4 Multipliers
Christian Schoder
Remzi Barış Tercioğlu
4.1 List of acronyms
Institutions
OECD Organisation for Economic Co-operation and Development
UNU WIDER World Institute for Development Economics Research
Abbreviations
CPAT Climate Policies Assessment Tool
EET Excise Taxes
EGT Energy Taxes
EVT Total Environmental Taxes
EXT Excise Taxes
GCS Public Consumption
GDP Growth Domestic Product
GIS Public Investment
HIC High-Income Countries
LIC Low-Income Countries
LMIC Lower Middle-Income Countries
MFMod Macro-Fiscal Model
PINE Policy Instruments for the Environment
PIT Personal Income Taxes
TRS Transfers
UMIC Upper Middle-Income Countries
VAT Sales Tax
4.2 Introduction
Carbon pricing has effects on the baseline GDP growth forecasts. For the reference projection of GDP growth, the user can choose between the World Economic Outlook 2020, the World Economic Outlook 2021, and manually entering the growth forecasts. CPAT adjusts these growth forecasts endogenously depending on different carbon pricing and revenue recycling scenarios. Two channels are captured: First, a carbon tax has both direct and indirect effects on GDP. The latter arises when the carbon tax revenues are recycled as a reduction of other taxes and/or as an increase of government spending. These effects are quantified by the fiscal multipliers. Second, the change in GDP affects energy consumption and, therefore, the effective carbon tax revenues. This is captured by the income elasticities of energy demand.
Regarding the first channel, an increase in carbon pricing and the subsequent recycling of the carbon tax revenues into higher government spending and/or lower taxes cause GDP to change with the direction and magnitude depending on the respective spending and tax multipliers. For the fiscal multipliers of energy excise taxes (EET), personal income taxes (PIT), sales taxes (VAT), public investment (GIS), public consumption (GCS), and transfers (TRS), CPAT provides four different sources:
- Income group
- Global
- Manual
- Estimated
Multipliers indicate by how many % GDP responds on impact and in every subsequent year up to a horizon of 10 years to an increase of a fiscal policy instrument by 1% of GDP. The baseline multipliers can be adjusted upwards and downwards by adding/subtracting one empirical standard deviation. This acknowledges the uncertainty around empirical estimates as well as the fact that multipliers tend to be higher during times of economic contraction than expansion. It gives the CPAT user additional flexibility in choosing the appropriate set of multipliers. The present documentation reports the respective methodologies associated with each multiplier source.
When applying fiscal multipliers in CPAT to estimate the GDP effects of a carbon pricing scenario, the following caveats should be noted: First, fiscal multipliers are a link output effects to policy changes in a reduced from. The advantages are that many countries are covered and that values are comparable between countries. Nevertheless, GDP effects of policy interventions may depend on the state of the business cycle and the design of the policy. These are details which multipliers abstract from. Second, income group and global multipliers are derived from an economic model which is empirically estimated. While the model does not impose a strong prior on the multipliers, there is some small remaining influence of the model assumptions on the multipliers. Moreover, the values are averaged over countries because the country-specific multipliers are very volatile. Finally, the estimated multipliers also need to be interpreted with caution. While they are based on a solid dataset and a state-of-the-art methodology, the dataset only includes 75 countries with more than 10 observations. The results have been averaged over various characteristics (see below for details) and extrapolated to countries which are not covered by the data set. Estimating country-specific multipliers is not feasible given the small number of observations for each country.
4.3 Income group specific multipliers
These multipliers have been extracted from the World Bank’s estimated macro-structural model MFMod. Details on MFMod can be found in Burns et al. (2019). The model is estimated for each country and country-specific fiscal multipliers are then computed. To ensure robustness and reduce volatility of multipliers across countries, they are averaged over the countries of an income-group. This leads to four sets of multipliers: One set each for high-income countries (HIC), upper middle-income countries (UMIC), lower middle-income countries (LMIC), and low-income countries (LIC).
MFMod also provides standard errors for these multiplier estimates which are used to adjust the multipliers up- or downwards depending on user preferences.
4.4 Global multipliers
Like above, these multipliers have been extracted from the World Bank’s estimated macro-structural model MFMod and are averaged variants of the income group specific multipliers.
4.5 Manual multipliers
The ‘Manual input tab’ allows the user to enter specific multiplier values for the fiscal instruments mentioned above.
4.6 Estimated multipliers
Estimated multipliers are obtained from a large panel of high-, middle-, and low-income countries.
4.6.1 Methodology for estimating dynamic multipliers using panel data
A thorough discussion of the underlying methodology is provided by Schoder (2022) who exploits the global dataset to study how environmental tax multipliers vary over the business cycle. To obtain dynamic multiplier estimates we employ the local projection method proposed by Jordà (2005) and extended to panel data by Jordà, Schularick, and Taylor (2015) and Jordà, Schularick, and Taylor (2020). As in Dabla-Norris and Lima (2018), we estimate for every horizon \(h = 0,1,\ldots,H - 1\),
\(y_{i,t + h} - y_{i,t - 1} = \alpha_{i,h} + \delta_{t,h} + \text{Δs}_{i,t}\beta_{h} + \Delta x_{i,t}\gamma_{h} + \epsilon_{t,i,h}\)
where \(y_{i,t}\) is the dependent variable. We are interested in explaining 100 times the log of real per-capita GDP in percent. Note that we are estimating cumulative multipliers. Hence, for each horizon we use the change of these variables relative to \(t - 1\) as the dependent variable. \(s_{i,t}\) is the identified shock variable. For each tax instrument considered, it is the cyclically adjusted tax revenue-GDP ratio in percent. Hence, \(\beta_{h}\) has the interpretation of a cumulative multiplier. In particular, \(\beta_{h}\) tells us, under the identifying assumption made and discussed below, by how many percent(age-points) output (employment) increases in \(t + h\) relative to \(t - 1\) if discrete policy increases tax revenues by 1% of GDP. \(x_{i,t}\) is a vector of control variables. \(\alpha_{i,h}\) and \(\delta_{t,h}\) are country and time fixed effects, respectively. To account for heteroskedasticity and autocorrelation we apply the method proposed by Driscoll and Kraay (1998) for estimating a robust covariance matrix of parameters for a panel model. \(\varepsilon_{t,i,h}\) is the error term.
4.6.2 Data set
To create the data set, we employ various sources: The OECD PINE data set provides revenue data for total environmental taxes (EVT) and energy taxes (EGT). From UNU WIDER, we take data on personal income taxes (PIT), excise taxes (EXT), value added taxes (VAT), government consumption (GCS), transfer payments (TRS), public investment (GIS). Data on GDP, employment, GDP deflator, government spending, and population are taken from the World Bank’s World Economic Indicators database. We also use data on total final energy consumption, total final diesel consumption, total final gasoline consumption, diesel and gasoline supply prices, and implicit diesel and gasoline tax rates from a data set compiled to inform CPAT. To remove outliers, we cut off the 1% and 99% percentiles of the changes in the tax revenue-GDP ratios.
4.6.3 Cyclical adjustment of tax revenues and public spending
To address the simultaneous equation bias in the estimates of the tax and spending multipliers which may arise from the feedback of output into tax revenues and spending, we follow the cyclical adjustment approach which assumes that there is a given instrument-specific constant output gap elasticity which can be used to remove the cyclical element from the tax revenues. For instance, the tax revenue-GDP ratios \(\frac{T}{Y}\) have been cyclically adjusted as
\(\frac{T^{*}}{Y^{*}} = \frac{T}{Y}\left( \frac{Y^{*}}{Y} \right)^{\eta_{\text{YT}} - 1}\)
where \(Y^{*}\) is trend GDP obtained from the HP filter of log GDP and \(\eta_{\text{YT}}\) is the output gap elasticity of the tax revenues. Price, Dang, and Botev (2015) estimated the latter, among other, for PIT, VAT and indirect taxes for OECD countries and Dudine and Jalles (2017) for a large sample of high and low-income countries. For countries without elasticities available, we took the averages as the best guess. Note, that there are no output gap elasticities available for environmental taxes. Hence, for EGT, the output-gap elasticities are estimated following the approach proposed by Price, Dang, and Botev (2015) using total energy consumption as a proxy for the tax base. For the spending instruments, the elasticities have been approximated by the values estimated by Price, Dang, and Botev (2015).
4.6.4 Estimation results
For each tax and spending instrument and for various subsamples, this section presents the estimation results which, in the subsequent section, are used to compute country-specific multipliers. Note that in this section tax multipliers are not taken as the negative.
Estimates for the multipliers at horizons larger than eight are restricted to zero when the standard errors become very large, and the sign of the estimate contradicts economic theory. This is to reduce the noise captured by the estimates for larger forecast horizons.
The following tables report the fiscal multipliers from the year of the policy change until 10 years after. The multiplier for each horizon indicates the percentage change of GDP (relative to the year before the policy change) in response to a permanent increase in the policy instrument by 1% of GDP.
4.6.4.1 Pooled panel
GCS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | 0.72 | 1.06 | 0.91 | 0.76 | 0.58 | 0.4 | 0.05 | 0.03 | 0.1 | 0.12 | 0.09 |
TRS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | 0.66 | 0.72 | 0.63 | 0.55 | 0.28 | -0.04 | -0.28 | 0 | 0 | 0 | 0 |
GIS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | 0.47 | 0.76 | 0.76 | 0.52 | 0.48 | 0.49 | 0.16 | 0.07 | 0.51 | 0.55 | 0.53 |
PIT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | -0.19 | -0.57 | -0.86 | -1.26 | -0.87 | -0.77 | -0.73 | -0.94 | -0.87 | -0.71 | -0.53 |
EGT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | -0.58 | -0.76 | -1.24 | -0.5 | -0.37 | -0.64 | -1.21 | -0.77 | -0.23 | 0 | 0 |
EVT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | -0.75 | -1.04 | -0.84 | -0.15 | 0.17 | 0 | -0.4 | -0.03 | 0.25 | 0 | 0 |
EXT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | -0.24 | -0.35 | -0.17 | -0.3 | -0.26 | 0.01 | -0.18 | -0.15 | 0 | 0 | 0 |
VAT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
ABW | -0.47 | -0.62 | -0.32 | -0.55 | -0.83 | -0.88 | -1.1 | -0.9 | -0.9 | -0.9 | -0.9 |
4.6.5 Income levels
GCS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | 0.90 | 1.24 | 1.01 | 0.77 | 0.39 | 0.11 | -0.26 | -0.21 | -0.24 | -0.27 | -0.32 |
LIC & LMIC | 0.44 | 0.76 | 0.72 | 0.73 | 0.94 | 0.95 | 0.67 | 0.52 | 0.76 | 0.86 | 0.85 |
TRS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | 0.90 | 1.03 | 0.86 | 0.73 | 0.43 | 0.10 | -0.12 | -0.23 | -0.34 | -0.53 | -0.78 |
LIC & LMIC | 0.11 | -0.02 | 0.10 | 0.12 | -0.10 | -0.38 | -0.68 | -0.60 | -0.65 | -0.89 | -1.11 |
GIS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | 0.81 | 1.22 | 1.07 | 0.97 | 0.78 | 0.84 | 0.42 | 0.23 | 0.66 | 0.66 | 0.49 |
LIC & LMIC | -0.03 | 0.06 | 0.28 | -0.15 | 0.01 | -0.10 | -0.26 | -0.19 | 0.23 | 0.36 | 0.60 |
PIT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | -0.22 | -0.61 | -0.90 | -1.33 | -0.79 | -0.63 | -0.74 | -1.09 | -1.07 | -0.77 | -0.63 |
LIC & LMIC | 0.00 | -0.30 | -0.62 | -0.74 | -1.50 | -1.79 | -0.64 | 0.31 | 1.05 | -0.04 | 0.64 |
EGT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | -0.85 | -1.03 | -1.24 | -0.20 | 0.28 | 0.00 | -0.83 | -0.08 | 0.93 | 4.04 | 5.78 |
LIC & LMIC | 0.23 | 0.03 | -1.22 | -1.46 | -2.67 | -2.87 | -2.44 | -3.12 | -4.27 | -3.72 | -3.14 |
EVT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | -1.04 | -1.06 | -0.72 | 0.22 | 0.91 | 0.76 | 0.21 | 0.55 | 1.46 | 3.28 | 4.68 |
LIC & LMIC | 0.47 | -0.97 | -1.39 | -1.72 | -3.01 | -3.43 | -3.03 | -2.57 | -5.32 | -4.47 | -3.79 |
EXT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | -0.28 | -0.46 | -0.28 | -0.33 | -0.71 | -0.44 | 0.05 | 0.22 | 0.74 | 1.47 | 1.91 |
LIC & LMIC | -0.19 | -0.15 | 0.02 | -0.23 | 0.66 | 0.95 | -0.65 | -0.93 | -1.48 | -1.97 | -1.12 |
VAT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
HIC & UMIC | -0.42 | -0.37 | -0.02 | 0.02 | 0.20 | 0.05 | -0.29 | -0.64 | -1.12 | -1.47 | -1.49 |
LIC & LMIC | -0.58 | -1.07 | -0.86 | -1.62 | -2.76 | -2.66 | -2.65 | -2.70 | -2.60 | -3.13 | -3.76 |
4.6.6 Regions
GCS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 1.52 | 1.76 | 1.35 | 1.32 | 1.10 | 0.94 | 0.63 | 0.92 | 1.42 | 1.08 | 0.79 |
South East Asia & Pacific | 0.67 | 1.08 | 1.01 | 1.06 | 1.11 | 1.36 | 0.85 | 0.78 | 0.67 | 1.18 | 1.04 |
Africa | 0.25 | 0.29 | -0.07 | -0.50 | -0.43 | -1.12 | -0.94 | -0.90 | -0.99 | -0.54 | -0.56 |
Eastern Europe | 0.68 | 1.43 | 1.30 | 0.89 | 0.64 | 0.40 | -0.07 | 0.16 | 0.39 | 0.14 | 0.16 |
Central Asia & Middle East | 1.58 | 1.80 | 1.58 | 1.77 | 1.30 | 0.95 | 0.48 | 0.27 | 0.54 | -0.11 | 0.35 |
Western Europe & North America | 0.32 | 0.53 | 0.58 | 0.24 | -0.14 | -0.32 | -0.69 | -0.96 | -1.00 | -1.09 | -1.26 |
TRS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 0.79 | 0.92 | 0.72 | 0.66 | 0.56 | 0.25 | -0.04 | -0.05 | -0.18 | -0.53 | -0.67 |
South East Asia & Pacific | 0.81 | 0.97 | 1.12 | 1.27 | 0.83 | 0.38 | 0.08 | -0.44 | -0.33 | -0.33 | -0.23 |
Africa | 0.09 | 0.00 | -0.04 | -0.35 | -0.53 | -0.96 | -1.26 | -1.21 | -1.21 | -1.55 | -1.69 |
Eastern Europe | 0.85 | 0.83 | 0.58 | 0.50 | 0.25 | 0.15 | -0.08 | -0.05 | -0.39 | -0.50 | -1.03 |
Central Asia & Middle East | 0.47 | 0.72 | 0.92 | 1.03 | 1.13 | 1.08 | 0.98 | 1.14 | 1.27 | 0.95 | 0.35 |
Western Europe & North America | 0.70 | 0.67 | 0.43 | 0.00 | -0.68 | -1.40 | -1.64 | -1.76 | -1.78 | -2.05 | -1.87 |
GIS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 0.44 | 0.60 | 0.36 | 0.22 | 0.06 | 0.38 | 0.63 | 0.25 | 0.36 | 0.36 | 0.81 |
South East Asia & Pacific | 0.35 | 0.39 | 0.33 | 0.14 | 0.18 | 0.16 | -0.71 | -0.48 | 0.59 | 1.04 | 1.07 |
Africa | 0.13 | 0.42 | 0.88 | -0.15 | -0.15 | -0.35 | -0.57 | -0.28 | -0.15 | -0.25 | -0.36 |
Eastern Europe | 0.93 | 1.61 | 1.40 | 1.26 | 1.13 | 1.23 | 0.83 | 0.32 | 0.40 | 0.28 | -0.48 |
Central Asia & Middle East | 0.84 | 1.71 | 1.73 | 1.88 | 1.98 | 1.71 | 1.44 | 0.95 | 1.53 | 1.16 | 1.55 |
Western Europe & North America | 0.43 | 0.52 | 0.52 | 0.63 | 0.63 | 0.57 | 0.53 | 0.39 | 0.34 | 0.32 | 0.12 |
PIT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 1.20 | 1.43 | 1.30 | -0.09 | -0.08 | -0.07 | -0.92 | -1.05 | -1.23 | -2.08 | -0.99 |
South East Asia & Pacific | -0.69 | -1.05 | -0.76 | -1.55 | -1.48 | -1.68 | -1.78 | -2.06 | -1.95 | -1.91 | -1.01 |
Africa | -0.53 | -0.40 | 0.08 | -0.32 | -1.44 | -1.39 | -0.66 | -1.34 | -0.92 | -1.35 | -2.42 |
Eastern Europe | 0.16 | -1.65 | -3.27 | -4.30 | -3.67 | -3.78 | -4.12 | -4.24 | -2.78 | -1.10 | -1.82 |
Central Asia & Middle East | -1.42 | -3.05 | -4.38 | -6.20 | -6.64 | -6.78 | -6.36 | -5.23 | -5.08 | -5.01 | -5.59 |
Western Europe & North America | -0.18 | -0.23 | -0.33 | 0.05 | 1.04 | 1.34 | 1.51 | 1.07 | 0.79 | 0.76 | 1.12 |
EGT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 0.37 | 0.53 | -0.58 | -0.43 | -2.44 | -5.66 | -8.87 | -7.66 | -7.57 | -5.90 | -2.52 |
South East Asia & Pacific | 0.46 | 0.19 | -0.45 | -0.34 | 0.21 | -1.01 | -2.00 | -0.20 | -3.51 | -3.22 | -6.00 |
Africa | -0.09 | 0.29 | -0.93 | -1.40 | -2.41 | -1.98 | -1.26 | -2.11 | -3.15 | -0.25 | 1.23 |
Eastern Europe | -0.82 | -0.86 | 0.59 | 2.80 | 4.96 | 7.35 | 6.74 | 6.85 | 7.85 | 11.15 | 13.64 |
Central Asia & Middle East | 1.17 | 3.98 | 2.31 | 3.77 | 2.51 | -2.06 | -3.98 | -7.78 | -5.14 | -4.77 | -4.25 |
Western Europe & North America | -2.06 | -4.91 | -6.36 | -7.42 | -8.73 | -9.23 | -8.95 | -7.28 | -6.08 | -3.91 | -3.83 |
EVT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 0.10 | 0.07 | -0.67 | -0.09 | -1.45 | -4.15 | -6.28 | -4.81 | -4.63 | -3.05 | -0.54 |
South East Asia & Pacific | 0.95 | 0.34 | -0.19 | -0.56 | -0.10 | -0.41 | -0.15 | 1.09 | -1.16 | -0.86 | -2.39 |
Africa | -0.45 | -1.79 | -1.88 | -2.90 | -3.84 | -3.17 | -2.70 | -1.59 | -3.20 | -0.94 | 0.50 |
Eastern Europe | -2.40 | -2.95 | -1.95 | 0.25 | 2.46 | 4.57 | 4.66 | 4.30 | 4.71 | 7.43 | 10.25 |
Central Asia & Middle East | 1.07 | 2.66 | 1.63 | 2.60 | 2.07 | -1.06 | -2.03 | -4.23 | -2.53 | -2.95 | -1.58 |
Western Europe & North America | -0.80 | -1.04 | -0.04 | 0.14 | 0.70 | 0.55 | -0.13 | 0.53 | 1.83 | 2.77 | 1.96 |
EXT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | -0.38 | -1.43 | -2.25 | -2.60 | -3.71 | -4.68 | -5.55 | -7.38 | -7.90 | -7.45 | -7.57 |
South East Asia & Pacific | 0.75 | 0.58 | 0.94 | 1.05 | 0.53 | 0.72 | -0.24 | 0.50 | 0.02 | 0.32 | 0.49 |
Africa | 0.34 | 0.13 | 0.31 | -0.14 | 1.38 | 1.00 | -0.08 | -0.46 | -0.86 | -1.08 | -0.72 |
Eastern Europe | -0.82 | -0.68 | 0.38 | 1.01 | 0.91 | 2.21 | 2.58 | 3.17 | 3.37 | 4.74 | 7.06 |
Central Asia & Middle East | 0.20 | 1.03 | -0.27 | -1.70 | -3.70 | -3.50 | -3.08 | -3.30 | -3.16 | -3.80 | -2.43 |
Western Europe & North America | -1.06 | -1.22 | -0.78 | -0.67 | 0.28 | 0.84 | 1.34 | 1.79 | 2.83 | 3.32 | 3.29 |
VAT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Latin America & Caribbean | 0.44 | 1.03 | 1.78 | 2.05 | 1.71 | 0.66 | 0.50 | 0.76 | 0.58 | -0.34 | 0.20 |
South East Asia & Pacific | 0.48 | 0.06 | -0.59 | -0.96 | -1.39 | -2.00 | -1.35 | -1.14 | -1.72 | -2.49 | -3.01 |
Africa | -0.41 | -0.32 | 0.39 | -0.41 | -0.47 | -0.66 | -1.35 | -1.08 | -1.59 | -1.09 | -1.35 |
Eastern Europe | -0.78 | -1.07 | -0.60 | -1.00 | -1.89 | -1.82 | -1.97 | -2.72 | -2.72 | -3.56 | -3.95 |
Central Asia & Middle East | -0.97 | -1.98 | -3.32 | -3.53 | -2.75 | -0.97 | -1.21 | -1.32 | -2.21 | -2.86 | -4.59 |
Western Europe & North America | -0.64 | -0.64 | -0.10 | 0.00 | 0.10 | -0.14 | -0.51 | -0.93 | -1.19 | -1.28 | -1.20 |
4.6.7 Debt levels
GCS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | 0.81 | 1.14 | 1.09 | 0.92 | 0.53 | 0.30 | 0.11 | 0.10 | 0.12 | 0.11 | 0.08 |
High debt | 0.59 | 0.92 | 0.68 | 0.52 | 0.61 | 0.39 | -0.02 | -0.12 | -0.04 | 0.09 | 0.21 |
TRS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | 0.76 | 0.75 | 0.68 | 0.59 | 0.45 | 0.28 | 0.12 | 0.15 | -0.07 | -0.18 | -0.50 |
High debt | 0.52 | 0.70 | 0.57 | 0.49 | 0.17 | -0.18 | -0.54 | -0.67 | -0.68 | -1.08 | -1.27 |
GIS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | 0.74 | 1.09 | 0.76 | 0.57 | 0.63 | 0.69 | 0.43 | 0.10 | 0.36 | 0.37 | 0.29 |
High debt | 0.18 | 0.31 | 0.62 | 0.29 | 0.28 | 0.12 | -0.10 | -0.05 | 0.38 | 0.42 | 0.54 |
PIT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | -0.10 | -0.63 | -1.00 | -1.56 | -1.27 | -1.12 | -1.10 | -1.19 | -1.02 | -0.92 | -1.01 |
High debt | -0.57 | -0.44 | -0.43 | -0.31 | 0.64 | 0.69 | 0.79 | 0.06 | -0.07 | 0.66 | 1.84 |
EGT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | -0.61 | -0.38 | -0.52 | 0.26 | 0.54 | -0.23 | -1.06 | -0.46 | 0.41 | 3.70 | 5.69 |
High debt | -0.84 | -1.94 | -3.11 | -2.47 | -2.32 | -0.68 | -0.51 | -0.47 | -0.40 | 0.63 | 0.49 |
EVT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | -0.75 | -0.56 | -0.37 | 0.48 | 0.50 | -0.18 | -0.69 | -0.52 | 0.03 | 2.45 | 4.39 |
High debt | -0.95 | -2.45 | -2.17 | -1.93 | -0.49 | 1.10 | 0.98 | 1.74 | 1.75 | 2.10 | 1.46 |
EXT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | -0.05 | -0.06 | 0.30 | 0.16 | -0.55 | -0.28 | -0.18 | 0.11 | 0.51 | 0.95 | 1.7 |
High debt | -0.74 | -1.13 | -1.24 | -1.49 | -0.03 | 0.44 | -0.10 | -0.68 | -0.97 | -0.78 | -0.2 |
VAT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Low debt | -0.54 | -0.81 | -0.63 | -0.98 | -1.35 | -1.27 | -1.54 | -1.86 | -2.16 | -2.68 | -2.86 |
High debt | -0.53 | -0.39 | 0.37 | 0.33 | 0.16 | -0.17 | -0.23 | -0.33 | -0.57 | -0.83 | -1.03 |
4.6.8 Trade openness
GCS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | 0.79 | 1.25 | 1.16 | 0.99 | 0.92 | 1.03 | 0.83 | 0.87 | 0.93 | 0.87 | 0.88 |
Low imports | 0.62 | 0.78 | 0.54 | 0.32 | -0.03 | -0.54 | -1.16 | -1.24 | -1.11 | -1.01 | -1.02 |
TRS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | 0.74 | 0.72 | 0.69 | 0.61 | 0.20 | -0.19 | -0.42 | -0.53 | -0.54 | -0.63 | -1.06 |
Low imports | 0.55 | 0.69 | 0.54 | 0.46 | 0.32 | 0.12 | -0.13 | -0.11 | -0.32 | -0.65 | -0.71 |
GIS multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | 0.42 | 0.74 | 0.61 | 0.52 | 0.41 | 0.33 | -0.14 | -0.13 | 0.34 | 0.39 | 0.30 |
Low imports | 0.58 | 0.81 | 1.07 | 0.54 | 0.65 | 0.78 | 0.66 | 0.42 | 0.82 | 0.88 | 0.97 |
PIT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | -0.73 | -1.66 | -2.55 | -2.61 | -1.70 | -1.89 | -1.61 | -3.13 | -3.19 | -1.68 | -1.98 |
Low imports | 0.04 | -0.09 | -0.16 | -0.67 | -0.52 | -0.28 | -0.35 | -0.06 | 0.03 | -0.33 | 0.00 |
EGT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | 0.86 | 0.88 | 1.20 | 3.04 | 5.22 | 6.43 | 7.25 | 8.76 | 10.49 | 13.98 | 15.78 |
Low imports | -1.09 | -1.35 | -2.12 | -1.81 | -2.45 | -3.37 | -4.58 | -4.57 | -4.61 | -2.58 | -0.98 |
EVT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | -1.32 | -1.56 | -0.85 | 0.81 | 3.54 | 4.59 | 4.96 | 5.27 | 5.81 | 7.63 | 9.07 |
Low imports | -0.51 | -0.83 | -0.84 | -0.55 | -1.22 | -1.94 | -2.71 | -2.33 | -2.21 | -0.67 | 0.66 |
EXT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | -0.60 | -0.23 | 0.28 | 0.20 | 0.46 | 0.98 | 1.70 | 2.65 | 2.88 | 3.81 | 4.97 |
Low imports | -0.03 | -0.41 | -0.43 | -0.58 | -0.67 | -0.54 | -1.23 | -1.72 | -1.53 | -1.48 | -1.09 |
VAT multipliers
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
High imports | -0.50 | -0.9 | -0.66 | -1.21 | -1.97 | -2.02 | -2.28 | -2.58 | -2.66 | -3.06 | -3.47 |
Low imports | -0.45 | -0.4 | -0.05 | -0.04 | 0.05 | -0.04 | -0.25 | -0.48 | -0.92 | -1.35 | -1.49 |
4.6.9 Computation of country specific dynamic multipliers
In the final step, the raw multipliers for the pooled panel and various subsamples which are reported above are processed to compute country-specific dynamic multipliers for up to 10 years ahead. To obtain country-specific multipliers, we take a weighted average over the respective multipliers from each sample/subsample which the country is part of.
The construction of the weights follows this reasoning: The effective sample size decreases with the length of the estimation horizon. Hence, the longer the multiplier horizon, the more likely the estimate is blurred by noise. This is especially true for the sub-samples as they have fewer observations than the full sample in the first place. To account for this, the weights for the contribution to the multiplier average of the pooled estimates start from zero but increase linearly with the multiplier horizon.
Since, the multipliers for the spending categories (GCS, GIS, TRS) are not very volatile across horizons or across subsamples (especially at short horizons), we assign relatively more weight to the subsamples compared to the panel. The remaining four categories (income levels, region, debt level, trade openness) have equal weights:
Weights for the contribution to the multiplier average of the pooled estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0 | 0.083 | 0.166 | 0.25 | 0.333 | 0.416 | 0.5 | 0.583 | 0.666 | 0.75 | 0.833 |
Weights for the contribution to the multiplier average of the income level, region, debt level or trade openness estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.25 | 0.229 | 0.208 | 0.187 | 0.166 | 0.145 | 0.125 | 0.104 | 0.083 | 0.062 | 0.041 |
The weights for PIT and VAT multipliers are:
Weights for the contribution to the multiplier average of the pooled estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.5 | 0.541 | 0.583 | 0.625 | 0.666 | 0.708 | 0.75 | 0.791 | 0.833 | 0.875 | 0.916 |
Weights for the contribution to the multiplier average of the income level, region, debt level or trade openness estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.125 | 0.114 | 0.104 | 0.093 | 0.083 | 0.072 | 0.062 | 0.052 | 0.041 | 0.031 | 0.020 |
EGT multipliers are very volatile across horizons and subsamples. Hence, we assign a higher weight to the pooled results:
Weights for the contribution to the multiplier average of the pooled estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.8 | 0.816 | 0.833 | 0.85 | 0.866 | 0.883 | 0.9 | 0.916 | 0.933 | 0.95 | 0.966 |
Weights for the contribution to the multiplier average of the income level, region, debt level or trade openness estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.05 | 0.045 | 0.041 | 0.037 | 0.033 | 0.029 | 0.025 | 0.020 | 0.016 | 0.012 | 0.008 |
The weights for EVT and EXT multipliers are:
Weights for the contribution to the multiplier average of the pooled estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.75 | 0.771 | 0.791 | 0.812 | 0.833 | 0.854 | 0.875 | 0.895 | 0.916 | 0.937 | 0.958 |
Weights for the contribution to the multiplier average of the income level, region, debt level or trade openness estimates
Unit | H1 | H2 | H3 | H4 | H5 | H6 | H7 | H8 | H9 | H10 | H11 |
---|---|---|---|---|---|---|---|---|---|---|---|
Weights | 0.062 | 0.057 | 0.052 | 0.046 | 0.041 | 0.036 | 0.031 | 0.026 | 0.020 | 0.015 | 0.010 |
Finally, the energy excise tax multipliers are approximated by the means of the EGT and EVT multipliers.
4.7 References
Burns, Andrew, Benoit Campagne, Charl Jooste, David Stephan, Thi Thanh Bui. 2019. “The World Bank Macro-Fiscal Model Technical Description.” World Bank Working Paper 8965.
Dabla-Norris, Era, and Frederico Lima. 2018. “Macroeconomic Effects of Tax Rate and Base Changes: Evidence from Fiscal Consolidations.” IMF Working Papers 18/220. International Monetary Fund.
Driscoll, John C., and Aart C. Kraay. 1998. “Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data.” The Review of Economics and Statistics 80 (4): 549–60.
Dudine, Paolo, and João Tovar Jalles. 2017. “How Buoyant is the Tax System? New Evidence from a Large Heterogeneous Panel.” IMF Working Papers 2017/004. International Monetary Fund.
Jordà, Òscar. 2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review 95 (1): 161–82.
Jordà, Òscar, Moritz Schularick, and Alan M. Taylor. 2015. “Betting the house.” Journal of International Economics 96 (S1): 2–18.
Jordà, Òscar, Moritz Schularick, and Alan M. Taylor. 2020. “The effects of quasi-random monetary experiments.” Journal of Monetary Economics 112 (C):22–40.
Price, Robert, Thai-Thanh Dang, and Jarmila Botev. 2015. “Adjusting fiscal balances for the business cycle: New tax and expenditure elasticity estimates for OECD countries.” OECD Economics Department Working Papers 1275.
Schoder, Christian. 2022. “Regime-Dependent Environmental Tax Multipliers: Evidence from 75 countries.” Journal of Environmental Economics and Policy, forthcoming, 2022. https://doi.org/10.1080/21606544.2022.2089238