5 MARCO EMPÍRICO
5.3 PLANTEAMIENTO DE LA MICROSIMULACIÓN
The first objective was to investigate the effect of changes in total government spending and total tax on output, private consumption, private investment, price level and interest rate. Theoretically, output can remain stable and respond positively or negatively to an increase in total government spending. However, many studies have found a positive response of output to a positive shock in total government spending (e.g., Blanchard & Perotti, 2002; Fatas & Mihov, 2001; Dungey & Fry, 2009). However, negative responses have also been found (e.g. Barry & Devereux, 2003; Giavazzi & Pagano, 1990). In Thailand, Chainakul (2002) and Charoenkittayawut (2001) found that the effect of an increase in total government spending on output was positive. In contrast, Chang et al. (2002) found that output hardly responded to changes in total government spending.
In this study it was found that an increase in total government spending positively affected output. The magnitude of the peak fiscal multiplier was about 0.17 in the second quarter; i.e. an increase in total government spending of 1 baht resulted in an increase in output of 0.17 baht. Price levels and interest rates also positively responded to a positive shock in total government spending, although they negatively responded at the beginning (in the first three quarters), as found in many studies employing U.S. data (e.g. Burriel et al., 2009; Caldara &
A negative response of output to an increase in total tax has generally been found in many studies (e.g. Blanchard & Perotti, 2002; Caldara & Kamps, 2008). However, positive responses from an increase in total tax were also found by Dungey and Fry (2009) and Tubtimtong et al. (2009). In this study it cannot be concluded that the effect of a positive shock in total tax corresponds to the German view of expansionary fiscal contraction (EFC), even though the output response was positive (though insignificant) for the first five quarters after the shock. Because a positive shock in total tax also resulted in an increase in total government spending, it is possible that the positive response of output resulted from the increase in total government spending. It was observed that the total tax multiplier became negative after the fifth quarter (though, again, this was insignificant) after the positive effect of total government spending disappeared. Accordingly, it is possible that the effect of a positive shock in total tax was offset by an increase in total government spending, corresponding to the Keynesian idea that the multiplier of government spending is larger than that from tax (Keiser, 1967). This result also supports Sangsubhan and Basri (2012), who claimed that an expansion in government spending can increase output more than a reduction in tax in Thailand. This is also in line with De Castro and De Cos (2006).
Finally, both private consumption and private investment responded positively to an increase in total government spending, though this was insignificant in the case of private investment, as found in many studies employing Thai data (e.g. Chainakul, 2002; Charoenkittayawut, 2001). However, an increase in total government spending crowded out private investment later (though this was insignificant), as found by Fakthong (2006). In the case of an increase in total tax, private consumption responded positively, but private investment responded negatively to the shock, as found in many studies employing U.S. data (e.g. Badinger, 2006; Burriel et al., 2009). However, the effect of total tax on both private investment and private consumption was insignificant.
The second research objective examined the effects of government spending and tax, by type, on output, price level and interest rate. Many studies using Thai data have found a positive effect of output from an increase in capital spending (e.g. Suwanrada, 1999; Sriboonma, 1999). Some studies investigated the result from both capital and current spending. Yongpitayapong (2004) found that the multipliers of an increase in current spending and capital spending were the same, at 1.33. In contrast, the present study found that the effect from capital spending was larger than the effect from current spending, which is supported by Vorasangasil (2008), with a peak fiscal multiplier for capital spending of about 0.60 in the second quarter, while the peak fiscal multiplier for current spending was about 0.09 in the first
quarter. Moreover, the positive effect of output from an increase in current spending was insignificant.
In the case of taxation shock by type, some studies have found that output responded negatively to an increase in personal income tax shock (e.g. Arin & Koray, 2005, 2006). However, Arin (2003) found that output responded positively to a positive personal income tax shock in Japan. In the present study it could not be concluded that the effect on output of a positive shock to personal income tax corresponded to the EFC viewpoint, although the response of output was positive for 20 quarters after the shock, though insignificant. Since total government spending also responded positively to a positive (though insignificant) shock from personal income tax until the 21st quarter after the shock, it is possible that an increase in output came from an increase in total government spending, or that the effect of an increase in total government spending offset the effect of an increase in personal income tax, as found in the case of total tax shock. This is possibly because the tax base of personal income tax in Thailand is very narrow. Furthermore, a positive shock in personal income tax also raised the price level and interest rate, though insignificantly, consistent with the positive multiplier after the shock.
In terms of the indirect tax and corporate income tax shocks, De Castro (2003) and Arin and Koray (2005) found a negative effect of these positive shocks on output. In contrast, positive multipliers from these shocks were found in Canada by Arin and Koray (2006), and in Canada, Germany and Italy by Arin (2003). In the present study, output responded negatively to a positive shock from indirect tax, while the response of output to corporate income tax was insignificant.
The third research objective was to examine the duration of the effects of fiscal policy shocks. The positive shocks from total government spending and capital spending significantly affected output for about the first six quarters, while the positive output from an increase in current spending remained for a shorter time (until the third quarter), but it was insignificant. This contrasts with the work of Vorasangasil (2008), who found that the effect from current spending was longer. However, in her study only data after the financial crisis were employed, and during that period capital spending was largely focused on improving old facilities, such as streets and buildings (Vorasangasil, 2008). This is why the effect of capital spending in her study disappeared earlier. In the case of tax, an increase in all kinds of tax, except indirect tax, did not have a significant effect on output, while an increase in corporate