The nexus between taxes and economic growth has been extensively explored in the theoretical and empirical literature. The theoretical foundation of this relationship can be traced as far back to Solow (1956) and Swan (1956). One of the main predictions from this work was that growth simply depends on the accumulation of physical and human capital investments. Taxes may exert only temporary effects on the growth rate of income in the transition to successive equilibrium growth paths. The Solow-Swan neoclassical growth model therefore predicts that steady state growth is not affected by tax policy. However, endogenous growth models contend that taxes have a great impact on economic growth through the return on capital accumulation and the volume of investments in R&D (see, inter alia, Barro, 1990, 1991; King and Rebelo, 1990; Jones et al., 1993; Stokey and Rebelo, 1995; Barro and Sala-i-Martin, 1995; Mendoza et al., 1997).
Following the seminal work of Barro (1990), the economic growth-taxes nexus has generated extensive body of empirical literature. These include studies for different geographic areas as well as various sample periods. Roughly, we can categorize past studies in this field into two broad strands. The first strand examines the relationship between the overall level of taxes and economic growth (4). A general conclusion from this strand of literature is that the empirical results of the previous studies are mixed and have not reached a consensus. While some studies document a negative relationship between taxation and growth (e.g., Plosser, 1992; Engen and Skinner 1992; Mullen and Williams, 1994; Bleaney et al, 2001, Folster and Henrekson, 2001; Padovano and Galli, 2002; Tomljanovich, 2004; Holcombe and Lacombe, 2004; Koch et al., 2005; Reed, 2008; Ferede and Dahlby, 2012), the others do not detect any significant correlation, neither in the long-nor in the short-run (Koester and Kormendi, 1989; Levine and Renelt (1992), Easterly and Rebelo, 1993; Mendoza, et al., 1997). On the other hand, Myles (2000) maintains in a survey that the tax impact on growth is very weak.
The second strand is composed of the studies which focus on the nexus between tax structure and economic growth. This nexus suggests that different types of taxes affect growth in diverse ways. Theoretically, many scholars (see, for example, King and Rebelo, 1990; Rebelo, 1991; Pecorino, 1993; Devereux and Love, 1994; Stokey and Rebelo, 1995) show that income taxes reduce the long-run growth rate while the growth effects of consumption taxes depend on model specification. The extant empirical evidence on the relationship between tax burden and growth is, however, mixed (see Kneller at al., 1999; Widmalm, 2001; Lee and Gordon, 2005; Gemell et al., 2006). These mixed results may be attributed to, among others (5), the limitations of empirical approaches used. One major problem with the cross-country approach commonly employed in the aforementioned studies is that it fails to recognize the short-run dynamic paths that the individual economies may take to their long-run equilibrium (Ojede and Yamarik, 2012). In other words, the existence of a significant relationship in some countries does not necessarily imply that this exists in other countries as well. Such heterogeneity across countries is due to differences in the level of tax authorities' enforcement power, black economy existence, GDP magnitude, internal market size, access to outside markets, labor mobility, and zoning, environmental and other regulation (Mueller, 2003; Karagianni et al., 2012; Ojede and Yamarik, 2012). These differences suggest that the tax structure-growth relationship may be country-specific; therefore, it is necessary to recognize the heterogeneous nature of the countries under investigation.
In recognition of this situation, in a newly emerging strand of literature, researchers have increasingly turned to time-series analysis that enables them to control for the presence of country-specific heterogeneity and cope with the endogeneity problem and/or causal mechanisms. However, most empirical studies dealing with causality between taxation and economic growth rely only to traditional linear Granger causality tests. This means that researchers often neglect a possible nonlinear relationship between these variables because the traditional Granger causality test, designed to detect linear causality, is ineffective in uncovering certain nonlinear relations (Baek and Brock 1992, Hiemstra and Jones 1994). Recent empirical evidence, however, suggests that this relationship is very likely to be nonlinear and the growth effect of taxation is stronger for low average marginal tax rate levels (Bania et al., 2007; Arin et al. 2013, Jaimovich and Rebelo, 2017). In a number of earlier empirical studies, this type of nonlinear behavior has been parsimoniously captured by nonlinear granger causality tests (Karagianni et al., 2012; Tiwari and Mutascu, 2014). Nevertheless, these studies focus exclusively on the tax and growth experience of the USA. In this paper we extend the analysis to a panel of 23 countries with different levels of development and with considerable variability in terms of magnitude of taxation. Furthermore, we follow Tiwari and Mutascu (2014) by applying linear and nonlinear Granger causality tests in investigating the causality between the two variables studied. In particular, besides the linear Granger causality test of Toda and Yamamoto (1995), the nonlinear Granger test proposed by Kyrtsou and Labys (2006) is also applied to capture both linear and nonlinear Granger causality between tax structure and economic growth.
As emphasized by Arachi et al. (2015), the examination of nonlinear relationships between tax structure and economic growth is very relevant topic, and it is motivated by both theoretical and empirical insights (6). Indeed, most economic and financial time series exhibit a nonlinear behavior over time and tend to interact with each other in a nonlinear fashion. This recognition has been confirmed by, among others, the occurrences of severe economic and financial crises (e.g., the 1997-1998 Asian financial crisis, the 2007-2008 US subprime crisis, and the 2008-2009 global financial crisis), wars and other extreme events (e.g., the September 11, 2001 terrorist attack, the Second Gulf war in 2003, the 2006 oil price shock, and the Arab Spring movements), sudden changes in macroeconomic policies, fiscal and economic reforms, increased complexity of financial markets, structural change, and reallocation shocks. All the aforementioned factors may cause unexpected changes in the behavior of economic and financial variables, which particularly induce financial structural breaks, asymmetric responses to shocks, and leverage effects (Ajmi et al. 2013, Atil et al. 2014, Bildirici and Turkmen 2015). Under these circumstances, tax policy and economic growth are likely to exhibit a nonlinear pattern, and their joint dynamics imply a more complex than just a simple and stable relationship (Bertola and Drazen, 1993; Giavazzi et al., 2000; Gupta et al., 2005). In view of this, nonparametric analysis techniques are more suitable because they place direct emphasis on prediction without imposing a linear functional form (Saafi et al. 2015a). The failure in most previous studies to account for asymmetry and nonlinearity between taxation and economic growth may have resulted in incorrect inferences about the existence/non-existence of the taxation--growth relationship.
This study aims to examine whether there is a nonlinear and asymmetric causal relationship between tax burden and growth in 23 OECD countries for the 1970-2014 period. Specifically, this research makes three main contributions. First, it takes a novel approach in examining the countries under investigation, deviating from the common use in the related literature of cross-country and panel regression analysis to the use of separate regression models for each country. Through this approach, we can control for any differences in the financial and economic environment across countries. This is a crucial concern because tax burden varies a great deal across countries. Notwithstanding its significance, there has been limited empirical research that has adopted country-specific time series data to investigate the effect of tax structure on economic growth. Second, this study considers a broader set of tax structure indicators to quantify the impact of taxation on growth and, further, to examine the sensitivity of the results. Finally, as far as the authors are aware, this is the first study to employ the nonlinear causality test of Kyrtsou and Labys (2006) based on the bivariate noisy Mackey-Glass process (hereafter M-G) to explore the nonlinear relationship between tax structure and economic growth. According to Kyrtsou and Labys (2006), Hristu Varsakelis and Kyrtsou (2006), and Hristu Varsakelis and Kyrtsou (2008), the main advantage of the M--G approach for nonlinear causality over simple VAR alternatives is that the nonlinear M--G terms are better able to capture more complex dependent dynamics in a time series. In addition, unlike the standard symmetric methods, the asymmetric (7) version of Kyrtsou and Labys test allows for a potential difference between the effects of positive shocks compared to negative ones. Because of these advantages, the test has recently been applied in several causality studies (for instance, Kyrtsou and Labys, 2006; Hristu Varsekelis and Kyrtsou, 2008; Kumar, 2009; Kumar and Thenmozhi, 2012, Ajmi et al., 2013; Bildirici and Turkmen, 2015; Choudry and Osoble, 2015; Saafi et al. 2015a, 2015b, 2016; Sotoudeh and Worthington, 2016; Jain and Biswal, 2016). It is expected that the analysis in this study will add new insights to the existing empirical literature that will help the policymakers to embrace sound economic policies in order to sustain...
Untangling the causal relationship between tax burden distribution and economic growth in 23 OECD countries: Fresh evidence from linear and non-linear Granger causality.
To continue readingREQUEST YOUR TRIAL
COPYRIGHT TV Trade Media, Inc.
COPYRIGHT GALE, Cengage Learning. All rights reserved.
COPYRIGHT GALE, Cengage Learning. All rights reserved.