Abstract
This article applies newly developed asymmetric impulse response functions and asymmetric variance decompositions to investigate the dynamic relationship between government spending and the GDP at constant prices in Sweden. The estimated results show that an innovation in the government spending does not lead to a significant response in the GDP regardless of whether or not the asymmetric property is taken into account in the estimation of the impulses. The asymmetric variance decompositions also provide support for this conclusion. This might support the view that the Ricardo equivalence theorem is valid in the case of Sweden.
Keywords:
Notes
1 See Barro (Citation1974) for mathematical representation of this theorem.
2 Defining shocks in this way was suggested by Granger and Yoon (Citation2002) that introduced hidden cointegration. However, the authors dealt with only stochastic trends and the deterministic trend parts were excluded.
3 The optimal lag order, k, determined by minimizing an information criterion suggested by Hatemi-J (Citation2003, Citation2008).
4 The source of the data is the Swedish statistical bureau.
5 It should be mentioned that the statistical software component produced by Hatemi-J (Citation2014b) is used to transform the data into cumulative positive and negative components. These components are used in EViews 7 in order to estimate the asymmetric generalized impulses and variance decompositions.