Uruguay: Selected Issues
Author:
International Monetary Fund. Western Hemisphere Dept.
Search for other papers by International Monetary Fund. Western Hemisphere Dept. in
Current site
Google Scholar
Close

Selected Issues

Abstract

Selected Issues

Uruguay: Estimates of Fiscal Multipliers1

Fiscal multipliers are estimated for Uruguay using VAR models and local projection (LP) method. The results suggest that government consumption has only short-term output effects with cumulative multipliers of less than one and vanishing within five years whereas government investment has longer effect with multipliers steadily increasing to levels above two. On the other hand, tax shocks appear to have a negative short-term impact on output with impact multipliers between 0 and −1 and dissipating within two years.

A. Introduction

1. Fiscal multipliers measure the short-term impact of discretionary fiscal policy on output. They are usually defined as the ratio of a change in output to a discretionary change in government spending or tax revenue. Multipliers are expected to depend on country characteristics, and be lower with openness to trade, labor market rigidity, exchange rate flexibility, and high debt levels or in case of full employment (Batini and others, 2014).

2. The literature on fiscal multipliers suggests that first year multipliers generally lie between 0 and 1, and are higher in advanced economies (AE) than in emerging market economies (EMEs) and low-income countries (LICs). By looking at different studies, Mineshima and others (2014) show that first-year multipliers in AEs are on average 0.75 for government spending and 0.25 for government revenues. Due to factors such as expenditure inefficiencies, the limited empirical works available suggest multipliers are smaller in EMEs and LICs than in AEs (Estevão and Samake, 2013; Ilzetzki and others, 2013; Ilzetzki, 2011; and Kraay, 2012).

3. Beyond the first year, the persistence of the fiscal multipliers may differ depending on the fiscal instrument used. In case of indirect taxes, government consumption, and transfers, the model-based literature shows that permanent discretionary changes have only short-lived output effects (Anderson and others, 2013; Coenen and others, 2012). In contrast, permanent discretionary changes in public investment or corporate taxes have longer effects on output with multipliers steadily increasing to their long-term values (Coenen and others, 2012).

4. Uruguay is in the middle of a fiscal consolidation. When the current government took office in 2015, it set out a five-year budget plan that envisaged a reduction in overall fiscal deficit from 3.5 percent of GDP in 2014 to 2.5 percent of GDP in 2019 and an improvement in the primary balance of 1.6 percent of GDP. Half of this improvement has been expected to come from a reduction in current and capital expenditures whereas about one-fifth from higher central government revenues with the rest mainly coming from profits of public enterprises. The progress towards the set goals is proceeding with an estimated 3.3 percent of GDP overall fiscal balance in 2017.

5. Estimation of fiscal multipliers for Uruguay could give important insights in the context of the ongoing fiscal consolidation. Previous studies that estimate fiscal multipliers for Uruguay are scarce. In that regard, this paper contributes to the discussion about the short- to medium-term possible adverse effects of the fiscal consolidation by looking at the different fiscal instruments at the disposal of policy makers and their impacts on output.

6. The paper is organized as follows. In section B, we discuss the vector-autoregressive model (VAR) models and local projection (LP) method used to estimate fiscal multipliers. We present estimates of government spending and revenue multipliers in section C and check robustness of the results in section D. In the final section, we make some concluding remarks.

B. Methodology

7. We apply VAR models and local projection (LP) method (Jorda, 2005) to estimate fiscal multipliers for Uruguay. While VAR models are the conventional methodology in the literature, the LP method has become prominent in recent times. Using the two methods in conjunction not only enriches the analysis, but also strengthens the results and policy implications.

8. Our data ranges from 1999Q1 to 2017Q2. We use seasonally adjusted quarterly data for GDP and the fiscal variables that are converted into real terms using the CPI deflator. The fiscal variables include government consumption, government investment and tax revenues. In all analysis, government consumption excludes interest payments2, tax revenues are cyclically adjusted and changes in log of real effective exchange rate (REER) are used as control variables. Data sources for the variables are Haver and the Central Bank of Uruguay.

9. As our baseline, we use VAR models to estimate fiscal multipliers for government consumption, government investment and tax revenue. The reduced form VAR specification is:

Y t = a + D ( L ) Y t 1 + u t

Where a is a constant, Yt is a three-dimensional vector of endogenous variables, D(L) is an autoregressive lag polynomial and ut is the reduced form residual. Yt consists of the fiscal variable (government consumption, government investment or tax revenue), GDP and changes in REER. Lag length is set to be 2 based on AIC, SBC and LR test.

10. We follow the Blanchard and Perotti (2002) identification, which assumes that the fiscal variables are not affected by shocks to GDP within the same quarter. Although this assumption is straightforward for government spending components, it requires cyclical adjustment in case of revenues.

11. Alternatively, we use Jorda’s (2005) LP method to compute estimates of multipliers for Uruguay. The method entails the estimation of a series of regressions for each horizon h, such that:

X t h = a h + D h ( L ) Z t 1 + b h s h o c k t + e t + h , f o r h = 0 , 1 , 2 , ...

Where x is the variable of interest (GDP or the fiscal variable), z is a vector of control variables (which includes in our case two lags of the fiscal variable, GDP and changes in the real effective exchange rate), Dh(L) is a polynomial in lag operator and “shock” is the identified shock to the fiscal variable. The coefficient bh is the response of x at horizon t+h to the shock at time t.

12. Employing the Blanchard and Perotti (BP) identification to the LP method gives same contemporaneous responses of output to fiscal shocks as are derived through the VAR models. Since the set of controls, z, includes lagged measures of GDP, the fiscal variable and changes in REER, with BP identification, the shock is simply given by current value of the fiscal variable. However, when we extend the horizon the impulse response functions are constructed differently. For the LP method, these are sequence of bh estimates in a series of single regressions for each horizon whereas for VAR models, they are iterated forward from the estimated parameters of the VAR for horizon 0.

13. For both the VAR models and LP method, we use variables normalized by trend GDP obtained using HP. The usual practice of using the log of the variables would require converting the estimated elasticities by the sample average of the ratio of GDP to the fiscal variable to obtain multipliers. However, Ramey and Zubairy (2016) show that the variability of this ratio in time biases the multiplier estimates, usually upwards. To avoid this bias, we divide all the variables by an estimate of trend GDP and express them in the same units, thus directly estimating the multipliers.

14. Following Ramey and Zubairy (2016), we compute cumulative multipliers as the cumulative of the output response divided by the cumulative of the fiscal variable response. Many papers define multipliers as the ratio of the output response to the initial fiscal variable shock. However recent literature3 argues multipliers should instead be calculated as ratio of the integrals of the output response to the fiscal variable response because the integral multipliers address directly the policy question of measuring the cumulative GDP gain relative to the cumulative spending during a given period (for example, a budget year).

C. Empirical Results

15. We estimate fiscal multipliers for government consumption, government investment and tax revenues using the VAR models and the LP method. In general, the estimated multipliers are broadly similar across the two approaches for Uruguay (text table). Figure 1 presents cumulative orthogonalized impulse responses over 20 quarters to the different fiscal shocks from the VAR models. Caution is warranted when interpreting the results since the confidence bands are wide in most cases.

Figure 1.
Figure 1.

Cumulative Orthogonalized Impulse Responses to Fiscal Shocks

(VAR models)

Citation: IMF Staff Country Reports 2018, 024; 10.5089/9781484339824.002.A003

Source: Staff estimates

Uruguay: Estimates of Fiscal Multipliers

(cumulative)

article image
Source: Staff estimates

16. Government consumption has only short-term output effects with multipliers less than one and vanishing within five years. As expected, output responds positively to government consumption shocks, but the effect dies out within the first few years. Looking at the estimates of the multipliers computed from the VAR models and LP method, the impact multipliers for government consumption are less than one, reach their peak at the second or third quarter and dissipate afterwards. The government consumption multipliers for Uruguay are in line with the estimates obtained for Paraguay (David, 2017) and slightly higher than the ones for Peru (Vtyurina and Leal, 2016).

17. Government investment has longer effect on output with multipliers steadily increasing above two. Impulse response analysis shows that output responses are more persistent to government investment shocks. Both the VAR models and LP method suggest that while impact multipliers for government investment are less than one, they tend to increase gradually through time and reach levels above two. These results are similar to the government investment multipliers obtained for Paraguay (David, 2017) that increase from 0.1 to 2.1 in 5 years and for Peru (Vtyurina and Leal, 2016) that reach from 0.5 to 1.1 in 3 years.

18. Tax revenue shocks have negative short-term impact on output. The impulse responses from the VAR and LP models point to a negative response of output to tax revenue shocks in the short run. The impact multipliers implied by these impulse response functions and the LP method are between 0 and −1. Under the VAR model, the impact vanished within two years, whereas the LP method shows this effect switching to positive numbers after a year or so. While the impact tax revenue multipliers compare well with the ones estimated for Brazil (Matheson and Pereira, 2016), Paraguay (David, 2017) and Peru (Vtyurina and Leal, 2016), their short persistence differs from the constant or increasing magnitude in tax revenue multipliers obtained in these countries.

D. Robustness

19. We undertake some robustness checks around the baseline specification. We re-estimate alternative models in which the GDP deflator is used instead of the CPI deflator to convert nominal variables and/or where changes in nominal exchange rates are used rather than changes in REER. We use both the VAR models and the LP method under these robustness check specifications.

20. The results from the robustness exercises (not shown) are in line with the baseline specifications. They confirm the main findings that government investment multipliers are larger than government consumption and tax revenue multipliers, and that government consumption and tax revenue shocks have short-term impact on output whereas government investment shocks have much longer one.

E. Concluding Remarks

21. Uruguay is making progress towards reaching its overall fiscal deficit target of 2.5 percent of GDP in 2019. Attaining the deficit target is important to maintain public debt sustainability, improve investor confidence and enhance credibility of government policies. At the same time, policymakers need to consider options for minimizing the negative impact of fiscal consolidation on growth through an appropriate choice of fiscal policy instruments.

22. This paper finds that fiscal consolidation pursued through a combination of revenue increases and lower government consumption would have a more modest impact on growth than reductions in government investment in the short term. In particular, policymakers in Uruguay need to weigh the effects of reducing public investment (cut by close to 1 percent of GDP since 2015) on growth and consider a reorientation of budget spending from government consumption to capital spending.

References

  • Anderson, D., B. Hunt, M. Kortelainen, M. Kumhof, D. Laxton, D. Muir, S. Mursula, and S. Snudden, 2013, “Getting to Know GIMF: The Simulation Properties of the Global Integrated Monetary and Fiscal Model,” IMF Working Paper 13/55 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Batini, N., Eyraud, L., Forni, L., and Weber, A. (2014) “Fiscal Multipliers: Size, Determinants, and Use in Macroeconomic ProjectionsIMF Fiscal Affairs Department Technical Notes and Manuals 14/04 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Blanchard, O., and R. Perotti, 2002, “An Empirical Characterization of the Dynamic Effects of Changes in Government Spending and Taxes on Output,” Quarterly Journal of Economics, Vol. 117, pp. 132968.

    • Search Google Scholar
    • Export Citation
  • Coenen, G., C. J. Erceg, C. Freedman, D. Furceri, M. Kumhof, R. Lalonde, D. Laxton, J. Lindé, A. Mourougane, D. Muir, S. Mursula, C. de Resende, J. Roberts, W. Roeger, S. Snudden, M. Trabandt, and J. in’t Veld, 2012, “Effects of Fiscal Stimulus in Structural Models,” American Economic Journal: Macroeconomics, Vol. 4, No. 1, pp. 2268.

    • Search Google Scholar
    • Export Citation
  • David, A., 2017, “Fiscal Policy Effectiveness in a Small Open Economy: Estimates of Tax and Spending Multipliers in Paraguay,” IMF Working Paper 17/63 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Estevão, M and I. Samake, 2013, “The Economic Effects of Fiscal Consolidation with Debt Feedback,” IMF Working Paper 13/136 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Fisher, Jonas D.M. and Ryan Peters, 2010. “Using Stock Returns to Identify Government Spending Shocks.” The Economic Journal 120: 414436.

    • Search Google Scholar
    • Export Citation
  • Ilzetzki, E., 2011, “Fiscal Policy and Debt Dynamics in Developing Countries,” Policy Research Working Paper Series 5666 (Washington: The World Bank).

    • Search Google Scholar
    • Export Citation
  • Ilzetzki E., E. G. Mendoza and C. A. Vegh, 2013, “How Big (Small?) Are Fiscal Multipliers?Journal of Monetary Economics, Vol. 60, pp. 23954.

    • Search Google Scholar
    • Export Citation
  • Jorda, O. (2005) “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review, 95: 161182.

  • Kraay, A., 2012, “How large is the Government Spending Multiplier? Evidence from World Bank Lending,” Quarterly Journal of Economics, Vol. 127, No. 2, pp. 82987.

    • Search Google Scholar
    • Export Citation
  • Matheson, T. and Pereira, J. (2016) “Fiscal Multipliers for BrazilIMF Working Paper WP/16/79 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Mineshima, A., M. Poplawski-Ribeiro, and A. Weber, 2014, “Fiscal Multipliers,” in PostCrisis Fiscal Policy, ed. by C. Cottarelli, P. Gerson, and A. Senhadji (Cambridge: MIT Press).

    • Search Google Scholar
    • Export Citation
  • Mountford, Andrew and Harald Uhlig, 2009. “What are the Effects of Fiscal Policy Shocks?Journal of Applied Econometrics 24: 960992.

    • Search Google Scholar
    • Export Citation
  • Ramey, V. and Zubairy, S. (2016) “Government Spending Multipliers in Good Times and in Bad: Evidence from U.S. Historical Data”, mimeo, University of California, San Diego.

    • Search Google Scholar
    • Export Citation
  • Uhlig, Harald, 2010. “Some Fiscal Calculus.” American Economic Review 100(2): 3034.

  • Vtyurina, S. and Leal, Z. (2016) “Fiscal Multipliers and Institutions in Peru: Getting the Largest Bang for the SolIMF Working Paper WP/16/144 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
1

Prepared by Yehenew Endegnanew.

2

Interest payment is excluded since the government has limited discretionary control over it.

  • Collapse
  • Expand
Uruguay: Selected Issues
Author:
International Monetary Fund. Western Hemisphere Dept.
  • Figure 1.

    Cumulative Orthogonalized Impulse Responses to Fiscal Shocks

    (VAR models)