Fiscal Policy and Growth
Saima Nawaz and Idrees Khawaja[1]
Abstract
The paper attempts to analyze the impact of fiscal policy on economic growth while accounting for the level of development and controlling for the state of institutions. We extend the Solow growth model by incorporating fiscal policy and institutions through total factor productivity. The empirical analysis includes a panel of 56 countries. Results from ‘fixed effects’ and SYS-GMM show that the impact of fiscal policy on growth is statistically insignificant at the aggregate level. However the disaggregated estimation for the developed and developing economies shows that fiscal policy has a positive association with economic growth in developed economies and negative association in developing economies. The findings thus inform that fiscal policy contributes positively to growth only in developed economies. The reason for this seems to be the conducive institutional environment in developed economies e.g. a more efficient public sector. Such an institutional environment allows fiscal policy to play a positive role in developed economies and the absence of such an environment contributes to the negative impact of fiscal policy in developing economies.
Key words: Fiscal Policy, Institutions, Economic Growth
- Introduction
The role of governmental policy in promoting economic growth has been the focus of attention even in the times of Adam Smith. Kyenes gave a fresh impetus to ‘role of government’ after the great depression of 1930s. He argued in favour of government’s interventions to smoothen business cycles and to accelerate economic growth. Since then the role of governmental policies, especially the fiscal policy as a determinant of economic growth has received much attention. Keynesians argue that government can overcome recession by borrowing from money market and then injecting this back into the economy. Government spending raises demand, generates employment and raises the profitability of the firms leading to higher output. This suggests that public spending, even of recurrent nature, may positively contribute to economic growth. On the supply-side, fiscal policy can promote a competitive environment ─ public spending on developing infrastructure, especially power generation, road construction, education and health has a direct impact on economic growth in the long run(Easterly & Rebelo, 1993; Fölster & Henrekson, 2001; Mauro, 1998).
After the great depression of 1930s, the Keynesian activism was adopted by almost every nation. Studies that find a positive relationship between fiscal expansion and growth as well as the ones that find just the opposite abound and the empirical literature on the growth effects of fiscal policy remains inconclusive. It has also been suggested that the effectiveness of the fiscal policy varies with income level of a country(Butkiewicz & Yanikkaya, 2011; Grier & Tullock, 1989). Moreover the composition of public spending, which reflects the priorities of the government, also varies across income levels and this variation might be making it difficult to establish a clear relationship between fiscal policy and growth at the aggregate level(Nawaz, 2013).
The foregoing discussion suggests that the available literature does not unambiguously portray the nature of the relationship between fiscal policy and economic growth. One reason for the lack of clarity could be that attention has, by and large, remained focused on examining the aggregate relationships. Therefore there is a need to further explore these relationships ata disaggregated level. For example the role of fiscal policy in determining growth can be examined with reference to the stages of development/income levels of the nations. Secondly, and more importantly, the quality of institutions seems to influence the impact that government spending generates on growth. Growth effects of fiscal policy also do not seems to have been analyzed in depth while controlling for the quality of institutions. For example, the extent of rent seeking and corruption influence the impact that public spending generates on growth. Similarly bureaucratic quality and rule of law influence, through variety of channels, influence the impact of public spending on growth. This study seeks to examine the impact of fiscal policy on growth while accounting for the level of development and controlling for the state of institutions. A theoretical framework is developed to examine the impact of fiscal policy on economic growth while controlling for institutions.
This paper contributes to literature in various ways. First, we extend the augmented Solow growth model by assuming that technological advancement depends not only upon constant rate of technological progress (as envisaged in the neoclassical model) but also upon fiscal policy and the quality of institutions. This provides a framework to analyze the impact of both the fiscal policy and institutions on economic growth. Second, this study constructs a comprehensive index of institutional quality by incorporating various dimensions of institutions such as political stability, quality of administration and democratic accountability. Third, this study addresses the various econometric issues such as endogeneity by using the System Generalized Method of Moments (SYS-GMM) approach. The empirical investigation uses panel data of 56 countries over 1981-2010. We use fixed effects model and a dynamic panel estimation based on the SYS-GMM. The fixed effects model tackles the cross-sectional heterogeneity and the SYS-GMM takes into account theendogeneity.
The rest of the paper is structured as follow: section 2 presents the theoretical framework incorporating the role of institutions and fiscal policy in economic growth; section 3 discusses data and methodology; section 4 is devoted to reporting and analyzing the estimation results and section 5 concludes.
- Model specification: A variant of Augmented Solow Growth Model
To put the discussion in a framework, we modify the augmented neoclassical growth model. Our starting point is the standard neoclassical growth model with labor, human capital and physical capital as the factors of production:
where is real output, is the stock of physical capital, is stock of human capital and is raw labor. We assume Cobb–Douglas production function with constant return to scale for the entire economy. This function for the country at time is given by:
where is the share of physical capital, is the share of human capital and is the share of labor. We also assume that and 1, i.e. there is decreasing return to all factors of production and constant return to scale at the aggregate level[2]. is the advancement of technology and the efficiency of the factor of production in the country that represents the level of total factor productivity. Equation 2 states that the total output of the economy depends upon the physical capital, labor employed, skill of the employed labor force and the level of total factor productivity. In the standard neoclassical growth model, technological progress is the sole determinant behind economic growth. The exogenous technological progress, evolves with the constant rate of given by , where represents the growth rate of technology.
The literature shows that fiscal policy and institutions influence output through total factor productivity. The fiscal policy could have a direct and quick impact on economic growth in the short run if public expenditure is used to produce goods. However, to assess the long term impact it is more important to examine the influence of fiscal policy on the productivity of private inputs. Public expenditure may influence economic growth by increasing the productivity of the factors of production. This could be through improvement in the quality and quantity of factors like human and physical capital. The literature on endogenous growth has established beyond doubt the role of human capital in enhancing the total factor productivity and hence the growth rate of the economies[3]. However given the relatively longer gestation period of the investment in human capital the private sector may shy away from making investment this kind of investment. The government can take care of this shyness either by subsidizing private investment in human capital accumulation or by investing itself to accumulate human capital. Lucas (1988)strongly advocates publicly provided education to eliminate or at least reduce the externalities involved in human capital accumulation. Thus fiscal policy has a strong role to play in influencing economic growth by encouraging the accumulation of human capital. One of the ways in which the public expenditure may influence the TFP is through enhancing the productivity of complimentary inputs of the private sector. For example, the investment in public infrastructure may encourage specialization in sectors with higher productivity. In this regard,Chatterjee, Sakoulis, and Turnovsky (2003)demonstrates that transport infrastructure permit firms to have access to high-tech and high-volume production because it improves communication which increases the size of the market. In a similar vein Agénor and Moreno-Dodson (2006) argue that the provision of efficient public infrastructure may reduce the cost of private sector on maintenance of its stock of physical capital. To sum up, all the standard arguments that are put forth in favor of production of public goods are valid here but with a caveat ─ the relevant literature distinguishes between productive and non-productive expenditures (e.g. leakages) and attributes the positive impact of public expenditures on growth to productive expenditures only, with optimal level of spending (Dhont & Heylen, 2009; Kneller, Bleaney, & Gemmell, 1999; Tanzi & Schuknecht, 2000). As institutions facilitate cooperative activity by establishing ‘rules of the game’ therefore the institutions may influence economic growth through their impact on factor productivity. del Mar Salinas-Jiménez and Salinas-Jiménez (2011)argue that institutions may influence the efficiency level of factor and thereby its productivity. Thus institutions may affect growth through their influence on the aggregate production function. Dawson (1998)suggests that even the countries with abundant resources may experience lower standards of living due to lack of institutional “infrastructure” to support a system of efficient resource allocation. The failure of a country to adopt good institutions could adversely affect its economic growth. A. O. Krueger (1974)argues that weak institutional framework promotes rent seeking, a socially unproductive activity. For example the prevalence of corruption would require that more time be devoted to rent seeking activities, thus leaving little time for productive activities. Similarly Olson (1982) argues that given rent seeking the production of public goods could be inefficient due to the existence of weak institutional framework ─ corruption may distort the allocation of resources and their productivity by altering the incentive structure. This distortion may cause loss in efficiency and productivity of the resources(Nelson & Sampat, 2001; North, 1990). Moreover corruption may increase the cost of importing and using superior technology to such levels that it becomes difficult to import better technology(Bernard & Jones, 1996).
Based on this discussion, we extend the neoclassical model by assuming that technological advancement depends not only upon constant rate of technological progress (as envisaged in the neoclassical model) but also upon fiscal policy and the quality of institutions. In modeling ‘’ we hypothesize that ‘’ is determined by constant rate of technological progress ‘’, fiscal policy ‘’ and quality of intuitions ‘’. Therefore, the ‘’ is assumed to evolve according to the following function[4]:
where and are the parameters determining the outcome of fiscal policy and institutions. In this framework, the technological improvement depends not only on exogenous technological growth; but also on fiscal policy which tend to increase the productivity of inputs (Everaert et al., 2014; Glomm & Ravikumar, 1997; González & Pazó, 2008; Nijkamp & Poot, 2004) and the quality of institutions which tend to enhance productivity(Nelson & Sampat, 2001; North, 1990)[5].
We assume that both the physical and human capital face a similar production function and the depreciation rates for each type of capital are represented by. The assumption is in conformity with Mankiw et al. (1992). The population growth rate is and the time-invariant country-specific saving rates for each type of capital are given by. The technological progress evolves with the constant rate. Given the steady state assumption, and by using steady state physical and human capital to effective labor ratio; the steady state income per capita is obtained as:
Putting the value of ‘’ from equation into equation, assuming uniform depreciation rate for both types of capitals[6] i.e. physical and human, and taking log of equation; the per capita income is given by:
Equation describes that the steady state output per capita depends upon the accumulation of reproductive capital, the stock of technology, direct impact of fiscal policy and institutions and the indirect effect of institutions through the fiscal policy. The variant of the Solow model described above provides three important hypotheses.
i)Institutions influence economic growth through total factor productivity;
ii)Fiscal policy influences output growth through total factor productivity;
Using the equation 5 we develop the empirical model to quantify the impact of institutions and fiscal policy on economic growth. Following Mankiw et al. (1992), we assume that the growth rate of technological progress ( is the same for all countries therefore follows a deterministic trend in equation and, where is constant and is the country specific shock. The empirical model, in compact form, can be written as:
where represent GDP growth rate of country at time, is the quality of institutions for country at time, represents fiscal policy, and represents the matrix of control variables while is the disturbance term which is assumed to be serially uncorrelated and orthogonal to the explanatory variables. The vector of control variables includes: human capital, physical capital, initial income, trade openness, macroeconomic stability and growth rate of population. These variables have been frequently used in growth literature [see e.g. (Barro & Lee, 1996; Levine & Renelt, 1992; Mankiw et al., 1992)].
- Data and Methodology
To measure the impact of fiscal policy on economic growth, we employ a panel of 56 countries over 1981-2010 with values taken at 5 year interval[7]. The panel data estimation is considered as an efficient analytical method for analysis, since it allows us to include data for different cross sections i.e. countries and time periods. The purpose of using the five year interval data is to overcome the business cycle effects and account for the missing values. This practice is fairly standard (Acemoglu, Johnson, Robinson, & Yared, 2008; Esfahani & Ramı́rez, 2003; Nawaz, 2015; Temple, 1999)
The choice of 56 countries is mainly based on the availability of data ─ these are the countries for which we could gather the data on all the variables used in the models estimated[8]. The set of 56 countries included in the overall sample has been divided into two sub-groups based on the level of income. Low and lower middle income countries mainly fall in the category of developing countries while the upper middle and high income countries are termed as developed countries. We combine the first two categories i.e. low income and lower middle income countries into one sub-group which we call ‘Low Income Countries/Developing Countries’. Similarly we combine upper middle and high income countries into another sub-group and refer to this as ‘High Income Countries/Developed Countries’. In Low Income sub-group, we have 22 countries while in High Income sub-group we have 34 countries. The data on all the variables, except institutional quality and human capital is from the World Development Indicators (WDI) published by the World Bank. The data on institutional quality is from the Political Risk Services (PRS) Group and the data on human capital is fromBarro and Lee (2013)[9]. We measure the fiscal policy by using government spending, government consumption and taxation revenues. All expenditures and revenues have been normalized by GDPs of the respective countries.
To measure the quality of institutions, researcher have either used all components of the index or taken a few components or even a single component that best suited the objective of their study.Knack and Keefer (1995)have used a composite index of institutional quality by using five indicators from ICRG data including: Rule of law, Corruption in government, Bureaucratic quality, Risk of expropriation of assets by the government and Repudiation of contract by the government[10]. Subsequently, various studies have used this measure (Hall & Jones, 1999; Rodrik, 1997)or have used a single indicator; for example Rodrik (1999)use only bureaucratic quality, Mauro (1995)employs only corruption and Iqbal and Daly (2014) use democracy and corruption. Various studies have constructed an institutional quality index by taking the sum of all the twelve indicators included in the ICRG dataset(Papaioannou, 2009; Younas, 2009).
We construct institutional quality index using the six indicators from ICRG dataset that supposedly influence growth and live up to the definition of institutions. The six indicators are: Government Stability (GS_I), Investment Profile (IP_I), Control over Corruption (CC_I), Law and Order (LO_I), Democratic Accountability (DA_I), Bureaucracy Quality (BQ_I).These six indicators capture three different dimensions of institutions. First, these indicators capture the political stability mainly through government stability indicator. Second, these measures capture the quality of administration of the country. Level of administrative quality is assessed using: i) control over corruption, ii) quality of bureaucracy, iii) investment profile, and iv) law and order. Third, democratic accountability captures the role of democratic institutions in economic growth[11]. Nawaz (2015)has used the same indicators to quantify their impact on economic growth. As the six institutional indicators referred above are likely to be correlated, therefore using these six indicators, we develop an index of institutional quality. Using Principal Component Method (PCM), we have developed an Index institutional quality.