced a tipping point in May 2021 when a controversial fiscal reform proposal by the government led to a 62-day national strike marked by violent protests and road blockades. The social unrest which resulted in 84 casualties (civilians and police) and cost more than USD 3 bn according to the Finance Ministry, led to the withdrawal of the reform proposal, precipitated the departure of two ministers (Finance and Foreign Affairs) and led to a spike in Covid-19 cases exacerbating the country’s 3^rd epidemic wave.

There has been no shortage of debate related to how the pandemic shock should be treated in credit models or credit assessment frameworks. Just like the Global Financial Crisis previously, the Covid-19 epidemic raised important questions concerning the appropriate fiscal response, its size, composition and duration. In any case, it became apparent early on that most sovereigns would be dealing with sizeable debt increases and would have to engage (at some point) in some form 1 S&P (May 2021) and Fitch (July 2021) downgraded Colombia from BBB- to BB+. Moody’s kept its rating unchanged (at Baa2 since 2014) during its July review. However, two investment grade ratings from the three main rating agencies are required to maintain an overall investment grade status. 2 IMF (2021). Article IV: Colombia of policy adjustment to consolidate fiscal accounts, while having to compose with a more pressing set of challenges (lower potential growth, accelerated digitalization of their economies, energy transition).

Considering the warnings relayed by many international organizations [3] relative to the risks of adjusting too quickly, a case could be made to temporarily move away from strictly focusing on government debt burdens and benchmarking exercises[4] when evaluating the trajectory of sovereign ratings. However, just as in 2007-8 previously, attention in sovereign creditworthiness assessments has already turned to the speed of consolidation – the latter often being equated with fiscal tightening due to the implicit assumption "[…] that fiscal tightening is the key test of a government’s determination to honour its debts, and is therefore necessary for a quick return of investor confidence and a rapid pickup in growth [5]".

Is Colombia’s ability and willingness to pay being adversely impacted by its decision to delay its fiscal adjustment? Is Colombia’s intrinsic credit quality worse today or comparable to that of 2007-2011 – the last time the country durably found itself in the BB+ category? Is the downgrade into speculative grade harbinger of worse things to come or a temporary backslide? Does the accumulation of public debt incurred by the pandemic response and ensuing fiscal expansion pose a risk to debt sustainability?

The government’s decision to maintain headline fiscal deficits in the range of [7-8.6%] over period 2020-2022 contributed to further putting public finances in the spotlight as Colombia was only one of the few emerging markets that decided to backload its fiscal adjustment plans to 2023. Perhaps more importantly, the country faced a tipping point in May 2021 when a controversial fiscal reform proposal by the government led to a 62-day national strike marked by violent protests and road blockades. The social unrest which resulted in 84 casualties (civilians and police) and cost more than USD 3 bn according to the Finance Ministry, led to the withdrawal of the reform proposal, precipitated the departure of two ministers (Finance and Foreign Affairs) and led to a spike in Covid-19 cases exacerbating the country’s 3^rd epidemic wave.

There has been no shortage of debate related to how the pandemic shock should be treated in credit models or credit assessment frameworks. Just like the Global Financial Crisis (GFC) previously, the Covid-19 epidemic raised important questions concerning the appropriate fiscal response, its size, composition and duration. In any case, it became apparent early on that most sovereigns (a) would be dealing with sizeable debt increases and (b) would have to engage (at some point) in some form of policy adjustment to consolidate fiscal accounts while (c) having to compose with a more pressing set of challenges (lower potential growth, accelerated digitalization of their economies, energy transition).

Considering the warnings relayed by many international organizations[1] relative to the risks of adjusting too quickly, a case could be made to temporarily move away from government debt burdens and benchmarking exercises[2] when evaluating the trajectory of sovereign ratings. However, just as in 2007-8 previously, attention in sovereign creditworthiness assessments has already turned to the speed of consolidation – the latter often being equated with fiscal tightening due to the implicit assumption […] that fiscal tightening is the key test of a government’s determination to honour its debts, and is therefore necessary for a quick return of investor confidence and a rapid pickup in growth.[3] Is Colombia’s ability and willingness to pay being adversely impacted by its decision to delay its fiscal adjustment? Is Colombia’s intrinsic credit quality worse today or comparable to that of 2007-2011 – the last time the country durably found itself in the BB+ category? Is the downgrade into speculative grade harbinger of worse things to come or a temporary backslide? Does the accumulation of public debt incurred by the pandemic response and ensuing fiscal expansion pose a risk to debt sustainability?

In an effort to address these questions, section "Pre-Covid-19 public finances snapshot: the pros and cons" looks at the state of public finances prior to the Covid-19 health crisis – outlining the main elements, which back then, supported or constrained Colombia’s public creditworthiness. Section "Covid-19 shock: what is the damage to public finances?" provides an overview of the fiscal response to the health crisis and its impact on fiscal metrics in an effort to assess the damage done to public finances by Covid-19. Section "Post-covid shock recovery: possible trajectory of debt metrics" projects the possible paths of the public debt ratio after estimating the trajectory of its underlying drivers. In light of the insights formulated, the conclusion attempts to answer the central question of this article, that is whether there are any cause for concern about the current trajectory of Colombia’s public finances.

Pre-Covid-19 public finances snapshot : the pros and cons

A good starting point to evaluate Colombia public finances is to first look at some of (A) the improvements in terms of public financial management as well as the (B) enduring challenges encountered by successive governments ahead of the health crisis.

[1] Many international institutions including the IMF, the World Bank, ECLAC have warned about how premature fiscal tightening in some cases could be self-defeating if it ends up weakening the recovery process, reducing growth and investment all of which ultimately can end up hurting the prospects of fiscal consolidation and reduce fiscal space.

[2] Whereby the performance metrics of a sovereign (in particular GDP per capita, fiscal deficit, debt to GDP and economic growth volatility) are evaluated against that of a typical sovereign in a given rating category (captured by the median). These types of benchmarking exercises are often used to justify rating actions (a change in outlook or an outright rating change).

[3] Unctad (2011). Trade and development report.

Pre-Covid-19 public finances snapshot: the pros and cons

A good starting point to evaluate Colombia public finances is to first look at some of the improvements in terms of public financial management as well as the enduring challenges encountered by successive governments ahead of the health crisis.

Sound institutional, regulatory and policy frameworks

Colombia has bolstered its fiscal policy frameworks and governance in significant ways over the past 25 years or so (box 1). In particular, the introduction of a fiscal rule with precise structural deficit6 targets and mechanisms to control spending and revenues has been particularly foundational in instilling policy discipline and helping to improve coordination between monetary and fiscal policy. There have also been important qualitative improvements in the form of advancing gender parity in the public sector and strongly improving data availability and transparency7 .

Improved fiscal management helped save a large portion of the commodity windfall induced by large increases in terms of trade over period 2004-20128 , as well as built up saving funds to cover against the materialization of contingent liabilities (estimated at some 2% of GDP).

Advances on the fiscal front have mirrored other policy improvements, regulatory and institutional developments:

• Democratic institutions have experienced an uninterrupted functioning over the past 50 years or so9 . Unlike many peer countries in Latin America, there have not been repeated coup attempts in Colombia since 1958 – following the end of the military regime of General Rojas Pinilla (1953-1958). Continuity of democratic governance has been shown to be an important long-term determinant of prosperity10.
• • Colombia has a central institution in charge of long-term planning – the National Planning Department (NPD) created in 195811 – helping to link public policies with an overall development vision. The NPD helps formulate national development plans (NDP) whereby long-term targets and objectives are laid out while medium-term priorities and goals are identified in the realm of socio-economic and environmental policies. The NDP is used to monitor the government’s performance against said targets and objectives.
• Several improvements in terms of policymaking are the product of structural reforms from the 1990’s following the adoption of a new constitution in 1991: independence of the Central Bank (1991), creation of a string of development banks, export promotion and investment agencies (1989-1991), adoption of a flexible exchange rate (1999), waving of exchange controls12, adoption of inflation targeting regime (1999), privatization of state enterprises, tariffs reduction, waving of import restrictions, relaxation of foreign direct investment (FDI) rules, strengthening of financial supervision and regulation following the banking crisis of 1998-199913. As part of the reforms of the Central Bank (BanRep), it was determined that monetary authorities would not be permitted to monetize the public debt (ie BanRep cannot buy sovereign debt instruments in the primary market). • Colombia’s accession to the OECD (in 2018 but formalized in 2020) is a testimony to its efforts towards structural reforms and improved policy management. Adhesion to the organization implies in particular to take steps in improving governance (anticorruption, better access to information strengthening the rule of law, and protection of human rights).
• This prudent policy framework helped underpin financial and macroeconomic stability—both important stepping-stones in the development of the local sovereign debt market (chart 6). Indeed, compared to some peers in the region, Colombia has managed over the past two decades to post solid growth performances, control inflation (chart 7), avoid debt restructuring (unlike Argentina, Uruguay, Ecuador)

Structural public debt management reforms have also helped reduce the vulnerability of public accounts to financial shocks.

Active debt management initiatives have in particular helped to :

• Progressively alter the composition of public debt, reducing dependence on debt denominated in foreign currency to cover financing needs (chart 8). The greater reliance on local debt instrument has been fostered by the development of the local sovereign bond market (3rd largest in Latin America after Brazil and Mexico). The foreign-currency debt portfolio has hovered around 35% of the public debt stock.

• Diversify the sovereign’s investors’ base. There are no dominant holders of the public debt (chart 9). However, non-residents still account for some 35% of holders in the local public debt market (25% if one only counts bond debt).

• Improve coordination with monetary authorities and lower levels of government. Debt managers meet with the Central Bank to build projections on current and future liquidity needs and actively manage subnational debt (unlike in many other countries, national debt managers are actively involved in the management of debt issued at lower levels of government.)

• Build up the risk management framework in order to better account for the impact of contingent liabilities and to the risk as tail risks and provide a backstop for external payments in foreign currency of both public and private counterparties. FX reserves reached USD 58.9 bn, some 13.8 months of import cover versus 6 months in 2008 (chart 12). The accumulation of reserves has helped bring about a fall in the net external public debt to GDP ratio.

The more challenging part: upholding fiscal credibility

Even if authorities have continued to show a commitment to the fiscal rule, they have faced more challenges in recent years in upholding fiscal targets and credibly communicating a path on how to reach future fiscal targets. There are three main sets of factors either cyclical or structural, that help to explain these difficulties:

• Strong and long-standing opposition in Congress and the population to comprehensive tax reform. Despite well-known shortcomings of the tax code (limited redistributive impact, special regimes, high marginal tax rates15), opposition to reforming the current tax system tends to be strong. This has forced successive governments to adopt a piecemeal approach to tax reform. As a result, Colombia has had close to 20 tax reforms since 1990 including some important ones over the past decade (appendix 1). Tax reform proposals have had a tendency to be watered down in Congress and rarely deliver larger gains than 1% of GDP. Historically, the avenue to increase tax revenues has primarily taken place through frequent changes to the VAT rate and coverage16. Increases in VAT rate (currently at 19%) however tend to be controversial because of their adverse impact on the poor – who spend the largest portion of their income on consumer goods and services. VAT has therefore been subject to many exemptions (see below). Another important imbalance stands from the disproportionate reliance on corporate income tax as a source of tax revenues. Corporates in Colombia face the highest effective tax rate in the OECD at some 60%. The corporate income taxes along with the VAT / goods and services tax have been the main contributors to total tax revenues contributing respectively 24% and 30% of total tax revenues in 2019 – relatively high shares in comparison with the OECD (10% and 20%). Meanwhile, personal income tax only represented 6% of total tax revenues in 2019 versus 24% on average for the OECD.
• High levels of informality in the labour market and high thresholds for paying income tax have impeded an expansion of the tax base (chart 15). A significant portion of the labour market operates outside formal channels (62% of the labour force is estimated to work in the informal sector)17. The high level of informality is in large part explained by elevated startup costs for formal businesses and high labour costs driven by a comparatively high minimum wage18. The persistently high level of informality generates an estimated 1.4% of GDP in foregone government revenue every year. Another reason for the low tax base stems from the high threshold to pay income tax such that less than 10% of workers pay personal income tax (chart 16). Over time, there has been little political consensus on raising the taxable base on individual income.

• On top of tax evasion and loopholes, distortions in the tax system have reduced potential revenues by at least one third according to Colombia’s revenue services DIAN. i/ Many items are exempt from VAT with exemptions accounting for an estimated 7% of GDP19) ii/ Tax benefits to encourage investment in priority industries iii/ Imbalances in tax collection between sectors and company size (preferential special regimes for businesses). The ad-hoc expansion of tax exemptions (including VAT free days) are often deployed to help ease popular tensions and cope with political constraints following increases in nominal tax rates20. According to the Finance Ministry some 40% of potential VAT tax collection and 30% of income tax collection is subject to evasion.
• Loss of short term revenues from reforms: for instance, i/ the 2012 tax reform led to an estimated loss of 1.2 percentage points of GDP in revenues by suppressing a financial transaction and capital gains tax (0.8% of GDP) and wealth tax (0.4% of GDP). ii/ The 2018 fiscal reform has also been highly criticized for having led to permanent revenue losses of 0.8% of GDP as a result of corporate tax cuts21. The original 2018 Financing Law was however meant to be revenue-neutral in the medium term with a widening of the VAT tax base. However, that aspect of the reform never made it through Congress.

Optimistic biases in fiscal planning have contributed to dilute policy credibility about future consolidation paths [22]:

• Assumptions about economic growth. With some rare exceptions, economic growth has been systematically projected to grow above potential growth since 2012 in the government’s medium-term fiscal framework26. According to government estimates a 1-percentage point drop in real GDP growth costs the government around 0.4% of GDP in revenues.

Given the limited compressibility of current spending, the additional spending needs generated by external shocks and the underperformance of revenues (due to lower oil related revenues, underperforming GDP growth or unrealized efficiency gains), the authorities have been at times forced to cut subsidies, infrastructure spending but most importantly rely on extraordinary (one-off) sources of revenues to meet revised fiscal targets27. This has included i/ sale of assets ii/ divestment from assets (selling some of its stakes in utility company ISA and Ecopetrol). But, because these types of non-recurrent revenues do not offer a structural solution to achieving fiscal consolidation, they have tended to be perceived negatively in rating agencies’ credit assessments.

Covid-19 shock: what is the damage to public finances?

Initial fiscal conditions

Colombia entered the Covid-19 crisis with a weaker fiscal position, less policy scope and a weaker growth track record compared to the 2007-8 Global Financial Crisis (GFC). When the pandemic hit, the economy was starting to gain pace (3.3% growth in 2019) after years of sluggish growth following the 2014- 15 oil shock (the 5-year average growth rate stood at 2.4% in 2019 vs. 5.5% on average in the years preceding 2007). The economy faced a less supportive external environment (crude oil prices stood above USD 100/barrel in Sept 2007 vs. approximately USD 48 USD in February 2020). The average size of the fiscal deficit was comparable over the 5 years preceding the shocks: 3.6% of GDP (2003-2007) vs. 3.1% (2015-2019). On the flipside, however, the public debt ratio decreased by some 12 points of GDP in the 5 years preceding the GFC reaching 32% of GDP in 2007. By contrast, public debt continued to rise steadily between 2012-2019 gaining some 14 percentage points of GDP over the period, despite budget deficit plans being formulated in accordance with the fiscal rule28. This can be largely explained by “automatic debt dynamics” (appendix 4). Indeed, according to calculations by the IMF, "6.5 points of the increase was driven by cyclical components (oil and output) while another 6.5 percent of GDP was driven by a sharp depreciation of the exchange rate in response to the drop in oil prices"[29]. The public debt trajectory however masks the fact that in the interim period, the fiscal institutional framework strengthened, debt became more affordable, the country improved its capacity to withstand external shocks and more importantly, a peace agreement with the FARC was signed.

Impact on fiscal metrics

The health crisis affected fiscal balances and public debt through multiple channels :

• Reduced revenues (via collapse of economic activity, tax relief measures, drop in earnings from commodity exports),
• Automatic stabilizers (unemployment allowances),
• Social spending
• Currency depreciation (given that about a third of the debt is denominated in foreign currency),
• Steep fall in nominal GDP (the collapse of demand was particularly steep in Colombia compared to other regional peers [34],
• Lag in vaccination campaign (need to extend support to the economy over longer periods of time compared to developed markets).

In order to meet higher than usual financing needs (~10% of GDP) over the period 2020-22, the authorities are relying on 5 main sources of financing:

• The saving and stabilization fund (FAE, 1.2% of GDP)
• The national pension fund for local governments (FONCET),
• Issuance in the local bond market in the range of USD 13 bn to USD 16 bn on average per year (it asked banks to buy USD 2.5 bn worth of solidarity bonds in 2020),
• External financing (in the range of USD 10-11 bn per year – typically from CAF, IBRD, IBD, including some emergency financing from the IMF),
• The government also resorted to privatizations (USD 3.5 bn in 2021 and an expected USD 1.7bn in 2022) and introduced a temporary “solidarity tax” [35].

In 2020, the central government debt ratio increased by 13 percentage points to 59% of GDP (+13pp also for the general government debt and +15 pp for the non-financial public sector debt to 71% of GDP). As such, the debt burden increased almost by the same amount in 2020 then over the preceding 8 years with the debt to GDP ratio close to doubling since the instauration of the fiscal rule in 2011. The increase in the debt burden was larger than the Latin American average (+10pp), in line with the global average (+13pp) but smaller than the increase in advanced economies (+16 pp) [36]. The extension of guarantees, like in many other countries also increased contingent liabilities to the tune of 2.6% of GDP [37].

Post-covid shock recovery: possible trajectory of debt metrics

The growth rate of a country’s debt to GDP ratio depends essentially on the future paths of the primary balance, interest rates, the exchange rate and growth prospects (appendix 4). But in large part the debt to GDP ratio responds to the interaction of interest rates and output growth:

• If the economy’s growth rate is higher than the average interest rate on public debt, authorities may stabilize or even reduce their public debt to GDP ratio even if they run a primary deficit. Thereby "low rates of interest can allow for a fiscal expansion on a sustainable basis" [38].
• If the average interest rate on public debt is higher than the rate of growth of the economy then the authorities will have to run a primary surplus (which size can be determined arithmetically) to stabilize the public debt ratio otherwise debt will automatically grow over time.

In the first subsection, we start by evaluating the drivers of the public debt ratio in light of the social context in the country, political developments in recent years, and the effects of Covid-19. This will help us get a sense of their likely future direction. In the second subsection, we use the insights from the previous subsection to come up with macro-assumptions that will serve as inputs to generate a baseline scenario for the public debt path. The baseline scenario and shocks to the baseline scenario can serve as a basis of comparison against official projections (embodied in the government’s mediumterm fiscal framework).

Evaluating the drivers of the public debt ratio in a post-Covid-19 era:

Driver n°1 – Primary balance: in deficit for at least 5 years. In the same way that the 1991 constitution led to a sharp rise in spending [39], the government will have to compose with important fiscal pressures in coming years which will likely keep the primary balance in deficit through the medium-term.

The government opted to continue supporting the economy and vulnerable population through the end of 2022 in the forms of health spending, household transfers, support to firms and investment (table 1).

Pressures to increase social spending will persist:

For some observers, the absence of policy measures destined to counter the distributional impact of the pandemic on the most vulnerable is likely to further feed social unrest ultimately fueling a vicious cycle (lower output, increases inequality which triggers social unrest which leads to lower output etc.) [44]. Avoiding this trap should provide enough incentive for policymakers to keep support schemes in place for longer, and spend more on healthcare and education (currently in lower tier of countries in the region (chart 25)).

Other types of increasing spending pressures are likely to emanate from :

• costs of integrating migrants (as migration pressures are likely to subsist given the deepening of the socio-economic crisis in Venezuela and the effects of climate change),
• implementation costs of 2016 peace accords (there will be continued pressure both by local stakeholders as well as the international community for a better implementation of the negotiated agreement.),
• icontrol of narco-trafficking and combating organized crime (one of the many consequences of Covid-19 has been the increase in drug use, and organized crime, higher levels of poverty and disparities has also favoured recruitment into crime cells [45]),
• energy transition costs (as part of the COP26, Colombia has committed to reducing emissions by 50% by 2030),
• age-related spending pressures (the country’s demographic profile will require important changes to the pension and healthcare systems in the coming years).

Driver n°2 - Interest rates: the way is up There are multiple arguments militating in favour of public finances being affected by rising interest rates going forward:

• Rising inflation is driving up the short end of the yield curve

During the pandemic, the Central Bank of Colombia (BanRep) cut rates to unprecedented levels. However, inflation has been creeping back up (chart 26) in many emerging markets on the back of high commodity prices, supply chain disruptions, lagged effects of FX depreciation, consumption bounce back [46]. In Colombia in addition to these factors, a nationwide protest in May-June 2021 led to an increase in food prices as the unrest affected the provision of basic goods. Indexations practices and a sizable real increase in the minimal wage also contributed to increased inflation. So far, the BanRep has been one of the last large Central Banks in Latin America to start normalizing its policy stance (chart 27) but all points to faster tightening in the coming year both for domestic and external reasons (i.e. monetary policy normalization in advanced economies may affect yields, capital flows and debt pricing for EM)

Following the Covid-19 shock, the economy will need to adjust to a new macroeconomic equilibrium which will drive up the real policy interest rates. One consequence of the Covid-19 shock has been the temporary expansion of the twin deficit (especially in 2021 with a current account deficit at -5.7% of GDP). Adjusting to the accumulation of these macroeconomic imbalances will either require fiscal policy to be less expansionary (i.e. faster fiscal consolidation) or require a faster adjustment of monetary policy. The latter’s stance is still expansionary (real policy rates stood at approximately -3% in February much below the neutral (equilibrium) real interest rates estimated at 1.5% [47]). Given the current fiscal stance and outlook as well as the likely convergence of private savings to their pre-pandemic historical average (17% of GDP vs. current 20% of GDP), the BanRep may need to raise policy rate closer to equilibrium faster if it wants to close external imbalances [48].

Driver n°3 – Nominal exchange rate: some room for appreciation in the medium-term

Following, the Covid-19 crisis however both the nominal and real effective exchange rates have shown greater levels of decoupling with oil prices. Domestic political risks (social protests, changes in the political landscape), higher market perception of sovereign risk (widening CDS spreads), deterioration of external imbalances and slower tightening of monetary policy relative to regional peers are explanations put forward to account for the greater divergence.

Fundamental changes to the BanRep’s current exchange rate policy are not anticipated in the context of our simulations. Historically, since Colombia shifted from a crawling band (from 1960’s on) to a flexible exchange rate regime in 1999, the BanRep’s interventions in the FX market (discretionary or scheduled) have been motivated by three interventions were compatible with its mandate (control inflation and preserve financial stability.) However, it has not done so with an explicit target for the nominal or real exchange rate and we do not expect that this policy will change over the forecast period in the medium-term fiscal framework (approximately 10 years). For the simulations, we assume an appreciating trend for the nominal exchange rate [50] as most fair-value models mobilized to estimate the long-run equilibrium exchange rate – whether they are based on prices (purchasing power parity or PPP) or fundamentals (behavioural equilibrium exchange rate or BEER) – show evidence that the Colombian peso has been persistently undervalued in recent years.

Driver n°4 - Potential growth: weakened by the Covid-19 shock

Simulations: Upside risks on the debt ratio

In this subsection, we evaluate alternative assumptions to those envisaged in the medium-term fiscal framework. We translate information from the previous section into numerical form to gauge the possible evolution of public debt dynamics.

The benchmark: the MTFF scenario

The government has made some important revisions to its medium-term fiscal framework (MTFF) in June 2021 with fiscal deficits for the general government projected to be significantly larger going forward than in previous MTFFs. Part of the increases stem from the fact that deficit targets appeared unattainable in light of the pandemic55 but the plan goes much further. Important takeaways are summarized in box 3, table 2, charts 32 & 33. According to the MTFF, public debt ratio would peak at close to 70% of GDP in 2022-23 and bring the debt burden to about 60% of GDP by 2032. However in order to fully comply with the fiscal rule, the government projects that in addition to fiscal revenues of 1.2% of GDP internalized in the 2021 tax reform (which would only kick in 2023), it would need an additional fiscal adjustment amounting to some 0.6% of GDP per year over the period 2023-26 – with a smaller adjustment thereafter. This implies that a new package of reform will be needed in the short term.

Baseline and alternative scenarios

• In light of the elements discussed previously, we present in the notes of table 3 our baseline macro assumptions for each drivers of the public debt ratio. Note that we internalized the release of new data since the publication of the MTFF in June 2021. Real and nominal GDP in 2021 were for instance much stronger than anticipated in the MTFF while the exchange rate depreciation was relatively contained. We also evaluate alternative paths for the debt ratio considering adverse shocks to the baseline scenario:
• Primary balance shock: under this scenario, consistently larger primary deficits are projected assuming less supportive oil prices, higher social spending needs, sustained growth underperformance or a watering down of future tax reforms. The baseline primary balance scenario is shocked by an additional 0.8% of GDP over the forecast period (representing about half a standard deviation of the primary balance over the period 2010-20).
• Interest rate shock: we also consider higher inflation dynamics and tighter external financial conditions over the forecast period. As such we internalize a 100 basis points (bps) shock to the baseline local interest rate scenario. Baseline foreign borrowing costs are also shocked by a 50 bps reflecting tighter external financial conditions (equivalent in each case to approximately one standard deviation considering 10-year historical rates both local and foreign).
• Growth shock: the economy is assumed to grow below potential (estimated at 2.7%) for 2 years starting in 2024-2025 with baseline growth rates lowered by one standard deviation (3.5 percentage points over period 2010-2020). This implies a mild recession in both years.
• Exchange rate shock: we introduce two depreciative episodes to our otherwise optimistic assumptions for the USDCOP path (relative to official projections). A severe nominal depreciation of the exchange rate analogous in size to the one experienced in 2015 (36%) is introduced in 2025. A milder episode is introduced in 2028 corresponding to one standard deviation over the period 2010- 2020.

Despite using more conservative assumptions (to tame the optimistic bias alluded to earlier in the text) on real GDP growth and the primary balance (internalizing higher spending pressures), we see in table 3 that under the baseline scenario the debt to GDP ratio would peak at close to 67% of GDP in 2023-2024 – below official projections. This more favourable path is in large part due to the internalization of the strong economic rebound witnessed in 2021, which helped stabilize the debt trajectory much earlier than anticipated, improving the fiscal outlook. In the latter part of the projection horizon, the baseline assumptions weigh more heavily on the debt dynamics with the debt trajectory dropping less quickly than official projections.

Conclusion: any cause for concern then?

Our analysis[56] showed that achieving official medium-term fiscal consolidation goals and meeting fiscal targets could be challenging as Colombia will have a narrower space for fiscal consolidation in the next 5 years.

Given that diagnostic, how concerned should we then be about the trajectory of public finances in Colombia?

The situation is far from being critical and Colombia has a good ability to support its debt. Colombia faces a stable interest payments burden, highly manageable debt servicing costs and a relatively favourable interest-growth differential. In addition, Colombia may offer greater upside risks compared to some of its regional peers in keeping rates low and keeping higher growth potential (section 3). Admittedly, there has been a big increase in the public debt ratio and it will take a long time return to pre-pandemic levels but active debt management and the build-up of the country’s external asset position have helped reduce interest rate, currency and rollover risks. The medium-term outlook is also supported by low contingent liabilities (which are well identified and provisioned for) as well as a solid institutional framework (cf. section 1) and safeguards (no monetization of the public debt) should help to keep spending and the debt at sustainable levels. The effect of volatile commodity prices on public accounts is also partially alleviated through the fiscal rule which is likely to be further enhanced going forward.[57]

However, there are three main vectors of risk, which if left unattended, could heighten concerns over the trajectory of public finances in the medium to longer term:

(i) If no efforts are made to rebuild policy credibility by revising future fiscal targets on the basis of more realistic macroeconomic assumptions ;

(ii) If future administrations are unable to generate new permanent sources of revenues. A risk of fiscal slippage could materialize if revenues and growth disappoint or if future governments do not adhere to the current administration’s fiscal plans – a possible scenario in light of emerging shifts in the political landscape (cf section 3) ;

(iii) If little effort is made to mend the social expectation gap.

Without addressing these challenges, Colombia may find it difficult to appease credit rating agencies and bolster investors’ confidence. This would likely compromise Colombia’s ability to maintain market access at favourable terms.

In order to enhance its chances of fulfilling the MTFF aspirations of fostering a virtuous cycle between the social, the economic and the fiscal, Colombia’s first order of business should be to enhance social dialogue to improve the existing social pact. Economists at the IMF remarked that amidst increasing poverty and defiance towards governing institutions, the countries that are likely to do better post-pandemic (in terms of both growth outcomes and passing reforms) are the ones most capable of building “broad social consensus and political cohesion around several crucial dimensions of public finances [through] a fiscal pact”.[58] This proposed roadmap would, in the case of Colombia, help address all three aforementioned vectors of risk. It stresses the need to engage in a social dialogue on how to broaden the social safety net (ie mending the social expectation gap) and how to finance it (generate new permanent sources of revenues), to assess society’s preferences regarding the tax and expenditure implications behind these very consequential tradeoffs. This public dialogue should serve as a basis into the legislative process that should take place in the next couple of years to revise pensions, health, and educational systems as well as reforming tax frameworks to support it. The authors also stress the importance of improving the effectiveness and flexibility of fiscal responsibility frameworks (to increase credibility and reduce fiscal risk): the formal adoption of fiscal anchors, introduction or strengthening of fiscal councils, and enhancement of communication strategies will improve efficiency and reduce fiscal risk, thus increasing credibility, lowering interest rates, and widening much needed fiscal space. In that regard, changes to the fiscal rule in the most recent tax reform is one step in the right direction. A more conservative calibration of the medium-term plan and the creation of an independent fiscal council (as recommended by the OECD) tasked with costing and forecasting the impact of fiscal policy measures over medium to longer time horizons could also help bolster fiscal policy credibility.

[1] S&P (May 2021) and Fitch (July 2021) downgraded Colombia from BBB- to BB+. Moody’s kept its rating unchanged (at Baa2 since 2014) during its July review. However, two investment grade ratings from the three main rating agencies are required to maintain an overall investment grade status.

[2] IMF, article IV, 2021

[3] Many international institutions including the IMF, the World Bank, ECLAC have warned about how premature fiscal tightening in some cases could be self-defeating if it ends up weakening the recovery process, reducing growth and investment all of which ultimately can end up hurting the prospects of fiscal consolidation and reduce fiscal space.

[4] Whereby the performance metrics of a sovereign (in particular GDP per capita, fiscal deficit, debt to GDP and economic growth volatility) are evaluated against that of a typical sovereign in a given rating category (captured by the median). These types of benchmarking exercises are often used to justify rating actions (a change in outlook or an outright rating change).

[5] Unctad (2011). Trade and development report.

[6] The structural fiscal balance = the headline fiscal deficit without business-cycle effects (cyclically adjusted) + adjusting for other temporary factors beyond the cycle (ie disaster-related relief, commodity shock that temporarily increases or reduces natural resources revenues). It allows to disentangle the permanent and temporary influence on the budget balance in order to gauge the medium-term orientation of fiscal policy (cf Hagemann, M. R. P. (1999). The Structural Budget Balance. The IMF’s Methodology. International Monetary Fund.

[7] Colombia persistently registers the highest score in the region on the OECD’s OURdata index – an index measuring government data availability, data usability, the extent to which data are available in open, free and accessible formats. OECD (2020). Government at a Glance. Country report: Colombia.

[8] IMF (2014). Colombia Article IV. Note that “the economy saved about 90% of the commodity windfall, the highest share in the region.”

[9]Vergne, C. (2015). Colombie: l’enjeu des réformes structurelles et du processus de paix. Agence française de développement. (AFD)

[10] European Parliament Research Service (2021). Democratic institutions and prosperity: the benefits of open society. Briefing, re-thinking democracy.

[11] OECD, (2019). Production Transformation Policy Review of Colombia: Unleashing productivity. OECD Development Centre. The latest national development plan presented by President Duque in 2019 had an endowment of USD 325 bn with focuses on education, employment entrepreneurship and the environment.

[12] Exchange controls had been in place since the 1930’s. However, some restrictions were maintained. Colombia maintained an exchange restriction arising from the special regime for the hydrocarbon sector under which branches of foreign companies in the sector have to give up their exports proceeds or agree to government-imposed limits on their access to foreign exchange.

[13] Colombia set up a Financial Stability Committee to coordinate the action of its three main supervisory bodies (Ministry of Finance, the Financial Superintendence of Colombia (FSC) which supervises financial institutions and the Central Bank). The FSC carries its supervisory functions with no political interference

[14] According to estimates by the IMF, fiscal costs associated with the migration flows amount to some 0.5% of GDP in 2019.

[15] OECD survey (2013).

[16] Lozano, I., 2001. Colombia's public finance in the 1990s: a decade of reforms, fiscal imbalance, and debt. Subgerencia de Estudios Económicos, Banco de la República.

[17] Informality defined workers who do not contribute to social security. It is however down from 70% in 2007.

[18] The high minimum wage and high non-wage labour costs are important cost factors found to reduce formal employment. OECD (2019) notes : “ At 87% of the median wage of full-time formal employees, the minimum wage is higher in relative terms than in any other OECD country”. For a review of drivers of informality and barriers to formalization, cf OECD. Eco Survey (2019).

[19] Klein Felipe (2021). Colombia : a cloudy fiscal horizon. BNP- Paribas, Global Market. March 2021.

[20] Salazar, N. (2013). Political economy of tax reforms: The case of Colombia. Woodrow Wilson Center Update on the Americas. Washington, DC, United States: Wilson Center, Latin America Program

[21] Note that in 2021, many of the corporate tax cuts initially voted in 2018 were overturned at least temporarily (cf. appendix 1)

[22] According to Hudson, B., Hunter, D., & Peckham, S. (2019). Policy failure and the policy-implementation gap: can policy support programs help? Policy design and practice, 2(1), 1-14. Overly optimistic expectations represent one of the vagaries explaining the gap between policy formation and implementation.

[23] The central government primarily depends on two sources of oil-related revenues: dividends paid by Ecopetrol and taxes paid by oil companies. Royalties on the other hand benefit regional governments. Export tax receipts are also an indirect form of revenue since more than 40% of exports are oil-based. See EIFI reports for breakdown for source revenues emanating from the extractive industry.

[24] More conservative approaches to determine structural oil prices (for example, by subtracting one standard deviation of oil estimates from the estimates of long-term prices) would have led to 3.5 percent of GDP less debt accumulation. Different estimates of the output gap would also have had an impact on debt accumulation but less so. (IMF, article IV – 2021)

[25] IMF (2021), Colombia : Article IV. For instance by making greater use of electronic invoicing.

[26] Felipe Klein – blog post BNPP Global Markets - According to a report from the IDB, except for 2013, growth turned to be well below potential growth projections in all MTFF between 2012 and 2018.

[27] According to Fitch, extraordinary revenues were used to meet revised central government deficit targets in 2017 (-3.6% of GDP), 2018 (-3.1% of GDP) and 2019 (-2.5% of GDP)

[28]IMF, article IV, 2021.

[29]IMF, article IV, 2021.

[30] Mauricio Cardenas, Luca Antonio Rocco, Jorge Roldos, and Alejandro Werner (2021). Fiscal Policy Challenges for Latin America during the Next Stages of the Pandemic: The Need for a Fiscal Pact by. IMF Working Paper WP/21//77, March 2021.

[31] According to the safeguard clause, when the output gap is negative and the expected real growth rate of production is at least 2 pp lower than the long-term rate (estimated between 4.3% and 4.8% by the government), a countercyclical spending program can be initiated. These countercyclical spending measures should be gradually dismantled once the economic growth rate has returned to its long-term level or above for 2 years.

[31] For an empirical discussion of the impact of fiscal and monetary measures (cf. IMF, article IV 2021).

[32] According to calculations by the IIF, a 10% increase in commodity prices adds ~0.25% of GDP in fiscal revenues. IIF(2021), The Fiscal Challenge, Latam view (august 2021).

[33] World Bank (2021). “Recovering growth: Rebuilding Dynamic Post-COVID-19 Economies Amidst Fiscal Constraints”. Semi-annual report on Latin America and the Caribbean region – October 2021.

[34] For 4 months from mid-April 2020 public sector workers with salaries between COP 10 million and COP 15 million per month (ie between USD 2,500 and USD 3,700) contributed 10% of their salary to finance COVID-19 related measures, while salaries above 15 million pesos contributed 15 percent. People who made less than COP10 million per month could make voluntary contributions.

[35] Kose, M. A., Ohnsorge, F., & Sugawara, N. (2021). A mountain of debt: Navigating the legacy of the pandemic. Policy Research working paper,no. WPS 9800,COVID-19 (Coronavirus) Washington, D.C. : World Bank Group

[36] According to July 2021 figures in fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic.

[37] Unctad (2011)

[38] Salazar, N. (2013). The 1991 constitution granted not only fundamental human rights but also economic and social guarantees within its social rule of law framework, as well as creating mechanisms to ensure and protect those rights. These guarantees have resulted in an expanded public sector and a significant growth in public spending […] Increased spending linked to the new constitution was coupled with other expenditure pressures, particularly those related to the financial imbalance in the pension system and increased resource demands from the defence sector.

[39]ECLAC (2021). Panorama Social de América Latina 2020.

[40] An estimated 5 million jobs were temporarily affected by the shock with most jobs being lost in the informal sector and unemployment hitting close to 20% at the height of the pandemic. As in many other LA5 countries (Brazil, Mexico, Peru, Argentina) females, the youth and low-educated workers endured the brunt of the adjustment in the labour market but Colombia, compared to the other LA5 had the specificity of enduring a larger rise in unemployment relative to the drop in economic activity. See Silva, Joana, Liliana D. Sousa, Truman G. Packard, and Raymond Robertson (2021). Employment in Crisis. World Bank Group, Book - Open Knowledge Repository. The OECD anticipates that pre-pandemic employment levels will only be regained by mid-2022. OECD (2021). Colombia Economic Outlook. December 2021.

[41] There has been greater income and job losses amongst poorer households, with lower levels of education that are self-employed or in informal jobs – thereby further worsening inequality. Cf IMF, Selected issues, 2021. Also the expansion of transfer programs helped but did not fully offset income losses, as a majority of those who reported losing jobs due to COVID-19 were not enrolled in any government transfer program.

[42]Vergne (2015)

[43]Sedik, T. S., Xu, R., & Stuart, A. (2020). A Vicious Cycle: How Pandemics Lead to Economic Despair and Social Unrest. IMF Working Papers, 2020(216).

[44] UN report + article americas quarterly.

[45] IIF (2021), the inflation challenge, Latam views ( July 2021)

[46] Sebatian Nietto Perra (2021)

[47] The neutral real interest rate (or natural rate) is the rate expected to prevail when an economy is at full employment and inflation is stable (ie the economy operates at potential). At the natural rate, monetary policy is neither expansionary nor contractionary. The natural rate cannot be observed, its level and trend can only be estimated. (cf Brookings). Note that the BanRep has used various methodologies to estimate the natural rate with estimates ranging from 1.1% to 2% with an average at 1.5% (cf inflation report, September 2018). The IMF also estimates the neutral real rate between 1% and 2% (cf article IV, 2021).

[48] Felipe Klein (2021). Colombia: External deficits strike again. Global Markets – EM Economics, Focus Latin America. Indeed from a saving-investment perspective, the current account deficit is equal to the fiscal balance plus a private saving-investment gap. To avoid large misalignments of the current account deficit relative to equilibrium (consistent with fundamentals), policy decisions have to affect the supply and demand for savings which ultimately are a function of the real interest rate.

[49] Moreno J.F. and Rojas J.S (2015). Recent performance of the exchange rate in Colombia. Monetary policy report, BanRep.

[50] Vargas, H. (2011). Monetary policy and the exchange rate in Colombia. Borradores de Economía, 655. FX interventions are made with a view to controlling inflation (the main mandate of the Central Bank) and preserving financial stability.

[51] Vergne (2015)

[52] Cf World Bank Global Economic Prospect (June 2020) for a review of channels through which Covid-19 affected potential growth worldwide.

[53] IMF, Colombia : selected issues, 2020. OECD (2019). Venezuelan migration shock in Colombia and its fiscal implications. OECD Policy Note.

[54] There are multiple ways of estimating potential (trend) output. For a review of methodologies on the issue see i/ Leonardo Bonilla Mejía, José David Pulido (2020). New Estimates for Colombia’s Potential (Trend) GDP and Output Gap. Box 1 In BanRep Monetary Policy Report (January 2020) and ii/ IMF, Colombia selected issues, 2021

[55] The fiscal deficit for 2022 was originally at 2.5% of GDP. It is now projected at 7% of GDP. According to calculations by the IIF, assuming that a tax reform yielded 1% of fiscal savings in 2022, in order to achieve a fiscal deficit of 2.5% of GDP (the original target of the medium-term fiscal plan for 2022), primary spending would have to be cut by almost 6%, a difficult result to achieve given the pressures to prop up social spending. IIF (2021). Fiscal risk in Colombia. Economic views

[56] Considering i/ the government’ decision to backload its fiscal adjustment, ii/ the prolonged impact of Covid-19 variants on public accounts (section 2), iii/ the historical challenges in raising tax revenues (section 1 and 3), iv/ the downside risks to projected oil related revenues, v/ the optimistic bias of government assumptions regarding GDP growth and efficiency gains (section 1), vi/ increased spending pressures (section 3) in a post-Covid era (healthcare, social spending) as well as other fiscal pressures (migration, Peace, etc..),

[57] In its 2021 article IV, the IMF provides a number of recommendations based on international experience to further strengthen the fiscal rule and the latest 2021 tax reform introduced a reform of the existing fiscal rule (cf appendix 1).

[58] Mauricio Cardenas, Luca Antonio Rocco, Jorge Roldos, and Alejandro Werner (2021). Fiscal Policy Challenges for Latin America During the Next Stages of the Pandemic: The Need for a Fiscal Pact by. IMF Working Paper WP/21//77, March 2021.