Tracing the Link between Government Size and Growth: The Role of Public Sector Quality

The influence of government size on per capita GDP growth has attracted much interest from scholars during the past two decades. However, despite all the work conducted, there is no consensus among researchers on the importance of government size in affecting economic growth. Some authors find negative effects (Barro, 1991, De la Fuente, 1997, or Fölster and Henrekson, 2001) while others point out that the relationship is positive or not significant (Caselli et al., 1996, or Agell et al., 1997 and 2006). Another strand of the growth literature underlines the importance of economic and political institutions, where there is widespread agreement that good institutions are a precondition for economic growth (Hall and Jones, 1999, Acemoglu et al., 2001, or Rodrik et al., 2004). By taking into consideration both branches of the literature, this paper focuses on the empirical link between government size and growth and assesses whether this relationship depends on the quality of public sector institutions. There are good reasons to expect the presence of heterogeneity in the effect of public sector on economic growth. One could suppose that public spending is negative for the real economy when the public administration is inefficient, corrupt or pursues the private interests of politicians and officials. By contrast, when quality standards in the public administration are high (government responds to citizen demands, works diligently and satisfactorily fulfils its functions), government size does not necessarily hinder economic growth. We believe that the lack of consensus about the effect of government size may result –among other things- from the presence of heterogeneity in its relationship with economic growth. Consequently, we try to contribute to this literature by showing that the effect of government size on growth depends on the quality of public sector institutions.In first place, we estimate cross-section growth regressions for the period 1981-2005 using ordinary least squares (OLS) to analyze the effect of government size on long-term growth. To investigate the presence of heterogeneity, we estimate an interaction model where we include as regressors –along with other common variables in growth models- government size, bureaucratic quality and the product of both variables. In order to make a correct specification, we include all constitutive terms of the interaction in the equation. In addition, following the advice of Brambor et al. (2006), we conduct a conditional interpretation of the coefficient of government size, reporting the coefficients and standard errors of the marginal effect of government size for each value of bureaucratic quality. The results are subjected to several robustness checks, such as a wide range of control variables, the presence of outliers, the use of an alternative indicator of institutional quality and the use of two-stage least squares (2SLS). In second place, with the aim of overcoming the shortcomings of the cross-section approach as well as exploiting the time variation of the data, we estimate panel regressions with the system GMM estimator developed by Arellano and Bover (1995) and Blundell and Bond (1998). Likewise, we pay particular attention to the marginal effects of government size for each value of bureaucratic quality and carry out several robustness checks.This paper shows evidence of strong heterogeneity in the relationship between government size and growth, depending on the quality of public sector institutions. Government size affects growth negatively when the quality of the public sector is low, but not when public sector quality is high. In the latter case the evidence shows that there is no significant effect (either positive or negative). This finding holds both in cross-section and panel data analyses and is robust to a large number of robustness checks. The observed heterogeneity in the effect of government size on the real economy appears in line with previous work by Angelopoulos et al. (2008) and Rajkumar and Swaroop (2008). Our results contribute to shed light on the debate over whether government size is an obstacle to growth. In conclusion, the central message of this paper is that government size can be an obstacle to economic growth when public sector institutions are weak, but is neutral when bureaucratic quality is high.