MONETARY POLICY REPORT | SPRING 2021 (box 2)
(author: Róbert Ambriško)
Identifying the impacts of fiscal measures on economic activity is an important part of our analytical and forecasting work. To quantify the effect of fiscal policy on economic activity in the CNB forecast, we primarily work with the fiscal impulse calculated using the bottom-up method. One of the two key parameters used to calculate it is the value of the fiscal multiplier.[1]
To calculate the fiscal impulse, we previously (since 2005) used a single fiscal multiplier of 0.6, which we expertly adjusted where necessary. Specifically, in the previous forecast we worked with a fiscal multiplier of 0.6 lowered to 0.3 for changes in labour taxation (in response to the abolition of the super-gross wage and the increase in the tax deductible bonus for individuals).
However, experience has shown that this approach has its limits. The main motivation for enhancing our approach is to take account of the fact that different discretionary measures have different effects on economic activity. We know from the literature[2] and from our own economic research that revenue-side discretionary fiscal measures have less effect on GDP growth than expenditure-side ones with the same budgetary impact. The impacts of fiscal measures also differ across expenditure and revenue categories. Moreover, the structure of discretionary measures plays an important role at times such as the current coronavirus pandemic, when major discretionary changes are being made on the government revenue and expenditure side. To calculate the fiscal impulse, we therefore now use partial fiscal multipliers for the individual discretionary changes on the government revenue and expenditure sides.
We configured the partial fiscal multipliers – which we use for the first time ever in this year’s spring forecast – for the individual government revenue and expenditure categories using values (see Table 1) based on available empirical estimates for the Czech Republic.[3] The settings of the partial multipliers respect the prevailing view in the literature that expenditure multipliers are higher than revenue ones. We assume that the highest government expenditure multiplier effect is generated by additional government investment. This is because, in addition to affecting aggregate demand, government investment improves the production potential of the economy and boosts private investment growth. An increase in government investment thus has a bigger macroeconomic impact on real GDP than an increase in social benefits of the same amount. By contrast, the lowest multiplier effect is produced by a change in capital taxation on the government revenue side. The low multiplier results from the fact that investment responds to the change in capital taxation but the resulting capital stock adjustment response is small due to the additional costs of changing the investment amount.
The partial fiscal multiplier represents the degree to which fiscal discretion in a given revenue or expenditure category affects real GDP.
Table 1 – Values of partial fiscal multipliers for the main government revenue and expenditure categories
| Revenues | Multiplier |
|---|---|
| Social contributions paid by the employer | 0.4 |
| Excise duties | 0.3 |
| Labour taxation | 0.3 |
| Capital taxation | 0.1 |
| Expenditures | Multiplier |
| Government investment | 0.7 |
| Government consumption | 0.6 |
| Other expenditures | 0.5 |
Note: The government consumption multiplier is relevant to the wage component of government consumption entering the calculation of the fiscal impulse. Non-wage consumption expenditure is taken into account in the model forecast through the direct channel of government consumption.
The switch to calculating the fiscal impulse using partial fiscal multipliers means that the estimated effect of fiscal policy on economic activity in the past is slightly lower on average. This is because partial fiscal multipliers, which are mostly lower than the single multiplier of 0.6 used up to now, are applied to discretionary fiscal measures. The methodological change leads to a 0.3 pp decrease in the estimated positive fiscal impulse for 2020. However, the estimated impact of last year’s discretionary budgetary measures on economic activity remains high (at 1.5 pp on real GDP growth), reflecting the strongly expansionary (stabilising) effect of fiscal policy last year.
[1] This topic is analysed in the box The fiscal impulse in CNB forecasts in Monetary Policy Report – Winter 2021.
[2] See, for example, Coenen, G. et al. (2012): Effects of fiscal stimulus in structural models, American Economic Journal, Macroeconomics, 4(1): 22–68.
[3] The revenue-side and government consumption multiplier values are set on the basis of the CNB’s satellite fiscal DSGE model estimated for the Czech Republic (Ambriško, R., 2016: Growth-friendly fiscal strategies for the Czech economy, CERGE-EI Working Papers wp563). The government investment multiplier is set by expert judgement based on theory and available empirical estimates.