Does the Chilean government smooth taxes? A tax-smoothing model with revenue collection from a natural resource

Abstract

This article uses a present value test to examine whether the Chilean government has smoothed taxes optimally since 1973. An important portion of the Chilean government's revenue is the result of royalties it earns from the extraction of copper. An appropriate test for tax-smoothing therefore must recognize that this part of the government's revenue is not completely under its control. The results provide strong evidence for tax smoothing when royalties from copper are treated as not being under government control, but only weak evidence if they are treated as if they are under government control.

Notes

¹Vergara (2002) and Crispi and Vega (2003) show that under this rule Chilean fiscal policy is sustainable.

²Huang and Lin (1993), Olekalns (1996) and Cashin et al. (1998) are examples of tests for tax smoothing that do not take into account the royalties from a natural resource.

³This assumes that there is no tax-tilting effect (see e.g. Ghosh, 1995b; Olekalns, 1996). Pasten (2006, Chapter 3) fails to reject the hypothesis that there is no tax tilting in Chile for the period 1973 to 2002.

⁴We use only one lag in the VAR because (as pointed out below) in the estimated model additional lags are not statistically significant. Cover and Pasten (2009) show that it is straightforward to develop a test similar to that developed below if there is more than one lag in the VAR.

⁵Figures between 1973 and 2003 are available from GFS-IMF but no further revision has been made for years before 1992. The Chilean agency is no longer publishing data prior to 1992. This article uses the GFS-IMF data series because it is the only source long enough to yield reliable estimates.

⁶Results of these can be found in Cover and Pasten (2009).