COVID-19: How Tax Policy Responses Affected Uganda’s Economy

Abstract

We examined the impact of COVID-19-induced tax policy adjustments on Uganda’s gross domestic product. The analysis is based on a Structural Vector Autoregressive (SVAR) model of Ugandan quarterly data (2009 to 2021). We find that a one standard deviation positive tax policy shock has a negative effect on Uganda’s GDP. Likewise, a one standard deviation positive shock on the consumer price index has a negative effect on the GDP. Thus, we recommend that instead of fiscal provisions in tax cuts and deferrals to micro small and medium enterprises (MSMEs) and households, the government should focus more on raising expenditure on the private sector MSMEs and households, which would stimulate private demand and productivity and sustain domestic revenue collections, particularly from MSMEs. We also recommend the stabilization of food prices, which are the main drivers of the consumer price index in Uganda, to raise GDP growth. Our results provide new insights into the effects of tax policy responses on GDP amidst a global health crisis that has muted economic activities. JEL Classification: B22, C54, E62

Full text

Tanzanian Economic Review, Vol. 13 No. 2, December, 2023: 100–120
©School of Economics, University of Dar es Salaam, 2023  https://doi.org/10.56279/ter.v13i2.111
COVID-19: How Tax Policy Responses
Affected Uganda’s Economy
Ochen Ronald* & Lakuma Paul Corti§
Abstract
We examined the impact of COVID-19-induced tax policy adjustments on Uganda’s gross
domestic product. The analysis is based on a Structural Vector Autoregressive (SVAR)
model of Ugandan quarterly data (2009 to 2021). We find that a one standard deviation
positive tax policy shock has a negative effect on Uganda’s GDP. Likewise, a one standard
deviation positive shock on the consumer price index has a negative effect on the GDP.
Thus, we recommend that instead of fiscal provisions in tax cuts and deferrals to micro
small and medium enterprises (MSMEs) and households, the government should focus
more on raising expenditure on the private sector MSMEs and households, which would
stimulate private demand and productivity and sustain domestic revenue collections,
particularly from MSMEs. We also recommend the stabilization of food prices, which are
the main drivers of the consumer price index in Uganda, to raise GDP growth. Our results
provide new insights into the effects of tax policy responses on GDP amidst a global health
crisis that has muted economic activities.
JEL Classification: B22, C54, E62
Keywords: COVID-19, tax policy responses, gross domestic product, SVAR, Uganda
1. Introduction
Like in many other countries, the COVID-19-induced lockdowns constrained public
revenues in Uganda. Before the COVID-19 pandemic, the tax-to-GDP ratio had
grown from 11.5 percent in the financial year 2017/18 to 12.4 percent in 2018/19, but
it then dropped to 11.6 percent in 2019/20 during the COVID-19 crisis (World Bank,
2020b). Moreover, all major tax heads recorded shortfalls against their respective
targets for the year as collections were affected by the adverse effects of COVID-19
on economic activities (MoFPED, 2021). Nevertheless, this emanated from the
government’s adjustments in the fiscal policy to cope with the adverse effects of the
COVID-19 crisis on the economy. The tax policy measures instituted by the
government at the time were largely tax exemptions and deferrals to households and
private-sector businesses. These included deferred payments to corporate taxes,
employment taxes for firms in the formal sectors, presumptive taxes for micro small-
scale and medium enterprises (MSMEs), and personal income taxes on firms in
severely affected sectors like manufacturing, horticulture, and floriculture and
tourism; waiving of interest on tax arrears; tax deductions on donations and items
for COVID-19 response; and payment of VAT refunds (World Bank, 2020a).
* Centre for Population and Applied Statistics, Makerere University, Makerere University, Kampala,
Uganda: ochenronald@gmail.com
§ Economic Policy Research Centre, Makerere University, Kampala Uganda: plakuma@eprc.org

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
101
The empirical literature on the effects of fiscal policy on Uganda’s GDP is trivial.
That said, several studies elsewhere have examined the impact of fiscal policy on
economic growth (Adegboyo et al., 2021; Agu et al., 2015; M’Amanja et al., 2005;
Blanchard & Perotti, 2002). These studies, however, are not flawless. For example,
Blanchard and Perotti (2002) argued that when government expenditure increased
in the post-war period in the USA due to spending on defence, this induced tax
revenues to increase, hurting output. This phenomenon could contradict Uganda’s
case where tax revenues fell, and government spending increased due to the
COVID-19 fiscal stimulus packages to households and private sector businesses
(World Bank, 2020a; MoFPED, 2021). M’Amanja et al. (2005) used annual data to
study the link between fiscal policy and growth, instead of using quarterly data
which is preferable for studying such relationships. This is because when using
quarterly data, there’s no discretionary within the period response of fiscal policy
to shocks in output, unlike in annual data (Blanchard & Peroti, 2002). Agu et al.
(2015) adopt a literature review approach to study the effect of fiscal policy on
growth, which does not provide an in-depth analysis of the effect of fiscal policy
shocks on GDP. On the other hand, Adegboyo et al. (2021) found that tax revenue
does not affect economic growth in Nigeria, which conflicts with the economic
theory because tax revenue and expenditure are policy levers in fiscal policy for
influencing demand, and therefore output.
Another significant aspect that influenced Uganda’s GDP during the pandemic was
the consumer price index (CPI). Previous studies have found mixed results on this
aspect. For instance, Mahmoud (2013) found a positive relationship between CPI
and economic growth in Mauritania during the period 1990 to 2013. However, Kyo,
(2018) in Japan, found a negative relationship between CPI and economic growth.
Similarly, Mandeya and Ho (2021) found that CPI negatively harms economic
growth in South Africa.
In this paper, we address the theoretical and methodological gaps noted above and
then extend the existing literature by investigating the impact of COVID-19-
induced measures on tax policy on economic growth using the SVAR model, and
quarterly data from 2009 to 2001. Since none of the aforementioned reviewed
studies has attempted to investigate the impact of COVID-19-related fiscal
adjustment on economic growth, this presents a novelty to our study.
As aforementioned, the main objective of this paper is to examine the effects of
COVID-19-induced tax policy adjustments and consumer price index on Uganda’s
gross domestic product. To achieve this objective, it employs the structural vector
auto-regressive (SVAR) model to estimate the dynamic effects of tax policy shocks
on the GDP.
In the follow-up sections, we inspect the recent evolutions in Uganda’s real GDP
growth in section 2; and make a review of literature in section 3. Section 4 presents
the methods and data collection, while the results and discussions from the study
investigations are shown in section 5. Finally, section 6 concludes the study.

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
102
2. Evolution of the Growth of Uganda’s Real Gross Domestic Product
It is important to observe the outlook of the growth of Uganda’s economy before and
during the COVID-19 crisis to understand the rationale for our study. The adverse
effects of the COVID-19 pandemic have negatively affected Uganda’s GDP output.
According to the World Bank (2020a), the fall in Uganda’s real gross domestic
product growth in FY.2019/20 was due to COVID-19-related shocks. Before the
COVID-19 pandemic, Uganda’s GDP had been experiencing a positive trend in GDP
growth until it took a nosedive in FY.2019/20, before rebounding in the third quarter
of 2020/21. For instance, quarterly GDP sharply declined from 8.7 percent in the
second quarter of 2019/20 to 5.8 percent in the fourth quarter of the same period; and
later rebounded to 3 percent in the third quarter of FY.2020/21 (Figure 1).
Figure 1: Uganda’s Quarterly Gross Domestic Product
(FY.2018/19 – FY.2021/22)
Source: Authors’ construction using data from the Uganda Bureau of Statistics
3. Review of Literature
Here we present the theoretical underpinnings and empirical literature relating
fiscal policy to economic growth.
3.1 Theoretical Literature
The Keynesian, classical, and Ricardian schools of thought have significantly
contributed to the relationship between fiscal policy and GDP. Classical theorists
believe fiscal policy can foster sustainable long-term growth through carefully
designed tax systems and spending programmes (Hemming et al., 2002). For
example, the government’s expenditure geared toward enhancing the number of
factors of production positively impacts output growth (Barro & Sala-I-Martin, 1992;
Gerson, 1998). Further, the classical Keynesians expect the effects of fiscal
expansions on growth to be positive and negative for fiscal contractions since they
are traditionally associated with lower growth and recessions (Hemming et al., 2002).
They also argue that the effectiveness of any particular fiscal policy in stimulating
growth depends on the magnitude and sign of the fiscal multipliers (ibid.).
5.5 5.7
7.7 7.0 7.8 8.7
0.9
-5.8
-0.5 -0.3
3.0
13.3
3.5
7.6 5.9
-10.0
-5.0
0.0
5.0
10.0
15.0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3
2018/19 2019/20 2020/21 2021/22
GDP Growth Percentage (%)

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
103
Consequently, from the demand side perspective, the Keynesian view hinges on the
belief that marginal propensity to consume increases with income but at a lower
rate (hence the multiplier effect through increased savings). It holds that the larger
the increase in consumption, the larger the multiplier (Hemming et al., 2002). In
the Keynesian theory, fiscal expansion, therefore, has a multiplier effect on
aggregate demand, and hence on the outcome; implying that the multiplier is
greater than one (i.e., marginal propensity to save is greater than the marginal
propensity to consume); and it is larger for spending increase than for tax
reductions (Hemming et al., 2002).
However, neo-Keynesians have rational thought: they believe that consumers are
rational optimisers of their lifetime average income (i.e., permanent income) and
thus will not change their consumption in response to changes in current income
(Hemming et al., 2002). This, Ricciuti (2021) argues, causes a ‘Ricardian
equivalence’ between taxes and debt, which in its extreme form implies that a
reduction in government’s savings that is due to a tax reduction is entirely counter-
balanced with an increase in private savings; hence the aggregate demand remains
unchanged. Riccardo argued that a temporary increase in government spending
and/or tax reduction would have a stronger positive effect on growth due to a
smaller risk of unsustainable budgetary deficits (ibid.).
On the other hand, from the supply-side perspective, the key factors affecting the
potential effectiveness of short-term fiscal policy are the effects of changes in labour
income taxes on labour supply, and the effects of changes in profit taxes on savings
and investment (Hemming et al., 2002). The neo-classical theorists assume that
markets are efficient, and output growth can only be the result of supply-side
shocks, and should be uncorrelated to aggregate demand (Hemming et al., 2002).
Thus, Lucas and Stokely (1983) argue that under rational expectations, a fully
anticipated fiscal policy targeted at aggregate demand, but not at supply, will not
affect growth either in the short- or long-run.
3.2 Empirical Literature
A review of empirical literature in different countries and at a regional level shows
mixed views on the impact of fiscal policy on economic growth. Several studies in
Sub-Saharan African (SSA) countries exhibit both positive and negative effects of
fiscal policy on economic growth.
Many studies (e.g., Agu et al., 2015; Yusuf & Mohd, 2021; Mohammed & Ehikioya,
2015; Adeolu et al., 2012; Udo et al., 2022; and Tunji et al., 2020) done in Nigeria
using the Ordinary Least Squares (OLS), Error Correction Model, ARDL model,
and Generalized Least Squares techniques have found a positive relationship
between government expenditure and growth; and a positive effect of taxes on
growth. Likewise, a study done in South Africa by Ocran (2009) found a significant
positive effect of government consumption expenditure and tax receipts on
economic growth. Also, a study done in Ghana by Dodz et al. (2014), using the OLS,
found that fiscal policy affected the Ghanaian economy positively. Similarly, Itoro

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
104
and Ekere (2018) used a generalized method of moments (GMM) to analyse the
effects of fiscal and monetary policies on economic growth in a panel of 47 SSA
economies from 1996 to 2016. Their findings show that fiscal and monetary policies
affected economic growth positively in the sub-region.
However, a few studies have found otherwise. Adegboyo et al. (2021), using an
ARDL model, found that fiscal policies stimulate economic growth, while the short-
run results show that fiscal policies have an inconsistent impact on the Nigerian
economy. Salako and Oyeleke (2019), using a VECM, found that government
expenditure positively and significantly impacted the growth of real economic
activities, but the converse was the effect of public revenues on real GDP. Using a
VAR-VECM approach, Bodunrin (2016) found much more unique results: that
fiscal policy had no significant effect on real GDP in Nigeria. In Kenya, M’Amanja
et al. (2005) used the autoregressive distributed lag (ADL) model to investigate the
relationship between various fiscal policy measures on growth in annual data for
the period 1964–2002, and found that contrary to expectations, productive
expenditure had a strong negative effect on growth; while there was no evidence of
distortionary effects on growth of distortionary taxes.
There is an overwhelming evidence of a positive impact of fiscal policy on
economic growth in SSA countries. However, some isolated studies—particularly
in Nigeria—provide inconsistent results on the effects of fiscal policy on economic
growth (see, e.g., Salako & Oyeleke, 2019; Bodunrin, 2016). In addition, to the
best of our knowledge, there is no study in SSA that has used the SVAR and
quarterly data to model fiscal policy shocks on economic growth. Additionally,
none of the aforementioned studies considers COVID-19-induced fiscal policy
adjustments on GDP.
Elsewhere in Asia, like in SSA countries, there are mixed results of a positive
relationship between fiscal policy and economic growth in Indonesia and
Malaysia, for example (Ismal, 2011; Sriyana, 2002). Likewise, using an ARDL,
Rahimi (2021) found a positive and significant effect of fiscal policy on the
economic growth of Afghanistan. Ahmed (2011), on the contrary, found a
negative effect of federal tax on economic growth using OLS and annual data
from 1982 to 2010 to investigate the role of fiscal policy in enhancing the
economic growth of Pakistan.
Separately, many studies done in Europe and the USA (Ritcher et al., 2015;
Stoilova & Todorov, 2021; Mukhtarov et al., 2018; Hamdi & Sbia, 2013) found a
positive relationship between government spending and economic growth, and a
negative link with taxes. However, Hamza and Milo (2021) used a VAR and found
that total public expenditure significantly affects GDP. In the USA, Fu et al. (2003)
found that an increase in the size of the federal government led to slower economic
growth, but tax revenues are the most consistent indicator of fiscal policy. However,
this was contrary to what Blanchard and Perotti (2002) found: a positive and
negative impact of government spending and taxes, respectively, in the USA.

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
105
The bulk of the empirical literature reviewed focuses on the effect of public
expenditure and economic growth, but a few on both fiscal tools. However, by and
large, many studies attempted to examine the impact of fiscal policy on economic
growth, but only a few use SVAR and quarterly data to examine the impact of fiscal
policy on economic growth. To the best of our knowledge, only one study done in
the USA by Blanchard and Perotti (2002) has attempted to use the SVAR to model
fiscal policy effects on output; but there are none on Uganda.
4. Methods and Data
4.1 Methods
4.1.1 Model Specification and Empirical Strategy
Our study is motivated by the work of Blanchard and Perotti (2002), who used the
structural vector autoregressive (VAR) to model the impacts of fiscal policy on
output. To undertake our investigation, we estimated an unrestricted reduced form
vector auto-regressive (VAR) model in levels with a dummy variable for COVID-19
exogenously determined in the model (for dummies in VARs, see Kronborg, 2021).
The VAR model is expressed as follows:
     
  
        󰇛 󰇜 (1)
Where  is a    vector in logarithms of quarterly endogenous variables,
including the gross domestic product, tax shock, domestic tax revenues,
government expenditure and the consumer price index at time t; while  is an
  vector of the exogenous variable, in this case the dummy variable
capturing the COVID-19 period. COVID-19 is a dummy variable denoting 1 at
time t period during the COVID-19 pandemic, and 0 otherwise. The VAR model
appropriates p lags of its endogenous variables; and the matrices and vectors
(
 
 
    󰇜 are coefficients of the estimated VAR.
Further, note that we generated a new variable ‘Tax shock’, emanating from an
interaction of the domestic tax revenue variable with a dummy variable COVID-19
to integrate the period the government of Uganda undertook tax reliefs to private
sector MSMEs and households during the COVID-19 period.
On the other hand, we adopted a VAR model because it identifies the
contemporaneous effects of fiscal policy shocks on GDP, and it is best suited for
studying fiscal policy because budget variables are prone to exogenous fiscal shocks
concerning output (Blanchard & Perotti, 2002). Relatedly, we include government
expenditure in the model because both government expenditure and taxes affect
GDP; and since they are not independent, estimating the effects of one requires
including the other (ibid.).
That said, we first carried out pre-estimation diagnostic tests to check for
stationarity of the variables and the order of integration, preferably [I (1)]. To do
this, we used an ADF unit root test in levels and differences. We then estimated an
unrestricted VAR in levels with 4 lags as the rule of thumb for quarterly data. A

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
106
further check for optimal lags to use in the model selected 4 lags as asterisked by
Akaike Information Criterion (AIC), Hannan Quin (HQ), Schwarz Information
Criterion (SIC), and the Final Prediction Error (FPE) in Table A3. After post-
estimation residual diagnostic tests for serial correlation, normality and stability
were done on the estimated VAR model to determine the significance and stability
of the model.
Second, once the unrestricted reduced form VAR model satisfied the post-
estimation necessary conditions, we then imposed short-run restrictions on the
endogenous variables in the VAR model, thereafter augmenting it into a SVAR
model with short-run restrictions imposed on the contemporaneous relations on the
endogenous variables to estimate their dynamic effects on GDP. Further, the SVAR
model was chosen because it is useful for identifying purely exogenous structural
shocks to obtain the responses of the endogenous variables on GDP. Therefore,
imposing restrictions on the reduced form VAR in (1) augments it into an SVAR,
expressed as follows:
       󰇛 󰇜 (2)
Where  is a non-singular    matrix,  = ,  =  and  = 
denoting the structural shocks in the model uncorrelated with time t.
Finally, we ran accumulated impulse response functions using the Monte Carlo
standard errors with 100 repetitions of a Cholesky decomposition to trace the
contemporaneous effects of fiscal policy shocks on the model.
4.2 Data
The study used quarterly data spanning 12 years from 2009Q1 to 2021Q1, producing
49 observations. The data points of the GDP were inadequate, hence the scope covers
up until the first quarter of 2021. We used quarterly data because it is essential in the
identification of fiscal shocks (Blanchard & Perotti, 2002). Also, in part, the study scope
is crucial because it captures the timeframe when COVID-19-induced tax policy
adjustments were carried out by the government of Uganda. The data on domestic tax
revenue and total government expenditure were obtained from Uganda’s Ministry of
Finance Planning and Economic Development (MoFPED), while the GDP was
obtained from the Uganda Bureau of Statistics (UBoS), and the Consumer Price Index
(CPI) was obtained from the Central Bank of Uganda (CBU) (see Table A2 for details).
The data variables were transformed into natural logarithms. As such, LDTR is the
natural logarithm of domestic tax revenue; LGEXP is the natural logarithm of total
government expenditure; LGDP is the natural logarithm of gross domestic product;
and LCPI is the natural logarithm of the consumer price index.
Further, data on domestic tax revenues were used as a proxy for Uganda’s tax
policy; and government expenditures were sectoral allocations, which also
composed socioeconomic transfers to households during the pandemic. In addition,
the choice of the study variables was informed by economic apriori and empirical
literature from other studies, except for LCPI which introduces novelty to our

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
107
study. For example, most empirical literature (Blanchard & Perotti, 2002, 2016;
Salako & Oyeleke, 2019; Ritcher et al., 2015; and Stoilova & Todorov, 2021),
showed that the expected sign for LGDP is negative when reacting to the effect of
domestic tax revenue shocks, and positive to government expenditure shocks.
4.2.1 Descriptive Statistics of Data
Table 1 shows the descriptive statistics of the main variables used for our study.
Transforming the study variables naturalized them, hence we observe a uniformity
and small variation amongst them. Therefore, during the period 2009q1 to 2021q1,
the GDP averaged 10 percent, tax revenues averaged 7 percent, government
expenditure averaged 8 percent, while consumer price index averaged 6 percent.
The Jacque-Bera confirms the normality of all the pre-estimated variables at a 5
percent level of significance. Additionally, the graphical exposition of these
variables is shown in Figure A1.
Table 1: Descriptive Summary Statistics of the Series (2009q1–2021q1)
LGDP
LDTR
LGEXP
LCPI
Mean
10.16581
7.110268
8.118630
6.061241
Median
10.16806
7.123347
8.070092
6.082354
Maximum
10.49208
8.048651
9.052109
6.332327
Minimum
9.858525
6.000272
7.178418
5.627011
Std. Dev.
0.165841
0.548160
0.506578
0.209572
Skewness
0.098007
-0.279629
0.013253
-0.630710
Kurtosis
2.028425
1.965704
2.267695
2.226506
Jarque-Bera
2.005692
2.822680
1.096321
4.470173
Probability
0.366834
0.243816
0.578012
0.106983
Sum
498.1245
348.4031
397.8128
297.0008
SumSq. Dev.
1.320156
14.42302
12.31782
2.108180
Observations
49
49
49
49
4.2.2 Correlation Matrix
We explored the direction and the strength of the linear relationship between the
pairs of our data variables used in the study. The correlation matrix presented in
Table 2 indicates that the study variables are significantly positive and highly
correlated with each other, and the off-diagonal elements are one.
Table 2: Correlation Matrix
LGDP
LDTR
LGEXP
LCPI
LGDP
1.000000
0.883240
0.886421
0.889754
LDTR
0.883240
1.000000
0.946213
0.966775
LGEXP
0.886421
0.946213
1.000000
0.928556
LCPI
0.889754
0.966775
0.928556
1.000000
4.2.3 Unit Root Test
We conducted an Augmented Dickey -Fuller (ADF) unit root test of our study
variables in levels and at first differences with the Schwarz Info Criterion (SIC)

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
108
for automatic lag length selection and 4 maximum lags to examine the
stationarity properties of the data. The results shown in Table 3 indicate that all
the study variables are stationary after the first difference ; therefore, they are
integrated into order one I(1), satisfying the necessary condition to proceed with
the VAR model.
Table 3: Unit Root Results for the Variables Using
the Augmented Dickey-Fuller Test
Variable
Unit Roots in Levels
Unit Roots in 1st Difference
Order of
Integration
Constant
(t-Statistic)
Constant,
Linear Trend
(t-Statistic)
Constant
(t-Statistic)
Constant,
Linear Trend
(t-Statistic)
LGDP
-0.220850
-3.207235*
-4.054043***
-3.985715**
I(1)
LGEXP
-0.477677
-6.801951**
-6.761247**
-6.680750**
I(1)
LDTR
-1.888404
-0.847562
-3.288854**
-3.815888**
I(1)
LCPI
-2.492759
-2.037716
-3.065802**
-4.747312*
I(1)
Note: *** p<0.01, ** p<0.05, * p<0.1
5. Results and Discussions
In this section, we present the estimated SVAR model and the impulse response
functions established amongst the variables’ interactions, and later the robustness
checks of the VAR model.
5.1 The Estimated SVAR Model Results
As per the results presented in Table 4, the coefficients [C(2), C(3) and C(11)] for
the short-run restrictions imposed on the SVAR model show that tax policy shocks
and the consumer price index have a negative effect on GDP. On the other hand,
government expenditure has a positive effect on GDP. We further investigate these
results using accumulated impulse response functions on the SVAR model.
Table 4: Estimated Structural Vector Auto-regressive Model
Coefficient
Std. Error
z-Statistic
Prob.
C(2)
-0.101640
0.167206
-0.607872
0.5433
C(4)
0.072510
0.455350
0.159240
0.8735
C(5)
0.739181
0.404308
1.828264
0.0675
C(7)
0.443204
0.244425
1.813251
0.0698
C(8)
1.181872
0.224879
5.255593
0.0000
C(11)
-0.073987
0.041711
-1.773821
0.0761
C(1)
0.032283
0.003403
9.486833
0.0000
C(3)
0.036210
0.003817
9.486833
0.0000
Log-likelihood
435.0265
Estimated A Matrix:
1.000000
0.000000
0.000000
0.000000
0.000000
0.101640
1.000000
0.000000
0.000000
0.000000
-0.072510
-0.739181
1.000000
0.000000
0.000000
-0.443204
-1.181872
0.009462
1.000000
0.000000
0.073987
0.042042
0.003655
-0.027836
1.000000

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
109
Estimated B Matrix:
0.032283
0.000000
0.000000
0.000000
0.000000
0.000000
0.036210
0.000000
0.000000
0.000000
0.000000
0.000000
0.098208
0.000000
0.000000
0.000000
0.000000
0.000000
0.052702
0.000000
0.000000
0.000000
0.000000
0.000000
0.008682
5.2 Impulse Response Results
The impulse responses of the Cholesky decomposition over the study scope are
presented in Figures 2, 3 and 4. Figure 2 shows the short-run tax policy shock
impulse responses to other endogenous variables in the model. Crucial to note is
that a 1 standard deviation positive tax shock has a 0 effect on GDP in periods 1
and 2; but then a more pronounced negative effect on GDP is realised in the
subsequent periods.
This result is consistent with the findings of several studies (e.g., Blanchard &
Perroti, 1999; Mukhtarov et al., 2018; Hamdi & Sbia , 2013; Ritcher et al 2015;
Stoilova & Todorov, 2021; Ahmed, 2011), which argue that an increase in tax
revenues leads to a reduction in GDP. However, in the aforementioned studies,
taxes were raised leading to a reduction in GDP contrary to Uganda’s case
where tax reliefs provided to firms and households reduced its GDP. This could
have been due to inertia in the effectiveness of the tax reliefs to stimulate
Uganda’s economy that was already grappling with the adverse effects of the
COVID-19 pandemic.
The result also contradicts the Ricardian theory, which argues that a temporary
increase in tax cuts has a stronger positive effect on growth due to a smaller risk
of unsustainable budgetary deficits (Ricciuti, 2021). More so, the contradiction
with our study could be attributed to the tax shock dominating the tax relief to
private sector MSMEs and households in the short-run. Also, our results
vindicate the outlook of Uganda’s GDP shown in Figure 1. The study findings
further shows that a 1 standard deviation positive tax shock results have a
positive effect on government expenditure and domestic tax revenues, which are
persistent throughout the periods.
The Cholesky decomposition impulse responses in Figure 3 show that a 1
standard deviation positive shock in the consumer price index leads to a negative
effect on GDP. This outcome is expected because of the inverse relationship
between prices and GDP. Then, a positive shock in the consumer price index leads
to a negative effect on government expenditure: this was also expected because a
rise in prices would induce the government to cut expenditure to curb a rise in
the cost of living.

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
110
-.08
-.04
.00
.04
.08
1 2 3 4 5 6 7 8 9 10
Accumulated Res ponse of LGDP to TAXSHOCK
-.02
.00
.02
.04
.06
.08
1 2 3 4 5 6 7 8 9 10
Accumulated Res pons e of TAXSHOCK to TAXSHOCK
-.10
-.05
.00
.05
.10
.15
.20
1 2 3 4 5 6 7 8 9 10
Accumulated Respons e of LGEXP to TAXSHOCK
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Accumulated Res ponse of LDTR to TAXSHOCK
-.04
-.02
.00
.02
.04
.06
.08
1 2 3 4 5 6 7 8 9 10
Accumulated Res ponse of LCPI to TAXSHOCK
Accumulated Response to Cholesky One S.D. Innovations ± 2 S.E.
Figure 2: Impulse Responses of Tax Policy Shocks
Using the Cholesky decomposition
Source: Estimated Structural VAR Model Impulse Response Functions
This finding is in tandem with those of Kyo (2018), and Mandeya and Ho (2021),
who found a negative relationship between CPI and economic growth in Japan
and South Africa, respectively. However, it differs from that of Mahmoud (2015),
who found a positive relationship between CPI and economic growth in
Mauritania.

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
111
-.10
-.05
.00
.05
.10
1 2 3 4 5 6 7 8 9 10
Accumulated Response of LGDP to LCPI
-.06
-.04
-.02
.00
.02
.04
.06
1 2 3 4 5 6 7 8 9 10
Accumulated Response of TAXSHOCK to LCPI
-.3
-.2
-.1
.0
.1
1 2 3 4 5 6 7 8 9 10
Accumulated Res ponse of LGEXP to LCPI
-.10
-.05
.00
.05
.10
.15
.20
1 2 3 4 5 6 7 8 9 10
Accumulated Response of LDTR to LCPI
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Accumulated Response of LCPI to LCPI
Accumulated Response to Cholesky One S.D. Innovations ± 2 S.E.
Figure 3: Impulse Responses to Consumer Price Index Shocks
Using the Cholesky Decomposition
Source: Estimated Structural VAR Model Impulse Response Functions.
From Figure 4, a 1 standard deviation positive shock on government expenditure
has a substantial negative effect on GDP throughout the periods; but the consumer
price index responds positively to a 1 standard deviation positive shock on
government expenditure; even though an upward trend is observed from period 3
onwards. However, a 1 standard deviation positive shock on government
expenditure does not affect domestic tax revenues since it is close to zero
throughout the periods; but a positive shock in the government expenditure has a
downward negative effect on government expenditure.

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
112
-.3
-.2
-.1
.0
.1
.2
1 2 3 4 5 6 7 8 9 10
Accumulated Response of LGDP to LGEXP
-.12
-.08
-.04
.00
.04
.08
1 2 3 4 5 6 7 8 9 10
Accumulated Respons e of TAXSHOCK to LGEXP
-.3
-.2
-.1
.0
.1
.2
1 2 3 4 5 6 7 8 9 10
Accumulated Res ponse of LGEXP to LGEXP
-.10
-.05
.00
.05
.10
.15
1 2 3 4 5 6 7 8 9 10
Accumulated Response of LDTR to LGEXP
-.04
.00
.04
.08
.12
.16
1 2 3 4 5 6 7 8 9 10
Accumulated Respons e of LCPI to LGEXP
Accumulated Response to Cholesky One S.D. Innovations ± 2 S.E.
Figure 4: Impulse Responses to Government Expenditure Shocks
Using the Cholesky decomposition
Source: Estimated Structural VAR Model Impulse Response Functions
5.3 Robustness Checks
We checked the robustness of our unrestricted reduced VAR model to ensure that
the estimated residuals are white noise and satisfy the classical regression model
assumptions, and the results are reliable and valid. Specifically, we carried out
residual tests on the estimated model, including serial correlation,
heteroskedasticity, normality, and the stability of the model. We used the LM test

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
113
to check the null hypothesis of no serial correlation of the VAR residuals, and the
results indicate that we fail to reject the null hypothesis at a 5% level of
significance: thus, there is no serial correlation in the residuals (Table 4). We also
confirmed homoscedastic residuals at a 5% level of significance in Table A5 after
failing to the null hypothesis of no heteroskedasticity. The multivariate test for
normality using the orthogonalized Cholesky (Lutkephol), developed by Jarque and
Bera (1987), confirms that the null hypothesis of residuals is multivariate normal
at a 5 percent level of significance (Table A6). In addition, the graphical exposition
of the VAR residuals is normally distributed around the zero mean as realized in
Figure A2. Lastly, the AR inverse roots test in Figure A3 shows that all the roots
lie within the unit circle, hence the model is stable.
6. Conclusion
The empirical literature on the effects of fiscal policy on Uganda’s GDP is little.
More so, globally, research on the impact of tax policy responses on economic
growth during the COVID-19 crisis is still novel; yet it is essential for policy
purposes. This study sought to bridge this gap by examining the impact of the
COVID-19-induced tax policy adjustments on Uganda’s gross domestic product
from 2009q1 to 2021q1. We conclude that the COVID-19-induced tax policy
adjustments on Uganda’s economy had a negative effect on GDP in the short-run.
This could be due to inertia in the reaction of the tax policy adjustments of tax cuts
and deferrals in the forms of relief to the private sector MSMEs and households in
raising GDP; hence, the tax shock overtook the tax relief, resulting in a
spontaneous reduction in the growth of Uganda’s economy. In that regard, we
recommend that to stimulate GDP growth in a crisis like COVID-19, instead of
fiscal provisions in tax cuts and deferrals to MSMEs and households, the
government should raise the expenditures of the private sector MSMEs and
households. This would, in turn, stimulate private demand and productivity to
sustain domestic revenue collections, particularly from MSMEs. We also found that
the consumer price index has a negative effect on Uganda’s GDP, thus we
recommend the stabilization of food prices in the country since this is the main
driver of the consumer price index in Uganda. This is critical to cool down the CPI
upward pressures and raise GDP growth.
References
Adegboyo, O. S., Keji, S. A. & Fasina, T. O. (2021). The Impact of Government Policies on
Nigeria Economic Growth (Case of Fiscal, Monetary and Trade Policies. Future Business
Journal, 7(1): 59 https://doi.org/10.1186/s43093–021–00104–6.
Adeolu, M., James A. S. & Bolarinwa. (2012). Fiscal/Monetary Policy and Economic Growth in
Nigeria: A Theoretical Exploration. International Journal of Academic Research in
Economics and Management Sciences, 1(5): (September) 2012. www.hrmars.com.

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
114
Agu, U. S., Okwo, I. M., Ugwunta, O. D. & Idike, A. (2015). Fiscal Policy and Economic
Growth in Nigeria: Emphasis on Various Components of Public Expenditure. SAGE
Open, 5(4): 1–12. https://doi.org/10.1177/2158244015610171.
Ahmed, Z. (2011). Fiscal Policy and Economic Growth in Pakistan. International Journal of
Research in Commerce, Economics & Management. 1(5). www.ijrcm.org.in.
Barro, J. Robert, and Xavier Sala-I-Martin. (1992). Convergence. Journal of Political
Economy, 223–251. https://doi.org/10.1086/261816.
Blanchard, O. & Roberto, P. (2002). An Empirical Characterization of the Dynamic Effects
of Changes in Government Spending and Taxes on Output. Quarterly Journal of
Economics, 117(4) (November): 2002: 1329–1368. https://doi.org/ 10.1162/ 0033553
02320935043.
Bodunrin, O. S. (2016). The Impact of Fiscal and Monetary Policy on Nigerian Economic
Growth. Munich Personal RePEc Archive. MPRA Paper(92811. https://mpra.ub.uni-
muenchen.de/92811/.
Dodzi, H.K. E. & Enu, P. (2014). The Effect of Fiscal Policy and Monetary Policy on Ghana’s
Economic Growth: Which Policy Is More Potent? International Journal of Empirical
Finance. 3(2): 61–75. http://rassweb.org/ admin/ pages/ResearchPapers/ Paper% 203_
1497044433.pdf.
Fu, D., Taylor, L. L. & Yücel, K. M. (2003). Federal Reserve Bank of Dallas Fiscal Policy and
Growth Federal Reserve Bank of Dallas. Working Paper 0301. http://www.dallasfed.
org/assets/documents/research/papers/2003/wp0301.pdf.
Gerson, P. (1998). The Impact of Fiscal Policy Variables on Output Growth. International
Monetary Fund, IMF Working Paper Wp9801. https://www.imf.org/ external/ pubs/ ft/wp/
wp9801.pdf.
Hamza, B. & Milo, P. (2021). Fiscal Policy and Economic Growth: Some Evidence from
Kosovo. Journal of Governance and Regulation, 10(4): 130–136. https://doi.org/ 10.22495/
jgrv10i4art11.
Hemming, R., Kell, M. & Mahfouz, S. (2002). The Effectiveness of Fiscal Policy in Stimulating
Economic Activity - a Review of the Literature. IMF Working Paper wp02208.
https://www.imf.org/external/pubs/ft/wp/2002/wp02208.pdf.
Ismal, R. (2011). Assessing Economic Growth and Fiscal Policy in Indonesia. Journal of
Economics and Business. (Issue 1). XIV-2011(1): 53–71. https://www.u-picardie.fr/eastwest/
fichiers/art94.pdf.
Itoro, U-A. & Ekere, D. (2018). Fiscal Policy, Monetary Policy and Economic Growth in Sub-
Saharan Africa Fiscal Policy, Monetary Policy and Economic Growth in Sub-Sahara
Africa. Munich Personal RePEc Archive. MPRA Paper (91950. https://mpra.ub.uni-
muenchen.de/91950/.
Jarque, C.M. & Bera, A.K. (1987. A Test of Normality of Observations and Regression Residuals.
International Statistical Review, 55: 163–172. http://dx.doi.org/ 10.2307/ 1403192.
Kronborg, F. A. (2021). Estimating Foreign Shocks in VAR Model. Working Paper, Danish
Research Institute for Economic Analysis and Modelling (DREAM) www. dreamgruppen.dk.
Kyo, K. (2018). The Dynamic Relationship between Economic Growth and Inflation in
Japan. Open Journal of Social Sciences, 6: 20–32. https://doi.org/10.4236/jss.2018.63003.

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
115
Lucas, E. R. & Nancy, L. S. (1983). Optimal Fiscal and Monetary Policy in an Economy
without Capital. Journal of Monetary Economics, 12(1): 52–93. https://doi.org/ 10.1016/
0304–3932(83)90049–1.
Mandeya, T.M.S. & Ho, Shinyo. (2021). Inflation, Inflation Uncertainty and the Economic
Growth Nexus: An Impact Study of South Africa. Elsevier B.V. https://doi.org/
10.1016/j.mex.2021.101501.
Mahmoud, M.O.L. (2015). Consumer Price Index and Economic Growth: A Case Study of
Mauritania 1990–2013. Asian Journal of Empirical Research, 5(2)2015: 16–23.
www.aessweb.com.
M’Amanja, D. & Morrissey, O. (2005). Fiscal Policy and Economic Growth in Kenya. Credit
Research Paper No.05/06. The University of Nottingham, Center for Research in
Economic Development and International Trade (CREDIT). http://hdl.handle.net/10419/
65474www.econstor.eu.
Mukhtarov, S., Gasimov, I. & Rustamov, U. (2018). Evaluation of Fiscal Policy Impact on
Economic Growth: The Case of Azerbaijan ASERC. Journal of Social-Economic Studies,
1(1): 82–90. www.ajses.az.
Mohammed, I. & Ehikioya, L. I. (2015). Behavioral Pattern of Fiscal Policy Variables and
Effects on Economic Growth: An Econometric Exposition on Nigeria. International
Journal of Academic Research in Business and Social Sciences, 5(2): ISSN: 2222–6990.
https://doi.org/10.6007/ijarbss/v5–i2/1488.
Ministry of Finance Planning & Economic Development (MoFPED). (2021). Half Year
Macroeconomic & Fiscal Performance Report Financial Year 2020/21. https://www.
finance. go.ug /publication/ half-year-macroeconomic-fiscal-performance-report-financial-
year-202122. (Accessed July 26: 2022).
Ocran, M. K. (2009). Fiscal Policy and Economic Growth in South Africa. Journal of
Economic Studies, 38(5): 604–618 602(5). https://doi.org/10.1108/01443581111161841.
Rahimi, E. (2021). The Impact of Fiscal Policy on Economic Growth of Afghanistan. International
Journal of Research and Analytical Reviews, 8(1): (March) 2021. www.ijrar.org.
Ricciuti, R. (2021). Assessing Ricardian Equivalence. Journal of Economic Surveys, 17(1):
55–78. http://dx.doi.org/10.1111/1467–6419.00188.
Richter, C. & Paparas, D. (2015). Fiscal Policy and Economic Growth, Empirical Evidence
in European Union. Working Paper 2015.06. International Network for Economic
Research. http://dx.doi.org/10.13140/RG.2.1.1268.1045.
Salako, G. & Oyeleke, O. J. (2019). Fiscal Policy and Growth of Real Economic Activities in
Nigeria (1980–2016). Asian Journal of Economics and Empirical Research, 6(2): 108–
112. https://doi.org/10.20448/journal.501.2019.62.108.112.
Sriyana, J. (2002). Fiscal Policy and Economic Growth: An Empirical Evidence in Malaysia
and Indonesia. Jurnal Ekonomi Pembangunan. JEP 7(2): 143–155. https://doi.org/ 10.
20885/ejem.v7i2.647.
Stoilova, D. & Todorov, I. (2021). Fiscal Policy and Economic Growth: Evidence from Central
and Eastern Europe. Journal of Tax Reform, 7(2): 146–159. https://doi.org/ 10.
15826/jtr.2021.7.2.095.

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
116
Tunji, T. S., Nurudeen, A. A., Festus, F. A. & Oyeyemi, G. O. (2020). Fiscal Policy
Sustainability and Economic Growth of Nigeria. Solid State Technology, 63(3): 2020.
www.solidstatetechnology.us.
Udo, E. G., Akpan, E. & Akpan, O. M. (2022). Fiscal Policy and Economic Growth: An
Empirical Assessment in Fiscal Regimes in Nigeria (1970–2019). International Journal
of Social Science and Human Research, 05(2): (February) 2022: Page(612–624. https://
doi.org/10.47191/ijsshr/v5–i2–29.
World Bank. (2020a). Uganda – COVID-19 Economic Crisis and Recovery Development
Policy Financing, by Tihomir Stucka and Racheal K. Sebudde. Report No: PGD203:
Washington, DC: World Bank. https://documents1.worldbank. org/curated/ en/60932
1593741824637/pdf/Uganda-COVID-19–Economic-Crisis-and-Recovery-Development-
Policy-Financing.pdf, (Accessed July 26: 2022).
World Bank. (2020b). Digital Solutions in a Time of Crisis: Uganda Economic Update,
Fifteenth Edition, by Richard Walker, Tihomir Stucka, and Qurusm Qasim. Washington,
DC: World Bank. https://openknowledge.worldbank.org/handle/10986/34078 License: CC
BY 3.0 IGO. (Accessed July 26: 2022).
Yusuf, A. & Mohd, S. (2021). Asymmetric Impact of Fiscal Policy Variables on Economic Growth
in Nigeria. Journal of Sustainable Finance and Investment. https://doi.org/ 10. 1080/
20430795.2021.1927388.

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
117
Appendices
5.6
6.0
6.4
6.8
7.2
7.6
8.0
8.4
09 10 11 12 13 14 15 16 17 18 19 20 21
LDTR
9.8
9.9
10.0
10.1
10.2
10.3
10.4
10.5
09 10 11 12 13 14 15 16 17 18 19 20 21
LGDP
6.8
7.2
7.6
8.0
8.4
8.8
9.2
09 10 11 12 13 14 15 16 17 18 19 20 21
LGEXP
Figure A1: Graphical Exposition of the Series
.
5.6
5.7
5.8
5.9
6.0
6.1
6.2
6.3
6.4
09 10 11 12 13 14 15 16 17 18 19 20 21
LCPI

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
118
Table A2: Study Variables Definitions and Sources
Variables
Description
Sources
Gross Domestic
Product (LGDP)
GDP at constant prices measured in
Uganda Shillings Billions.
Ministry of Finance
Planning and Economic
Development.
Domestic Tax
Revenues (LDTR)
Domestic tax revenues collected by the
tax body, are measured in Uganda
Shillings in Billions.
Ministry of Finance
Planning and Economic
Development.
Government
Expenditure
(LGEXP)
Total expenditure of government on the
different programmes in the various
sector of the economy.
Ministry of Finance
Planning and Economic
Development
Consumer Price
Index (LCPI)
Consumer Price Index, (2009/10=100)
on all items index (weight = 1000).
Bank of Uganda.
Table A3: Lag Order Selection Criteria of the Estimated VAR
Lag
LogL
LR
FPE
AIC
SC
HQ
0
199.4480
NA
1.52e-10
-8.419911
-8.018430
-8.270243
1
338.6675
235.1262
9.59e-13
-13.49633
-12.09115
-12.97249
2
407.9304
101.5856
1.41e-13
-15.46357
-13.05469
-14.56557
3
460.3827
65.27392
4.76e-14
-16.68367
-13.27109
-15.41150
4
510.5330
51.26477*
2.02e-14*
-17.80147*
-13.38518*
-16.15512*
Table A4: VAR residual serial correlation using LM test.
Null Hypothesis: no serial correlation
Lags
LM-Stat
Prob
1
35.47868
0.0799
2
20.90621
0.6978
3
34.34716
0.1007
4
24.27515
0.5035
5
19.15749
0.7895
Table A5: VAR Residual White Heteroskedasticity Tests:
No Cross Terms
Chi-sq
df
Prob.
625.4842
615
0.3759
Table A6: VAR Residual Normality Tests, Orthogonalized: Cholesky (Lutkephol)
Hypothesis: Residuals are Multivariate Normal
Component
Skewness
Chi-sq
Prob.
Kurtosis
Chi-
sq
Prob.
Jacque-
Bera
Prob.
1
-0.30222
0.685027
0.4079
3.26304
0.13
0.718
0.8148
0.6654
2
0.55953
2.348061
0.1254
4.18163
2.62
0.105
4.9660
0.0835
3
-0.32418
0.788239
0.3746
2.44565
0.58
0.447
1.3644
0.5055
4
-0.48207
1.742972
0.1868
3.75963
1.08
0.298
2.8249
0.2435
5
-0.14900
0.166519
0.6832
2.88453
0.02
0.874
0.1915
0.9087
Joint
5.730818
0.3333
4.44
0.489
10.161
0.4264

COVID-19: How Tax Policy Responses Affected Uganda’s Economy
Tanzanian Economic Review, Volume 13, Number 2, 2023
119
-.08
-.06
-.04
-.02
.00
.02
.04
.06
10 11 12 13 14 15 16 17 18 19 20 21
LGDP Residuals
-.08
-.04
.00
.04
.08
.12
10 11 12 13 14 15 16 17 18 19 20 21
TAXSHOCK Residuals
-.15
-.10
-.05
.00
.05
.10
.15
10 11 12 13 14 15 16 17 18 19 20 21
LGEXP Residuals
-.15
-.10
-.05
.00
.05
.10
10 11 12 13 14 15 16 17 18 19 20 21
LDTR Res iduals
-.020
-.015
-.010
-.005
.000
.005
.010
.015
10 11 12 13 14 15 16 17 18 19 20 21
LCPI Residuals
Figure A2: Graphical Exposition of the Estimated VAR Residuals

Ochen Ronald & Lakuma Paul Corti
Tanzanian Economic Review, Volume 13, Number 2, 2023
120
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
Inverse Roots of AR Characteristic Polynomial
Figure A3: Model Stability Using the AR Inverse Roots Test