Año 2018
DOCUMENTOS DE TRABAJO FCEA ISSN 1909-4469 / ISSNe 2422-4642
Departamento de Economía
Fiscal Policy Effects and Capital Mobility in Some Latin American Countries José U. Mora Rafael A. Acevedo
No.39
Facultad de Ciencias Económicas y Administrativas, FCEA
Año 2018
DOCUMENTOS DE TRABAJO FCEA
No.39
ISSN 1909-4469 / ISSNe 2422-4642
Documento de Trabajo FCEA ISSN 1909-4469 / ISSNe 2422-4642 Año 2018 No. 39 Fiscal Policy Effects and Capital Mobility in Some Latin American Countries Autores: Jose U. Mora [
[email protected]] Departamento de Economía, Pontificia Universidad Javeriana – Cali, Colombia Rafael A. Acevedo [
[email protected]] Free Market Institute, Texas Tech University, USA. WEBSITE: wp_fcea.javerianacali.edu.co Comité editorial Alina Gómez Mejía Julián Piñeres Luis Fernando Aguado Correspondencia, suscripciones y solicitudes Calle 18 No. 118-250 Vía Pance Santiago de Cali, Valle del Cauca, Colombia Pontificia Universidad Javeriana Cali Facultad de Ciencias Económicas y Administrativas Teléfonos: (57+2) 3218200 Ext.: 8694 Correo electrónico:
[email protected] Sello Editorial Javeriano - 2018 Coordinador: Iris Cabra
[email protected] Concepto Gráfico: William Fernando Yela Melo Formato 28 x 21 cms. ©Derechos Reservados ©Sello Editorial Javeriano Enero de 2019 La serie de Documentos de Trabajo FCEA pone a disposición para el análisis, discusión y retroalimentación de la comunidad académica los avances y resultados preliminares del trabajo académico de los profesores de la Facultad de Ciencias Económicas y Administrativas. Estos documentos no han sido sometidos a procesos de evaluación formal por pares internos ni externos a la Facultad. Se espera que muchos de estos documentos posteriormente sean sometidos a evaluación en publicaciones especializadas.
Las opiniones expresadas en este documento son de exclusiva responsabilidad de los autores y no comprometen institucionalmente a la Facultad de Ciencias Económicas y Administrativas, ni a la Pontificia Universidad Javeriana Cali.
Fiscal Policy Effects and Capital Mobility in Some Latin American Countries
Jose U. Mora
Rafael A. Acevedo
Associate Professor Department of Economics Pontificia Universidad Javeriana - Cali E-mail:
[email protected]
Research Associate Free Market Institute Texas Tech University Founder - Director of Econintech.org E-mail:
[email protected]
Abstract This paper studies the relationship between the size of the fiscal multiplier and the degree of capital mobility in some Latin American countries. The Mundell (1963) and Fleming (1962) model establishes that this effect might be very large or very small (very close to zero) depending on the exchange rate regime and the degree of capital mobility; the potency of fiscal policy is inversely correlated with the degree of capital mobility. We use Mora (2013) model to argue that the multiplier might not be negatively correlated with capital mobility in these countries. In other words, the potency of fiscal policy could be reduced by the fact that capital mobility in Latin American countries tends to be quite low. The empirical findings support the hypothesis. We have found that the size of the fiscal multiplier tends to increase or (at least) to remain relatively stable, around 1.40, in these countries in the short run; however, in the long run, this effect tends to decrease significantly to 0.34. These results also suggest that fiscal policy is still very potent, but given their economic structure differences, could be larger if Latin American countries become more financially integrated with the rest of the world. Keywords: fiscal policy, business cycles, Latin America JEL: E62, E12, F41, O54
Efectos de la Política Fiscal y la Movilidad de Capitales en Algunos Países Latinoamericanos
Resumen Este trabajo estudia la relación entre el tamaño del multiplicador fiscal y el grado de movilidad de capitales en algunos países de América Latina. El modelo de Mundell (1963) y Fleming (1962) establece que el tamaño del multiplicador puede ser grande o pequeño (cercano a cero) dependiendo del régimen cambiario y el grado de movilidad de capitales; la potencia de la política fiscal está inversamente correlacionada con el grado de movilidad de los capitales a través de las fronteras. En este trabajo se usa el modelo de Mora (2013) y se argumenta que el multiplicador podría no estar negativamente correlacionado con la movilidad de capitales en estos países. El tamaño del multiplicador podría ser muy pequeño debido al hecho de que la movilidad de capitales en estos países es muy baja. Los resultados que se muestran en este trabajo apoyan esta hipótesis. Se encontró que el tamaño del multiplicador fiscal tiende a aumentar o a permanecer estable alrededor de 1,40 en el corto plazo en estos países. Sin embargo, en un horizonte de tiempo un poco más largo, este efecto tiende a disminuir significativamente a niveles alrededor de 0,34. Estos resultados sugieren que la política fiscal es muy potente, pero dadas las diferencias de estructuras económicas, esta potencia de la política fiscal podría ser mucho mayor sí los países latinoamericanos se integraran más financieramente con el resto del mundo. Palabras clave: Política fiscal, ciclos económicos, Latinoamérica JEL: E62, E12, F41, O54
1. Introduction In the aftermath of the 2008-2009 crisis, there was an important surge in developed and developing countries on the use of fiscal instruments as means for the recovering of their economies. This phenomenon brought back the attention to the issue of the relative effectiveness of fiscal policy and, more particularly, on the size of the fiscal multiplier. However, there is no a unique agreement. On one side, there are Economists who argue that the size of the multiplier is not a constant among countries. It varies across time and countries and these variations are determined by factors such as the exchange rate regime, the degree of openness and capital mobility, the foreign debt as a percentage of GDP, and the stability and flexibility of the financial system (Corsetti, Meier, & Muller, 2012). On the other side, there is no agreement within countries with respect to the size of the multiplier. In this sense, the best example is the U.S. regarding the different positions adopted by Economists with respect to the effectiveness of the 787,000 million dollars approved by the U.S. congress in february 2009. On one side, Christina Romer, Director of the Economic Council Advisors at that time, argued that the multiplier was relatively high and used a multiplier of 1.6 to compute the impact that this fiscal aid would have had on the economy (Romer & Bernstein, 2009). On the other side, Robert Barro (2009) pointed out that during peace time the multiplier was cero and henceforth the fiscal stimulus would be ineffective. Had been this the case, the U. S. economy would not have recovered 3.7 million jobs by the end of 2010 (Ilzetski, Mendoza, & Végh, 2013). Why is the multiplier of fiscal policy so important in a more economically integrated world? The answer lies on the extensions of the keynesian model and the contributions of Mundell (1963) and Fleming (1962). The open economy keynesian model predicts that the higher the trade openness of an economy, the smaller the size of the multiplier and the lower the potency of fiscal policy. More particularly, the theory suggests that the increase in real output induced by an expansionary fiscal policy leaks through imports reducing the effectiveness of fiscal policy if the marginal propensity to import is high. However, the effectiveness of fiscal policy is higher (the highest) if the marginal propensity to import is low (zero, the closed economy). The Mundell-Fleming model establishes that this effect might be very large or very small (very close to zero) depending on the exchange rate regime and the degree of capital mobility; the potency of fiscal policy is inversely correlated with the degree of capital mobility. More precisely, in small open economies under floating exchange rates and perfect capital mobility, a fiscal expansion causes an increase in income and interest rates that produces an appreciation of the domestic currency and, as a result, diminishes or eliminates the stimulus of the fiscal policy; by the same token and under floating exchange rates, the lower the degree of capital mobility, the higher the effectiveness of fiscal policy. Nevertheless, if the economy faces a fixed exchange rate regime, then the degree of capital mobility enhances the potency of fiscal policy.
1
This topic is fundamental for Latin American countries since fiscal reforms in these countries are very common. An implicit assumption in the Mundell-Fleming and Keynesian models is that the economy is fully diversified, no productive sector is more important than another, and domestic financial markets are fully integrated to the world financial system. Latin American countries, on the contrary, are very different. These are economies characterized by the dominance of one or two economic activities that contribute to the bulk of exports. In this sense, Mora (2013) in a theoretical model of aggregate demand and supply for small open economies, like Latin American economies, under the assumptions of floating exchange rate regimes and imperfect capital mobility found that the potency of fiscal policy is positively correlated with the degree of capital mobility. This result is at odds with the Mundel-Fleming and Keynesian models. As a result, the effectiveness of fiscal policy in Latin America might be limited by their low capital mobility. If this hypothesis holds, then any fiscal reform implemented to make an economy grow faster might produce little or no benefits at all on the overall economic activity; only higher inflation rates and domestic currency depreciations will prevail. Therefore, not only opening their economies to the rest of the world is desirable, but also integrating their financial markets to the world financial system, at least in the short run. This paper aims to study the relationship between the size of the fiscal multiplier, openness, and capital mobility in Latin American countries. We use panel data on information collected from the Penn World Table and World Bank for 12 countries for the period 1980 to 2014. The empirical results suggest that there are fixed effects and random effects that affect the size of the multiplier across countries and across time. This implies that fiscal policy might prove very affective at changing real output. These results are very important because the size of the multiplier and how it behaves over time gives us more precise ideas on when and how to use fiscal policy in order to stabilize or stimulate the Latin American economies over the different phases of their business cycles. This paper is organized as follows. The following section discusses the literature. Section three presents the theoretical model from which we derive our hypothesis. Next, section four discusses the empirical methodology and presents the estimated results. Finally, section five concludes.
2. Literature review During and after the recent financial crisis, several developed and developing countries have been using expansionary fiscal policies, increasing government spending or reducing taxes, to stabilize or stimulate their economies. These has brought back significant attention not only in terms of research but also in terms of economic policy on the impact of the fiscal multiplier, that is, on the size of the fiscal multiplier. As a result of the debate, there is no 2
consensus on this issue. Firstly, for most Economists the value of the multiplier is not the same constant for all countries. Its value depends on characteristics such as the openness to foreign trade, the percentage of debt to GDP, the exchange rate regime, the degree of capital mobility, and the health of the financial system. Corsetti, Meier, & Muller (2012) argues that these characteristics may affect the effectiveness of fiscal policy to change domestic output. And, second, even within the same country, Economists disagree on the size of the multiplier and its effectiveness, as pointed out by the debate between Romer & Bernstein (2009) and Robert Barro (2009) since it could be very low during peacetime as pointed out by Barro (2009) and Barro & Redlick (2011), position related to those who argue that the size of the multiplier would depend on the state of the economy. Although the literature on this topic is growing very rapidly, we will discuss the most relevant and recent papers for the present research. There is a great amount of papers that estimate the size of the fiscal multiplier for different countries, mostly for the U.S. (Whalen, 2015; Ramey & Zubairy, 2014; Owyang, Ramey, & Zubaire 2013; Auerbach & Gorodnichenko, 2012a; Mustea, 2015; and Batini, Eyraud, Forni, and Weber, 2014; among others), and for many other developed, developing and emerging countries or groups of economies (Auerbach & Gorodnichenko, 2012b; see Batini et al, 2014, for details). However, results for Latin America are scarce except for the contributions of Fraga, Briceño, and Heras (2016) for the most important Latin American economies, Puig (2014) for Argentina, Moura (2015) and Costa and Vaz Sampaio (2016) for Brasil, and Laverde (2010) for Colombia, who use mostly VAR techniques and DSGE models to compute the short and long run fiscal multipliers. The size of the multiplier depends on several characteristics such as the state of economic activity, response of monetary policy to fiscal policy, openness to international trade, exchange rate regime, foreign debt, and financial system health. Auerbach and Gorodnichencko (2015, 2012a) argue that the economic state might affect the size of the multiplier which implies that this could be higher during recessions, when unemployment is high, or when the interest rate is near the zero-lower bound (see also theoretical contributions from Woodford 2011, Eggertsson 2011, and Christiano et al. 2011, among others). Auerbach and Gorodnichencko (2012a) employ an SVAR model that allows differentiated responses across recessions and expansions and use the regime-switching model to compute the government spending multiplier for the U.S. They find that during recessions the multiplier could be as large as 2.5 and as low as 0.6 during expansions. These results could be attributed to the facts that when the economy is near full capacity, fiscal policy effectiveness could be mitigated by a reduction in private-sector spending; however, when resources are more unused, then the impact of the fiscal stimulus is magnified by additional spending in the private sector. Auerbach and Gorodnichenko (2012b) extend the analysis, using an expectations-augmented VAR, to a large number of the Organization for Economic Cooperation and Development (OECD) countries. Their findings confirm their previous results that multipliers are larger during recessions than during expansions. Other studies, including 3
Fazzari et al. (2014); and Baum et al. (2012), that use different types of VAR models also find fiscal multipliers that are larger in economic contractions than in expansions. On the contrary, Owyang et al. (2013) and Ramey and Zubairy (2014) construct long data sets for the US and Canada that extend beyond the great depression (even to late 1800s) and use a narrative approach to analyze the effectiveness of government spending in the presence of announcements about military events. They show that in the U.S. fiscal multipliers are not higher during periods of economic recessions whereas for Canada the evidence suggests that multipliers are significantly higher during periods of slack in the economy (Owyang et al., 2013). Ramey and Zubairy (2014), assuming that longer periods of data might contain potential information, create a data set for the U.S. that extends back to 1889. They find, first, that some of the most cited findings in the literature about the size of the multiplier are due to details related to the way they are computed and are not robust to plausible generalizations. Second, they find no evidence at all that the government purchases multiplier for the U.S. is high during periods of economic slack. And third, their findings about higher multipliers when the economy is near the zero lower bound are only related to the exclusion of the rationing periods of WWII but warn that these estimates are imprecise and not robust for generalizations. Hall (2009), Christiano, Eichenbaum, & Rebelo (2011), Davig and Leeper (2011), and Coenen et al. (2012) have found that the size of the fiscal multiplier might be limited by the response of the Federal Reserve when it increases the federal funds rate as a countermeasure due to inflationary pressure under normal economic conditions. However, they argue that there can be times when fiscal policy effectiveness might be higher; for instance, when monetary policy is restricted by the zero lower bound interest rate. The exchange rate, public indebtedness and health of the financial system might also affect the multiplier effect. Corsetti et al (2012) use a panel for 17 OECD countries covering the period from 1975 to 2008. They perform the estimation in a two-step procedure, identifying fiscal shocks as residuals from an estimated spending rule and then use these residuals to trace their macroeconomic effects under different conditions: exchange rate regime, public debt, and health of the financial system. Under a positive fiscal shock, the unconditional response confirms reported results in the literature. The conditional response differs systematically across the different exchange rate regimes, given that real appreciation and foreign deficits occur mainly under currency pegs. Additionally, output and consumption multipliers are relatively high during periods of financial crisis. They suggest that this happens because private spending is more likely to be restricted due to credit access during financial crisis. Karras (2011) investigates the implications of exchange rate regimes on the size of fiscal multiplier by means of panel data for 61 developed and developing countries using annual data from 1951 to 2007. Empirical results show that fiscal expansions are more effective under both exchange rate regimes in the short run than in the long run. Moreover, 4
the multiplier is higher under fixed exchange rates than under floating. He argues that consumption is in fact a key variable because government spending crowds out private spending under flexible exchange rates while under fixed exchange rates private consumption increases. Karras (2012) investigates whether fiscal policy effectiveness depends on trade openness using panel data from 1951 to 2007 for 62 developing and developed countries. Trade openness is defined as the sum of exports and imports as a percentage of GDP. The sample have a variety of countries that exhibit different degrees of trade openness among them and over time. The empirical estimates indicate that trade openness negatively affects the effectiveness of fiscal policy, particularly pointing out that an increase of 10% of GDP reduces the fiscal multiplier by approximately 5 to 6%. Similar results are found by Yang (2016) when openness and financial openness are considered. In this context, financial openness is in some sense a proxy for capital mobility as we use it in our paper. However, in our paper, we test the relationship for Latin American countries using a different specification. It is also important to point out that the different types of government spending might produce different impacts on the economy. For instance, Gechert (2015), based on meta-regression analysis over 104 papers that provide 1063 reported multiplier values, performs a panel regression analysis over the entire sample and different subsamples. On average, the estimated multiplier is around 0.73 and, depending on the specification, methods used in the estimation, and control variables, it could vary between 0.09 to 1.4. Particularly, this variation comes from public investment which increases the average multiplier by 0.6 whereas taxes and transfers lower it by 0.3 to 0.4 units. Ilzetzki, Mendoza, and Végh (2013) use the SVAR approach with quarterly data for 44 developed and developing countries for the period 1960:1 - 2007:4. Their results indicate that in most cases the effect of government consumption is very small in the very short run (first quarter). In open economies with floating exchange rates, the multiplier is negligible with no effect in the long run. However, the long run impact is significant if the economy is closed and has fixed exchange rates. Furthermore, a fiscal stimulus in high-indebted countries (more than 60% of GDP) might lead to strong negative effects on output. Finally, it is important to emphasize that the literature shows a large amount of different results for the size and effectiveness of the fiscal multiplier in the short and long run. Most of the results are for developed countries which by the characteristics of their economies differ from the ones of Latin America. Therefore, this paper intents to contribute to fill in the existing gap in the literature for studies on the size of fiscal multipliers in Latin America.
5
3. The theoretical model Although the Mundell-Fleming model (Mundell, 1963 & Fleming, 1962) constitutes an interesting starting point, it is an aggregate demand model because it rests on the assumption that the price level is constant, the aggregate supply is perfectly elastic at the equilibrium price level, and as a result there is no inflation. This is a very restrictive assumption if we want to analyze Latin American economies where inflation has been an important phenomenon during the last 38 years. This paper aims to expand this model allowing prices (and salaries) to vary and, at the same time, describe the vast of Latin American where one or two productive sectors dominate the economy.
3.1 Assumptions According to Mora (2013) suppose that there is a small open economy with two productive sectors: sector 1 extracts, produces, and exports a single raw material or commodity (name it crude oil, copper, coffee, or any other mineral or agricultural product) that is very important ∗ for the economies of the world and whose price, 𝑝1,𝑡 , is determined in international markets, and the domestic economy takes such price as given. Sector 2 produces a composite good that can only be sold and consumed in the domestic market at a price 𝑝𝑡 . Finally, given that the economy is small and it is not fully integrated to the world financial markets, let´s assume there is imperfect capital mobility between the domestic economy and the rest of the world. Contrary to the Mundell-Fleming model, let´s assume that prices are relatively flexible and that workers are unionized and use the consumer price index (CPI) to negotiate their wage increase every period. Suppose that the CPI is computed as follows: 𝐶𝑃𝐼𝑡 = (𝐸𝑃𝑡∗ )𝜈 (𝑃𝑡 )1−𝜈
(1)
where 𝜈, 𝑃𝑡∗ , 𝑃𝑡 , and 𝐸 are, respectively, the weight that domestic prices receive in the composition of the CPI index, the foreign price level, the domestic price level, and the nominal exchange rate (domestic currency per U.S. dollar). Wages in every sector are indexed according to the following equation: 𝑊𝑖,𝑡 = 𝜛𝑖 𝐶𝑃𝐼𝑡 = 𝜛𝑖 (𝐸𝑃𝑡∗ )𝜈 (𝑃𝑡 )1−𝜈
(2)
where 0 ≤ 𝜛𝑖 ≤ 1 is the weight used for salary indexation in sector 𝑖. If 𝜛𝑖 = 0 wages in sector 𝑖 are rigid and do not change; if 𝜛𝑖 = 1, then wages in sector 𝑖 are completely flexible and adjust instantaneously by the change in the CPI. Taking logs1 from the previous 1
From now on, lowercase letters are used for natural logs of variables, except when the contrary is indicated.
6
expression we obtain: 𝑤𝑖,𝑡 = 𝜛𝑖 + 𝜈𝑒 + 𝜈𝑝𝑡∗ + (1 − 𝜈)𝑝𝑡
(3)
Following Mora (2013), let´s assume that firms in both sectors are profit maximizers and use Cobb-Douglas constant returns to scale technologies. Under this optimizing behavior, the aggregate supply function would be. 𝑦𝑡 = (𝜃2 − 𝜃1 (
1−𝜈 𝜈
)) 𝑝𝑡 − 𝜈(𝜃1 + 𝜃2 )𝑝𝑡∗
(4)
−(𝜃2 + 𝜃1 )𝜛𝑖 − (𝜃2 − 𝜃1 (
1−𝜈 )) 𝑒 + 𝑘1 + 𝑘2 + 𝜃2 𝑙𝑛(1 − 𝜑) 𝜈
1 1 ∗ +𝜃1 (𝑙𝑛(1 − 𝜃 )) + 𝜃1 𝑝1,𝑡 + 𝑧1,𝑡 + 𝑧2,𝑡 𝜃 𝜑 ∗ where 𝑘1 and 𝑘2 are constants that represent the (logs of) capital stocks, 𝑝1,𝑡 is world price of the tradable good, and 𝑧1,𝑡 and 𝑧2,𝑡 are supply shocks in each productive sector, respectively. This function has a positive slope in the space (𝑦, 𝑝) as long as 𝜃2 − 1−𝜈 𝜃1 ( 𝜈 ) > 0. Likewise, if the supply function has a positive slope, then a depreciaton of the domestic currency causes an increase in production of the tradable (exporting) good and a contraction in output on the other sector, reducing aggregate total output since 1−𝜈
− (𝜃2 − 𝜃1 (
𝜈
)) < 0.
Aggregate demand is determined by the simultaneous equilibria in the real market, the money market, and the foreign market, given the last by the balance of payments balance. The real market equilibrium is given by the expression (in logs) ∗ 𝑦𝑡 = 𝜎𝑝 𝑝1,𝑡 + 𝛼 (𝑒 + 𝑝𝑡∗ − 𝑝𝑡 ) − 𝛾𝑟𝑡 + 𝜎𝑓 𝑦𝑡∗ + 𝑧𝐼𝑆
(5)
where 𝑦𝑡∗ is the log of foreign real output2, 𝑟 is the real interest rate, and 𝑧𝐼𝑆 is a real shock ( 𝑧𝐼𝑆 ~(0, 𝜎𝑍2𝐼𝑆 )). In this paper, this shock would be associated with a fiscal expansion. The money market equilibrium requires that demand equals supply for real balances: 𝑚𝑡 − 𝑝𝑡 = 𝛽0 𝑦𝑡 − 𝛽1 𝑖𝑡 + 𝑧𝑀
(6)
where 𝑚𝑡 is the natural logarithm of the stock of money, 𝛽0 , 𝛽1 > 0 are, respectively, the 2
The main trading partner for many Latin American economies is the U.S.
7
elasticity of real income, 𝑦𝑡 , and the semi-elasticity of the nominal interest rate, 𝑖𝑡 , and 𝑧𝑀 is an unanticipaded shock in the money market. Note that the interest rate is the ex-ante nominal rate and is determined as follows: 𝑖𝑡 = 𝑟𝑡 + 𝜋 𝑒 where 𝜋 𝑒 is the expected value in period 𝑡 of the inflation rate in period 𝑡 + 1. In order to simplify the analysis, suppose that inflation expectations are adaptive and, as a result, 𝜋 𝑒 = 𝜋𝑡−1 . Substituting this result into the previous equation and then into equation (6), the money market equilibrium will be given by: 𝑚𝑡 − 𝑝𝑡 = 𝛽0 𝑦𝑡 − 𝛽1 𝑟𝑡 − 𝛽1 𝜋𝑡−1 + 𝑧𝑀
(7)
Finally, we must obtain the foreign equilibrium. Considering imperfect capital mobility, foreign equilibrium will be given by the balance of payment equation as shown next ∗ ∗ ∗ ∗ ∗ 𝑝 𝐵𝑡 = 𝜎𝑝1,𝑡 1,𝑡 + 𝜎𝑥 (𝑒 + 𝑝𝑡 − 𝑝𝑡 ) + 𝜎𝑓 𝑦𝑡 − 𝜎𝑦 𝑦𝑡 + 𝑧𝐸𝑋 + 𝜆(𝑟 − 𝑟 ) = 0
(8)
∗ , 𝜎𝑥 , 𝜎𝑓 , and 𝜎𝑦 are, respectively, the balance of payment elasticities with where 𝜎𝑝1,𝑡
respect to the tradable good price, with respect to the real exchange rate (𝑒 + 𝑝𝑡∗ − 𝑝𝑡 ), with respect to foreign real output, 𝑦𝑡∗ , and with respect to domestic real income. 𝜆 > 0 stands for the degree of capital mobility. For instance, a value of 𝜆 equal to zero implies that there is no capital mobility between foreign and domestic financial markets; however, a very large value of lambda (𝜆 = ∞) implies that there would not be barriers to financial transactions and foreign and domestic assets would be perfect substitutes of one another. Again, as in previous cases, 𝑧𝐸𝑋 ~(0, 𝜎𝑍2𝐸𝑋 ) . Equations (4), (5), (7), and (8) conform a system of equations that represent the aggregate supply and aggregate demand of this economy. Since the system is linear, we can totally differentiate the system and use matrix algebra to find the effects of any shock to the economy.
3.2 Theoretical results According to this model, a domestic fiscal expansion will produce the following results. Let 1−𝜈
|𝐷 | = − (𝜎𝑥 𝛾 + 𝛼𝜆 + (𝜃2 − 𝜃1 (
𝜈
)) (𝛾𝜎𝑦 + 𝜆)) < 0
be the determinant of the
coefficient matrix of the system. Then, by Cramer´s rule, the results of a fiscal policy shock are as follows:
8
𝑑𝑦𝑡 = 𝑑𝑧𝐼𝑆
𝑑𝑝𝑡 = 𝑑𝑧𝐼𝑆
𝑑𝑒𝑡 = 𝑑𝑧𝐼𝑆
1−𝜈 −𝜆 (𝜃2 − 𝜃1 ( 𝜈 )) |𝐷 |
>0
(9)
1−𝜈 −𝜎𝑥 𝛽1 + (𝜃2 − 𝜃1 ( 𝜈 )) (𝜆𝛽0 − 𝜎𝑦 𝛽1 ) |𝐷 |
⋛0
(10)
1−𝜈 (𝜆 − 𝜎𝑥 𝛽1 + (𝜃2 − 𝜃1 ( (𝜆𝛽0 − 𝜎𝑦 𝛽1 )) 𝜈 )) ⋛0
|𝐷 |
𝑑𝑟𝑡 =− 𝑑𝑧𝐼𝑆
1−𝜈 (𝜎𝑥 + 𝜎𝑦 (𝜃2 − 𝜃1 ( )) 𝜈 |𝐷 |
>0
(11)
(12)
Equations (9) and (12) show the standard results of a government spending increase while equations (10) and (11) show ambiguous signs. However, these signs will depend upon the signs of 𝜆𝛽0 − 𝜎𝑦 𝛽1 and 𝜆 − 𝛽1 𝜎𝑥 . If the economy faces low capital mobility 𝜆𝛽0 < 𝜎𝑦 𝛽1 ,
𝜆 < 𝛽1 𝜎𝑥 ,
and, as a result,
𝑑𝑒𝑡 𝑑𝑧𝐼𝑆
>0
and
𝑑𝑝𝑡 𝑑𝑧𝐼𝑆
> 0.
This means that an
expansionary fiscal policy will cause an increase in the domestic price level and the nominal exchange rate. If, on the contrary, the economy faces high capital mobility, 𝜆𝛽0 > 𝜎𝑦 𝛽1 and 𝜆 > 𝛽1 𝜎𝑥 , then
𝑑𝑒𝑡 𝑑𝑧𝐼𝑆
< 0 and
𝑑𝑝𝑡 𝑑𝑧𝐼𝑆
< 0. These results imply that an increase in
government spending will cause a fall in the domestic price level and the nominal exchange rate. Even though these results are important, the main contribution of this paper comes from the magnitude of the result in equation (9), which is the fiscal policy effect on output. To see how capital mobility affects the size and sign of this effect, let´s take the derivative of
𝑑𝑦𝑡 𝑑𝑧𝐼𝑆
with respect to 𝜆. 1−𝜈 1−𝜈 𝑑𝑦𝑡 ) [𝜃2 − 𝜃1 ( 𝜈 )] [𝜎𝑥 𝛾 + 𝛾𝜎𝑦 (𝜃2 − 𝜃1 ( 𝜈 ))] 𝑑𝑧𝐼𝑆 = >0 2 𝜕𝜆 1−𝜈 [𝜎𝑥 𝛾 + 𝛼𝜆 + (𝜃2 − 𝜃1 ( (𝛾𝜎𝑦 + 𝜆)] 𝜈 ))
𝜕(
(13)
This is a very important result with significant implications for Latin America. It states that the degree of capital mobility enhances the potency of fiscal policy. Therefore, it is expected 9
that since most countries in the region are characterized by domestic financial markets not well integrated to the financial world, which implies a relatively low capital mobility, then the potency of fiscal policy might be limited. Besides, as equations (14) and (15) below show, the lower the degree of capital mobility, the higher the costs in terms of inflation and depreciation of the domestic currency associated with a fiscal expansion, which are consistent with some events in Latin American economies. 𝑑𝑝𝑡 ) 𝑑𝑧𝐼𝑆
𝜕(
𝜕𝜆
𝜕
𝑑𝑒𝑡 𝑑𝑧𝐼𝑆
𝜕𝜆
=
1−𝜈 1−𝜈 2 )][−𝛽0𝜎𝑥 𝛾−𝜎𝑥 𝛽1 −𝛼𝜎𝑦 𝛽1 ][𝜃2−𝜃1( )] [−𝜎𝑦 𝛽1 −𝛽0𝛾𝜎𝑦 ]−𝜎𝑥 𝛽1𝛼 𝜈 𝜈 2 1−𝜈 [𝜎𝑥 𝛾+𝛼𝜆+(𝜃2 −𝜃1( ))(𝛾𝜎𝑦 +𝜆)] 𝜈
[𝜃2 −𝜃1(
