Abstract
Selected Issues
Impact of Public Investment on Growth in Uruguay1
A. Introduction
1. Uruguay’s infrastructure stock is among the highest in the region, but it has barely grown for over the last three decades.2 After a public investment push in the late 1970s that boosted the country’s infrastructure stock substantially, public investment has dropped to around 4 percent of GDP per year for the past three decades— fluctuating at around 2–2.5 percent of GDP in recent years. Although the investment level is not low by regional comparison, a higher base level of infrastructure stock means more resources are needed for simply maintaining the existing stock. As a result, the quantity and quality of infrastructure have not improved, or have even declined, over the past decades. In particular, compared to both the high-growth peer countries and main trade rivals, Uruguay lags behind in the quality of transport infrastructure.
Public Capital Stock
(As percent of GDP, 2015)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Source: IMF Fiscal Affairs Department.Public Investment
(As percent of GDP)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Source: IMF Fiscal Affairs Department.2. Investment in infrastructure is expected to increase, as various public-private partnerships (PPPs) start to be implemented. These include, for example, the US$800 million central railway project—which will accompany the investment in a new paper pulp plant by the Finnish forestry company UPM—several projects in road transportation, and projects to improve education infrastructure. The total infrastructure investments over 2019–23 is projected to be around US$2.3 billion (around 4 percent of 2018 GDP).
Public Investment
(As share of total investment)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: IMF Fiscal Affairs Department.Index of Public Capital
(1980=100)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: IMF Fiscal Affairs Department.Projected Infrastructure Investment Amount
(2019–2023, in USD mn)
Projected Infrastructure Investment Amount
(2019–2023, in USD mn)
| Year Amount |
2019 304 |
2020 800.7 |
2021 724.7 |
2022 420.7 |
2023 19 |
Projected Infrastructure Investment Amount
(2019–2023, in USD mn)
| Year Amount |
2019 304 |
2020 800.7 |
2021 724.7 |
2022 420.7 |
2023 19 |
3. This paper examines the impact of the planned infrastructure investments on Uruguay’s real GDP. Three methods are used to estimate the direct and indirect impact of the planned infrastructure investments on GDP growth and level: the direct value-added approach the fiscal multiplier approach, and the general equilibrium approach.
B. Direct Value-Added Approach
4. The direct impact of the planned infrastructure investments can be calculated as their contribution to the aggregate investment, minus their contribution to imports. As a starting off point, we can calculate the direct impact of these investments on GDP by adding the projected investment amounts to each projection year’s aggregate investments, while adding the imported components (e.g., imported machinery and materials) of the investments to the aggregate imports. The chart below shows the direct impact on GDP under the alternative assumptions of 30 percent of imported components and 70 percent of imported components. The results show that for 2020, the year of the highest infrastructure investment amount, the impact on GDP ranges from 0.4 to close to 1 percent of GDP. And since this approach does not assume any indirect, lagged, or knock-on effect on the rest of the economy, the projected impact goes back to zero after the investment projects are completed in 2023.
Impact of Infrastructure Investment on GDP
(Direct value-added approach, percentage difference from baseline GDP)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: BCU and IMF staff calculations.C. Fiscal Multiplier Approach
5. Fiscal multipliers for various government spending categories have been extensively studied in the literature. As the PPP infrastructure investments are ultimately government investment on infrastructure, the multiplier literature would shed light on their potential impact on economic activity. Taking the literature as a whole, one may conclude that the size of such multipliers varies greatly across country and spending category, and it is influenced by some country-specific conditions. This section reviews the relevant literature, identifies a possible range of public investment multipliers for Uruguay, and then applies the multipliers to the projected infrastructure investment data to estimate their potential impact on growth.
Public and Private Capital Stocks of Uruguay
(In bn USD, 2010 price)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: BCU and IMF staff calculations.6. The results from various research papers are considered for their application to Uruguay. The econometric estimates of the public investment multipliers for Uruguay are less reliable because the public investment has not much moved much in the past three decades, and there are not sufficient public investment announcements to identify exogenous public investment shocks. While there is a wide literature on fiscal multipliers, the paper paid attention to studies that 1) focuses on the multiplier of government investment, and 2) focuses on either emerging market countries similar to Uruguay or the Latin America region. The appendix presents the summary findings of seven research papers fitting these criteria.
7. The size of the multiplier depends on a range of factors, and ranges between 0.1–0.5 percent for the short term and 0.4–1.5 percent for the long term. The results laid out in the appendix table reveal the following patterns. First, the public investment multiplier is generally higher in advanced economies than in emerging and developing economies. This is likely due to the higher investment efficiency, as well as the stronger interconnections among different sectors, in advanced countries. In addition, studies show that the investment multiplier tends to be larger during recession and in countries with higher public investment efficiency. If debt levels are low, multipliers are also larger when the investment is debt-financed rather than budget neutral, reflecting demand effects of fiscal stimulus. If, however, debt levels are already high, multipliers could be lower or even negative under extreme circumstances. Some studies find that the size of the multiplier is negatively correlated with a country’s trade openness and public debt level. Finally, the investment multipliers in Latin American economies are found to be smaller compared to those in other parts of the world.3
8. Compared to the baseline, the planned infrastructure projects are estimated to increase real GDP level for the 2019–24 period. Three sets of multiplier estimations are selected from the literature and applied to the planned public investment series in Uruguay for the next few years. Specifically, the multiplier estimations are sourced from 1) IMF’s World Economic Outlook (2014), 2) IMF’s Regional Economic Outlook for Latin America (2018), and 3) Furceri & Li (2017). The results are dubbed as scenarios 1–3, respectively, in the chart below. Two of the three scenarios suggest that the GDP growth with infrastructure investment will be higher compared to the baseline, throughout the projection period.4 By the end of 2024, the real GDP level under the scenarios with infrastructure investment is 1 to 1.8 percent higher than the baseline.
Impact of Infrastructure Investment on GDP
(Multiplier approach, percentage difference from baseline GDP)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: BCU and IMF staff calculations.D. General Equilibrium Approach5
9. The IMF’s Global Integrated Monetary and Fiscal Model (GIMF) is calibrated for Uruguay (Box 1).6 The calibration is based on the recent values of basic macroeconomic variables. Regarding the share of liquidity constrained households in the economy, simulations are run for both 50 percent—close to the estimated share for Uruguay—and 25 percent—the average share for advanced economies.
Summary of GIMF Properties
GIMF is a multi-country Dynamic Stochastic General Equilibrium (DSGE) model with optimizing behavior by households and firms, and full intertemporal stock-flow accounting. In the model, frictions in the form of sticky prices and wages, real adjustment costs, liquidity-constrained households, along with finite-planning horizons of households, help to capture the important role for monetary and fiscal policy in economic stabilization.
The non-Ricardian features of the model provide non-neutrality in both spending-based and revenue-based fiscal measures, which makes the model particularly suitable to analyze fiscal policy questions. In particular, fiscal policy can stimulate the level of economic activity in the short run, but sustained government deficits crowd out private investment and net foreign assets in the long run.
Firms employ capital and labor to produce tradable and nontradable intermediate goods. There is a financial sector a la Bernanke, Gertler and Gilchrist (1999), that incorporates a procyclical financial accelerator, with the cost of external finance facing firms rising with their indebtedness.
GIMF is multi-region, encompassing the entire world economy, explicitly modeling all the bilateral trade flows and their relative prices for each region, including exchange rates. The version used in this paper comprises 6 regions: The United States, the euro area, Uruguay, Latin America excluding Uruguay, emerging Asia (including China), and, as a single entity, the remaining countries.
10. Simulation results suggest a significant impact of planned infrastructure investments on Uruguay GDP. These results should be interpreted as the upper bound because: i) simulations assume that all investments (roads, rail roads and education projects) are equally productivity enhancing; ii) there are no efficiency losses; iii) the model has strong financial accelerator properties (through equity and lending channels), amplifying the short-term demand effects, which are not present in Uruguayan economy; and finally iii) to the extent that a part of the investment is done by foreign companies the wealth effects will be smaller than implied by the model. Key contributing factors affecting the simulation results in the short-term are the share of liquidity constrained households which magnifies the impact of short-term investment. Infrastructure investment is assumed to be debt-financed and paid back over time through a small fiscal adjustment either by raising consumption taxes (VAT) or cutting transfers. Finally, the infrastructure investment would increase the level of public capital stock. But maintaining the stock at a high level requires maintenance costs. If the maintenance costs are minimal (0.08 percent of GDP) public capital depreciates through time—consistent with Uruguay’s past experience.
11. Specifically, four scenarios are simulated. In scenario 1, it is assumed that the government will finance its availability payment obligations to its private sector partners for these projects by raising consumption taxes (which could be in the form of reducing exemptions). Scenario 2 assumes that the availability payments will be financed through reducing public transfers. Both scenarios assume that around 50 percent of the households in the country are liquidity-constrained. Scenario 3 assumes that the share of liquidity-constrained households is around 25 percent, while the availability payments are financed through raising consumption tax. In scenario 4, it is assumed that maintenance spending after the implementation of investment projects is 2/3 lower than the level required to keep the public capital stock constant, while in all other scenarios, the level of maintenance spending is assumed to cover the capital depreciation entirely.
12. The short-term implications on GDP level are similar, mainly affected by the path of planned investment and its indirect demand effects. By 2024, the GDP level is projected to be roughly 1½–1¾ percent higher than the baseline (i.e. the scenario without the planned public investments).
13. The real GDP level is permanently higher, but its magnitude in the long-term will depend on whether the public investments are maintained. If the public capital stock is maintained, in the long-term real GDP could be 2 percent higher than in the baseline. If, however, public capital stock declines over time as a result of insufficient investment spending on maintenance, the impact on real GDP will be much smaller, at around 1 percent. Using consumption tax for fiscal adjustment is more distortionary compared to using transfers, and therefore the impact on real GDP are marginally lower in the former scenario.
Impact of Infrastructure Investment on GDP Level
(General equilibrium approach, percentage difference from baseline GDP)
Citation: IMF Staff Country Reports 2020, 052; 10.5089/9781513530864.002.A002
Sources: BCU and IMF staff calculations.E. Conclusion
14. The level of infrastructure investment in Uruguay is projected to increase in the next few years. Although public investment spending level is not low compared to regional peers, the public capital stock in Uruguay has barely increased over the past decades, as maintaining the large initial stock would have required higher maintenance investment. The planned infrastructure projects are expected to close the infrastructure gaps, both in term of quality and quantity, and therefore help increase productivity, reduce export costs, boosts economic catchup of poorer regions in the country, and therefore raise the level of real GDP.
15. These infrastructure investments are expected to have a positive impact on real GDP level, though the magnitude of the impact depends on many factors. While GDP growth rates will be higher when the infrastructure investment takes place, the growth impact will decline with the completion of projects. But the level of real GDP is expected to be permanently higher compared to the baseline reflecting productivity effects. This paper assesses the direct and indirect GDP level effects from the planned infrastructure investments by using three approaches: i) direct value-added approach, ii) multipliers; and iii) simulation of a general equilibrium model. While all three approaches suggest a positive effect on growth, the estimated magnitude of impact varies widely. The short-term effects are highly dependent on the size and presence of the indirect demand channels. Assuming some degree of indirect impact, as is the case with the multiplier and the general equilibrium approaches, the increase in GDP level compared to the baseline, by the end of the five-year projection period, ranges from 1 percent to 1¾ percent. The longer-term effects will depend on if these investments are maintained through higher public investment and can range between 1 percent to 2 percent.
References
Abiad, Furceri and Topalova, 2015. IMF Working Paper, No. 15/95: “The Macroeconomic Effects of Public Investment: Evidence from Advanced Economies.”
Alesina, Barbiero, Favero, Giavazzi and Paradisi, 2017. NBER Working Paper Series, No. 23385. “The effects of fiscal consolidations: Theory and evidence.”
Arslanalp, Bornhorst, Gupta and Sze, 2010. IMF Working Paper, No. 10/175: “Public Capital and Growth.”
Bernanke, Ben, Gertler, Mark and Gilchrist, Simon, 1999, The financial accelerator in a quantitative business cycle framework, ch. 21, p. 1341–1393 in Taylor, J. B. and Woodford, M. eds., Handbook of Macroeconomics, Vol. 1, Part C, Elsevier.
Bom andLigthart, 2014. Journal of Economic Surveys, Vol. 28, No. 5, pp. 889–916. “What have we learned from three decades of research on the productivity of public capital?”
Furceri and Li, 2017. IMF Working Paper, No. 17/217. “The Macroeconomic (and Distributional) Effects of Public Investment in Developing Economies.”
Ilzetzki, Mendoza and Vegh, 2013, Journal of Monetary Economics. Vol. 60, pp. 239–254. How big (small?) are fiscal multipliers?”
IMF, REO, 2018, April. “Fiscal Multipliers: How Will Consolidation Affect Latin America and the Caribbean?”
IMF WEO, 2014. “Chapter 3: Is it time for an infrastructure push? The macroeconomic effects of public investment.”
Kraay, 2014: American Economic Journal: Macroeconomics 2014, 6(4). pp. 170–208 “Government Spending Multipliers in Developing Countries: Evidence from Lending by Official Creditors.”
Kumhof, M., D. Laxton, D. Muir and S. Mursula (2010), “The Global Integrated Monetary Fiscal Model (GIMF), Theoretical Structure”, IMF Working Paper Series, WP/10/34.
Ramirez and Nazmi, 2003. Review of Development Economics, 7(1), pp. 115–126. “Public Investment and Economic Growth in Latin America: An Empirical Test.”
Annex I. Literature on Public Investment Multiplier
|
|
|
|
|
|
|---|---|---|---|---|
| Bom and Ligthart, 2014: What have we learned from three decades of research on the productivity of public capital? | A Meta-analytical technique survey: 68 studies and 578 estimates. Time series for 22 countries, 6 cross-section studies. | 1983–2008 | - In the short-run, the output elasticity of public capital installed at the central level of government is 0.083. - In the long-run, the elasticity increases to 0.122. - After correcting the publication bias, average output elasticity of public capital in the short-run increases to 0.106. |
- All countries considered in the studies are advanced economies, where 44 percent of the studies are based on the US data. - Focuses on papers that use production function approach only. - If public capital consists only of its core component and is installed by lower levels of government, the short-run multiplier becomes 0.154 and long-run multiplier reaches to 0.193. - Business cycles do not affect elasticity. - Public capital is more productive in the long-run. - Output elasticity of public capital has declined over time. |
| IMF, REO, April 2018: Fiscal Multipliers: How Will Consolidation Affect Latin America and the Caribbean? | Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Guatemala, Jamaica, Mexico, Paraguay, Peru, and Uruguay. | 1989–2016 | The multiplier for public investment in the contemporaneous year is estimated to be around 0.5 and reaches to almost 1.5 after 2 years. | - Expenditure vs. investment: Public investment multiplier reaches almost 1.5 - Degree of slack: No significant impact. - Sovereign default risk: a higher risk decreases multiplier to 0.6 after two years. With low default risk, the multiplier is 1.1. - Expenditure vs. tax: expenditure multiplier lies between 0.5 and 1.1. No difference between spending cut and tax hike multipliers. |
| Ramirez and Nazmi, 2003: Public Investment and Economic Growth in Latin America: An Empir ical Test | Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Mexico, Peru, and Uruguay | 1983–1993 | - An estimated coefficient of 0.014 for the effect of public investment on growth rate of real per-capita GDP. | |
| Furceri and Li, 2017: The Macroeconomic (and Distributional) Effects of Public Investment in Developing Economies | 79 emerging and low-income countries listed in IMF-WEO. | 1990–2013 | - A short-term public investment multiplier of 0.2 and a long-term public investment multiplier of 0.4. | - Business cycle: Multipliers are higher during recessions. Long-term multiplier is almost doubled. - Exchange rate regime: Multipliers are substantially higher under fixed-exchange rate regime. - Trade openness: Lower openness results in relatively higher multipliers. - Initial debt-to-GDP position: Lower ratios result in higher multipliers. - Investment efficiency: Multipliers are significantly stronger in countries with higher public investment efficiency. |
| Kraay, 2014: Government Spending Multipliers in Developing Countries: Evidence from Lending by Official Creditors | 102 Developing Countries | 1970–2010 | - One-year spending multiplier is estimated to be between 0.2 and 0.4. | - Multipliers are larger during recessions, with less international trade and under flexible exchange rate regime. |
| Ilzetzki, Mendoza and Vegh 2013: How big (small?) are fiscal multipliers? | 20 high income and 24 developing countries | 1960:Q1 to 2007:Q4 | Short-term multiplier (a quarter) High income: 0.57 Developing: 0.39 Long term multiplier High income: 1.5 Developing: 1.6 | - Multiplier is larger in industrial countries. - Fixed ER: short-term multiplier (0.15) long-term multiplier (1.4). - Close to trade: Short-term multiplier (0.61), long-term multiplier (1.1) - High-debt countries: multiplier is negative. |
| IMF WEO, 2014, Chapter 3: Is it time for an infrastructure push? The macroeconomic effects of public investment. | 36 Advanced and 153 Emerging Market and Developing Economies | 1985–2013 | Empirical Results for emerging and developing economies: The contemporaneous effect of a 1pp of GDP increase in public investment is a 0.25 percent increase in output and 0.5 percent four years after the shock. |
1. Degree of economic slack: output increases by about 1.5% in the same year and by 3 percent in the medium term in a low growth period. 2. Efficiency of public investment: high (low) efficiency: output increases by about 0.8 (0.2percent in the same year and by 2.6 (0.7 percent four years after. 3. The way the investment is financed: debt-financed shock: output increases by about 0.9 percent in the same year and by 2.9 percent four years after the shock. |
|
|
|
|
|
|
|---|---|---|---|---|
| Bom and Ligthart, 2014: What have we learned from three decades of research on the productivity of public capital? | A Meta-analytical technique survey: 68 studies and 578 estimates. Time series for 22 countries, 6 cross-section studies. | 1983–2008 | - In the short-run, the output elasticity of public capital installed at the central level of government is 0.083. - In the long-run, the elasticity increases to 0.122. - After correcting the publication bias, average output elasticity of public capital in the short-run increases to 0.106. |
- All countries considered in the studies are advanced economies, where 44 percent of the studies are based on the US data. - Focuses on papers that use production function approach only. - If public capital consists only of its core component and is installed by lower levels of government, the short-run multiplier becomes 0.154 and long-run multiplier reaches to 0.193. - Business cycles do not affect elasticity. - Public capital is more productive in the long-run. - Output elasticity of public capital has declined over time. |
| IMF, REO, April 2018: Fiscal Multipliers: How Will Consolidation Affect Latin America and the Caribbean? | Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Guatemala, Jamaica, Mexico, Paraguay, Peru, and Uruguay. | 1989–2016 | The multiplier for public investment in the contemporaneous year is estimated to be around 0.5 and reaches to almost 1.5 after 2 years. | - Expenditure vs. investment: Public investment multiplier reaches almost 1.5 - Degree of slack: No significant impact. - Sovereign default risk: a higher risk decreases multiplier to 0.6 after two years. With low default risk, the multiplier is 1.1. - Expenditure vs. tax: expenditure multiplier lies between 0.5 and 1.1. No difference between spending cut and tax hike multipliers. |
| Ramirez and Nazmi, 2003: Public Investment and Economic Growth in Latin America: An Empir ical Test | Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Mexico, Peru, and Uruguay | 1983–1993 | - An estimated coefficient of 0.014 for the effect of public investment on growth rate of real per-capita GDP. | |
| Furceri and Li, 2017: The Macroeconomic (and Distributional) Effects of Public Investment in Developing Economies | 79 emerging and low-income countries listed in IMF-WEO. | 1990–2013 | - A short-term public investment multiplier of 0.2 and a long-term public investment multiplier of 0.4. | - Business cycle: Multipliers are higher during recessions. Long-term multiplier is almost doubled. - Exchange rate regime: Multipliers are substantially higher under fixed-exchange rate regime. - Trade openness: Lower openness results in relatively higher multipliers. - Initial debt-to-GDP position: Lower ratios result in higher multipliers. - Investment efficiency: Multipliers are significantly stronger in countries with higher public investment efficiency. |
| Kraay, 2014: Government Spending Multipliers in Developing Countries: Evidence from Lending by Official Creditors | 102 Developing Countries | 1970–2010 | - One-year spending multiplier is estimated to be between 0.2 and 0.4. | - Multipliers are larger during recessions, with less international trade and under flexible exchange rate regime. |
| Ilzetzki, Mendoza and Vegh 2013: How big (small?) are fiscal multipliers? | 20 high income and 24 developing countries | 1960:Q1 to 2007:Q4 | Short-term multiplier (a quarter) High income: 0.57 Developing: 0.39 Long term multiplier High income: 1.5 Developing: 1.6 | - Multiplier is larger in industrial countries. - Fixed ER: short-term multiplier (0.15) long-term multiplier (1.4). - Close to trade: Short-term multiplier (0.61), long-term multiplier (1.1) - High-debt countries: multiplier is negative. |
| IMF WEO, 2014, Chapter 3: Is it time for an infrastructure push? The macroeconomic effects of public investment. | 36 Advanced and 153 Emerging Market and Developing Economies | 1985–2013 | Empirical Results for emerging and developing economies: The contemporaneous effect of a 1pp of GDP increase in public investment is a 0.25 percent increase in output and 0.5 percent four years after the shock. |
1. Degree of economic slack: output increases by about 1.5% in the same year and by 3 percent in the medium term in a low growth period. 2. Efficiency of public investment: high (low) efficiency: output increases by about 0.8 (0.2percent in the same year and by 2.6 (0.7 percent four years after. 3. The way the investment is financed: debt-financed shock: output increases by about 0.9 percent in the same year and by 2.9 percent four years after the shock. |
By Natasha Che.
The cross-country data on public capital investment and stock are from IMF Fiscal Affairs Department’s Public Capital database.
For scenario 2, growth would be higher than baseline from 2019 to 2022, and then drop lower than baseline for the subsequent two years due to higher base GDP.
GIMF simulations are run by Keiko Honjo.
For detailed documentation on the structure of the model see Kumhof and others (2010).
