A. Introduction

1. The structural tax reform approved in December 2016 aims to mobilize revenue to replace lost oil proceeds, while fostering employment and growth; in this chapter, we attempt to quantify the growth impact of key elements of the reform.

2. The recent literature provides economic models suitable to analyze the impact of the tax reforms in Colombia. BanRep’s Fisco (Rincón et al 2015), and IMF’s FSGM (2016) share key ingredients including a rich array of fiscal instruments and an important role for public capital to foster private sector competitiveness.

3. This papers applies the IMF’s FSGM model to analyze the impact that the proposed changes of direct and indirect taxes would have on GDP growth. In short, simulations show that: 1) the direct growth impact of the reform will start small and average about 0.3 percent of GDP between 2017–22, and 2) the contribution to growth comes from the reform’s boost to private investment and productivity. After providing an overview of the Colombia’s tax system, FSGM model and simulations, the chapter’s concluding remarks outline some areas in which the analysis could be expanded.

B. Brief Diagnosis of Colombia’s (pre-Reform) Tax System²

4. The importance of the structural tax reform approved in December 2016 has been highlighted in multiple recent studies. A key issue discussed in previous IMF consultations (2014, 2015, and 2016) is that without further revenue mobilization, public expenditure, including social and infrastructure expenditure, would have to decline significantly in order to comply with the medium-term deficit reduction mandated by the fiscal rule. Such expenditure restraint amid still large social and infrastructure needs in Colombia, is not desirable. OECD (2015) also argues for the need of an overall simplification of the tax system and strengthening of tax administration as its complexity contributes to widespread evasion. The final report of the ECTR (Expert Commission for Tax Reform) documents that the multi-tier income tax system is contributing to complexity, lack of progressivity and significant dispersion of (implicit) taxation across economic sectors.

5. The Colombian tax system has been subject to multiple reforms. Among the most recent ones, the 2012 structural tax reform aimed to improve labor formality and tax progressivity while being revenue neutral. It replaced part of payroll taxes with an earmarked corporate income tax (CREE) that would coexist with the traditional corporate income tax (CIT); and introduced a new tier in the personal income tax (PIT) akin to the US’s alternative minimum tax (AMT) denoted in Colombia as IMAN. The 2014 tax reform aimed to replaced lost oil revenue as the commodity supercycle started to wane. Its key measure was a temporary increase in CREE.

6. Colombia’s tax revenue is somewhat below the regional average, but with important compositional differences. Key features include:

• Total income tax revenue in Colombia is above the level in some of its peers, yet it comes almost exclusively from CIT. Due in part to the increase in CIT in the 2014 reform, CIT revenue exceeds Chile’s and Mexico’s. At the same time, the scheduled increase in CREE for the next few years would have placed Colombia’s combined CIT+CREE well above its peers (43 percent versus 27.2 percent average in LatAM). On the other hand, the relatively large exempt income bracket and deductions erodes the PIT’s tax base and hinders the progressivity of the overall tax system.

• VAT revenue suffers from weak productivity due in part a relatively large number of exempted goods. VAT revenue outperforms Mexico’s but significantly behind Chile’s and the OECD average. As reported by the OECD, the productivity of VAT in Colombia—VAT revenue ratio that measures VAT revenue vs what would theoretically be raised if VAT was applied at the standard rate to the entire potential tax base in a “pure” VAT regime and all revenue was collected as described in OECD (2015)—ranks among the lowest in the region. Further, Colombia’s mixed performance on VAT revenue is despite having the not-common feature of not crediting VAT paid on capital goods (collects about 0.6 percent of GDP).

• The complexity of the tax system represents a heavy burden to the tax authority and tax payers and contributes to tax evasion. The complexity of the system coupled with limited human and technological resources and a limited adoption of electronic invoicing constrain the impact of tax administration efforts from the tax authority (DIAN)—which nevertheless have obtained about 1 percent of GDP of additional revenue from tax administration over the last 3 years. Likely due in part to its complexity, there is a widespread problem of tax evasion in Colombia. Recent estimates (see OECD 2015) suggests that tax evasion in VAT and Income taxes combined represent up to 4 percent of GDP (which is about one fourth of the total tax revenue collected nowadays).

• Wide dispersion of effective tax rates. Extensive use of special treatment to some goods (e.g. in VAT) and sectors (e.g. hotels, construction, zona franca, etc.) have resulted in a wide dispersion of effective tax rates across sectors. Recent studies suggest taxation might be an important driver of resource misallocation in emerging markets (Fiscal Monitor April 2017).

Tax-to-GDP ratios, 2015

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Tax-to-GDP ratios, 2015

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

Tax-to-GDP ratios, 2015

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

Components of General Government Revenue (% GDP)

Colombia and comparator countries, 2015

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD Revenue Statistics.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Components of General Government Revenue (% GDP)

Colombia and comparator countries, 2015

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD Revenue Statistics.

• Download Figure
• Download figure as PowerPoint slide

Components of General Government Revenue (% GDP)

Colombia and comparator countries, 2015

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD Revenue Statistics.

• Download Figure
• Download figure as PowerPoint slide

Corporate tax rate comparison

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Corporate tax rate comparison

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Corporate tax rate comparison

(in percent of GDP)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

VAT revenue ratio (VRR) in LAC countries, 2014

(in percent; see text for definition)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

VAT revenue ratio (VRR) in LAC countries, 2014

(in percent; see text for definition)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

VAT revenue ratio (VRR) in LAC countries, 2014

(in percent; see text for definition)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

Source: OECD.

• Download Figure
• Download figure as PowerPoint slide

C. Key Elements of the 2016 Structural Tax Reform

7. The reform’s goal is twofold: secure enough revenue to protect key expenditure programs while complying with fiscal consolidation mandated by the fiscal rule as well as to foster employment and investment. To the first goal, the revenue aims to collect about 3 percent of GDP by 2022 which would allow to keep a broadly constant level of primary expenditure as a ratio of GDP. To achieve the second goal, the reform reduces the corporate tax burden including by lowering the CIT and allow to credit the VAT paid on capital goods against CIT.

8. The expected revenue yield will come from some key measures. An increase in the VAT general rate by 3 percentage points (from 16 to 19 percent) will generate about 1 percent of GDP; while anti-evasion measures and gains from formalization and growth will each generate about 0.6 percent of GDP. The reform also includes green taxes and some reductions in the exemptions for Personal Income Tax (PIT). It is important to note that the reform counts as net revenue the extension of the existing Financial Transaction Tax (FTT) that was scheduled to gradually disappear starting in 2019 (currently, the FTT generates about 0.8 percent of GDP).

Colombia: Key Elements of Structural Tax Reform

Source: National authorities and Fund staff estimates

^1/Includes revenue from taxation to non-profit entities; single tax for small taxpayers (monotributo)

Colombia: Key Elements of Structural Tax Reform

Expected revenue from selected measures (in percent of GDP)
		2017	2018	2019	2020	2021	2022
Total		0.7	0.7	1.7	2.5	2.8	3.1
VAT and consumption tax		0.7	0.8	0.8	0.9	0.9	0.9
	o/w deduct. VAT on capital goods.	0.0	−0.2	−0.2	−0.1	−0.1	−0.1
Green taxes		0.1	0.1	0.1	0.1	0.1	0.1
Personal income tax		0.0	0.1	0.1	0.1	0.1	0.1
Corporate income tax		−0.1	−0.3	−0.1	−0.1	−0.1	−0.1
Extension of Financial Transaction Tax		0.0	0.0	0.2	0.4	0.7	0.9
Formalization and Growth 1/		0.0	0.1	0.3	0.7	0.6	0.7
Anti-evasion measures		0.0	0.0	0.3	0.5	0.6	0.6
For simulation:
VAT excluding capital goods		0.7	1.0	1.0	1.0	1.0	1.0
CIT + VAT on capital goods		−0.1	−0.5	−0.2	−0.2	−0.2	−0.2

Source: National authorities and Fund staff estimates

^1/Includes revenue from taxation to non-profit entities; single tax for small taxpayers (monotributo)

View Table

Colombia: Key Elements of Structural Tax Reform

Expected revenue from selected measures (in percent of GDP)
		2017	2018	2019	2020	2021	2022
Total		0.7	0.7	1.7	2.5	2.8	3.1
VAT and consumption tax		0.7	0.8	0.8	0.9	0.9	0.9
	o/w deduct. VAT on capital goods.	0.0	−0.2	−0.2	−0.1	−0.1	−0.1
Green taxes		0.1	0.1	0.1	0.1	0.1	0.1
Personal income tax		0.0	0.1	0.1	0.1	0.1	0.1
Corporate income tax		−0.1	−0.3	−0.1	−0.1	−0.1	−0.1
Extension of Financial Transaction Tax		0.0	0.0	0.2	0.4	0.7	0.9
Formalization and Growth 1/		0.0	0.1	0.3	0.7	0.6	0.7
Anti-evasion measures		0.0	0.0	0.3	0.5	0.6	0.6
For simulation:
VAT excluding capital goods		0.7	1.0	1.0	1.0	1.0	1.0
CIT + VAT on capital goods		−0.1	−0.5	−0.2	−0.2	−0.2	−0.2

Source: National authorities and Fund staff estimates

^1/Includes revenue from taxation to non-profit entities; single tax for small taxpayers (monotributo)

9. The simulations to be presented next will focus on the impact of changes to VAT and CIT. In particular, the simulations will focus on the combined impact of the deductibility of VAT paid on capital goods together with the lower CIT. The next table describes the impact of the reform on corporate income taxation by the combined effect of CIT and CREE. The simulations will also include the impact of the increase in VAT so that to avoid double counting, the VAT revenue lost from allowing the deductibility of capital goods will be removed from such entry as described in the memo items in the table above.

Reform Changes to CIT (In Percent)

Source: National authorities

Reform Changes to CIT (In Percent)

		2017	2018	2019	2020	2021	2022
Combined No Reform		42	43	34	34	34	34
	CIT	34	34	34	34	34	34
	CREE	8	9	0	0	0	0
Combined Reform		40	37	33	33	33	33
	CIT	34	33	33	33	33	33
	CREE	6	4	0	0	0	0
Memo:
	Reform-No Reform	−2	−6	−1	−1	−1	−1

Source: National authorities

View Table

Reform Changes to CIT (In Percent)

		2017	2018	2019	2020	2021	2022
Combined No Reform		42	43	34	34	34	34
	CIT	34	34	34	34	34	34
	CREE	8	9	0	0	0	0
Combined Reform		40	37	33	33	33	33
	CIT	34	33	33	33	33	33
	CREE	6	4	0	0	0	0
Memo:
	Reform-No Reform	−2	−6	−1	−1	−1	−1

Source: National authorities

D. Evaluating the Growth Impact of the Reform: Literature Review and the FSGM Model

10. The literature on the growth impact of fiscal policy (fiscal multipliers) provide some useful benchmarks for the simulations in this paper. Batini et al (2014) offer a literature review and find that among the existing literature for emerging markets, the evidence suggests fiscal multipliers are smaller in emerging markets; and that tax multipliers and similar to expenditure multipliers (in the range of 0.1–0.4). The authors also identify structural factors that could lead to larger multipliers including, lower trade openness, more rigid labor markets and less flexible exchange rates.

11. A relevant benchmark of the estimates presented here is Rincon et al (2014) which estimates multipliers for Colombia using a dynamic stochastic general equilibrium (DSGE) model. The model (denoted FISCO) includes fiscal and monetary policy rules that endogenously react to shocks and policy changes to achieve medium-term targets. An important feature of the model is that public capital is modelled as an essential input in the production of some domestic goods. It finds a public investment multiplier is between 0.4 and 0.5. In contrast, public consumption multiplier depends heavily on the degree of response of monetary policy and of any other fiscal policy instrument used to achieve the fiscal rule targets and on average is about -0.02. Another useful reference is Vargas et al (2012) who employ a non-linear SVAR and estimate expenditure multipliers of about 0.4 and argue that data suggests multipliers have increased as the country’s overall fiscal position has improved since the 1990s.

12. The FSGM model shares some key properties of Rincon et al (2014) and provides a semi-structural framework to analyze policy changes.³ The Flexible System of Global Models (FSGM) is a semi-structural general equilibrium model that can be calibrated to specific countries. It incorporates a rules-based fiscal and monetary policy and a combination of OLG (Non-Ricardian) and liquidity constrained consumers (LIQ). Private consumption and investment are modelled structurally (micro-founded), while for other parts of the model (e.g. labor supply, price determination) a reduced-form is used.

13. Central to the analysis in this paper, the model poses that public capital (deviations from trend) positively affect total factor productivity and hence returns to private investment which is modelled through a Tobin’s Q model of quadratic adjustment costs. Intuitively, the model captures cases such as public investment in infrastructure that would allow better use of existing private resources. Similarly, to Rincon et al (2014), fiscal policy is modelled as the combination of different instruments including indirect and direct taxes as well as oil-related royalties and lump-sum transfers to households. The model includes a fiscal rule that ensures a stable debt to GDP ratio in the long-run, while monetary policy is guided by a Taylor rule.

E. Simulations

14. The simulations presented next aim to estimate the growth impact of key elements of the reform. Some remarks are in order. First, the simulations should be understood as the estimation of the joint impact of revenue and expenditure measures. The simulation includes the expected revenue yield for PIT, CIT and VAT (as described above) and also the fact that such revenue proceeds are devoted to public investment. Second, the estimation includes the endogenous response of monetary policy to the tax reform. That is, if the reform were to push GDP above its potential level, inflationary pressures would build and monetary policy will endogenously respond by hiking policy rates. Finally, the simulation depicts the additional growth versus a no-reform steady state, but not necessarily versus the latest observed growth rate.

15. To further illustrate the growth impact of public capital this section also includes a simulation of the impact of the 4G infrastructure agenda.⁴ To do so, the simulation includes the expected 4G capex between 2016–2021. Hence, the estimated growth yield is of the building phase of the agenda, and not necessarily of the additional GDP growth the agenda could generate once it is finished. At the same time, this way to model the 4G agenda is not without limitations as in practice the agenda is PPP-based and will be in practice accounted as private investment. By modelling it as public investment, the simulations include wider fiscal deficits. As the built-in fiscal rules aim to stabilize deficits, the results below will be somewhat biased by also incorporating some gradual fiscals tightening designed to restore deficits to steady state levels.

4G Infrastructure Agenda Capex (In Percent of GDP)

Source: National authorities and Fund staff estimates

4G Infrastructure Agenda Capex (In Percent of GDP)

	2016	2017	2018	2019	2020	2021
Capex	0.23	0.94	1.17	1.05	0.81	0.37

Source: National authorities and Fund staff estimates

View Table

4G Infrastructure Agenda Capex (In Percent of GDP)

	2016	2017	2018	2019	2020	2021
Capex	0.23	0.94	1.17	1.05	0.81	0.37

Source: National authorities and Fund staff estimates

• Growth impact. The simulations show the reform could boost growth by about 0.3–0.4 percent of GDP per year which builds up gradually as the reform total yield and hence the additional public investment start small (0.5–0.6 percent of GDP in 2017–18) and then increased slightly after (to 0.8 percent of GDP). The charts above also suggest, the reform would boost GDP above its trend level; in other words, the contribution of the reform to potential GDP (including the boost to TFP from additional public investment) will be milder than the impact on headline GDP.

• Private investment and consumption. The reform will change the composition of domestic demand with the increase in VAT leading to temporarily lower consumption and the combination of lower corporate taxation and additional public capital will boost private investment. After 2018, private consumption will recover and exceed a no-reform scenario boosted in part from additional labor income as firms will expand production and employment. The 4G agenda will reinforce the positive impact on investment and consumption.

• Public investment and total factor productivity. Both of these two factors will boost growth. The reform proceeds will allow additional public investment also strengthened by the 4G agenda. Additional public investment will lead to higher public capital and result in an increase cumulative increase in TFP levels.

• Inflation and monetary policy response. As the simulations suggest headline GDP will exceed potential GDP during the first years of the projection, inflation will tend to increase which will lead to an endogenous response of monetary policy. The simulations show monetary policy rates would be about 50bp and 125bp higher under a tax reform, and tax reform and 4g scenarios, respectively. Hence the growth impact discussed above should be understood as the join impact of fiscal and monetary policy in response to the reform.

GDP Growth

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

GDP Growth

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

GDP Growth

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Output Gap

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Output Gap

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Output Gap

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Consumption

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Real Consumption

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Consumption

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Investment (Private)

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Real Investment (Private)

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Investment (Private)

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Government Investment

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Real Government Investment

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Real Government Investment

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Trend total factor productivity

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Trend total factor productivity

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Trend total factor productivity

(% Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Core CPI Inflation

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Core CPI Inflation

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Core CPI Inflation

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Policy Interest Rate

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Policy Interest Rate

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

Policy Interest Rate

(%pt Difference)

Citation: IMF Staff Country Reports 2017, 139; 10.5089/9781484302187.002.A004

• Download Figure
• Download figure as PowerPoint slide

F. Concluding Remarks and Potential Extensions

16. The simulations presented in this chapter suggests the tax reform approved in 2016 stands to advance on the authorities’ objective to promote growth. The results are driven not only by the boost to private investment but also due to the additional public investment the reform will allow.

17. There are several dimensions where the above analysis could be refined and expanded.

• Elements of the reform. As mentioned above the simulations in this paper cover only some key elements of the reform. It could be useful to also model the impact of the extension of the FTT and green taxes. The former might impact growth through the credit channel while the latter could generate a distinction across industrial sector depending on their reliance on fossil fuels.

• Effect on risk premia. Something not considered in the analysis in this paper is the impact that the reform could have on Colombia’s risk premia and consequently on the cost of funding for firms and consumers. As some rating agencies had conditioned Colombia’s rating on the approval of the tax reform, the simulations could include a reduction on risk premia as part of the reform scenario. The model could likely generate higher private investment in that case.

• Quality/effectiveness of public capital. Given that the estimated growth impact is due in part due to formation of public capital, it could be useful to incorporate as part of the model’s parameters the effectiveness through which resources devoted to public investment become public capital.