The Distributional Impacts of Fiscal Consolidation in Uganda

Abstract

While Uganda is considered to be at low risk of debt distress, the stagnant tax effort and large planned capital expenditures might significantly alter this position. This paper employs the dynamic stochastic general equilibrium (DSGE) model to examine tax design issues that arise in addressing debt increases. The results suggest that Uganda may improve it debt position by permanently increasing tax rates by 5 percentage point. However, an
increase of consumption tax rates (VAT and Excise) by this magnitude to meet debt reduction is found to be relatively more distortionary affecting consumption, especially for the poor households, in both the short and long run leading to large temporary reductions in the gross domestic product (GDP).

Full text

International Relations and Diplomacy, October 2020, Vol. 8, No. 10, 438-459
doi: 10.17265/2328-2134/2020.10.003
The Distributional Impacts of Fiscal Consolidation in Uganda
Corti Paul Lakuma
Makerere University, Kampala, Uganda
Joseph Mawejje
World Bank Group, Kampala, Uganda
Musa Mayanja Lwanga
Uganda Bankers Association, Kampala, Uganda
Ezra Munyambonera
Makerere University, Kampala, Uganda

While Uganda is considered to be at low risk of debt distress, the stagnant tax effort and large planned capital
expenditures might significantly alter this position. This paper employs the Dynamic Stochastic General
Equilibrium (DSGE) model to examine tax design issues that arise in addressing debt increases. The results suggest
that Uganda may improve it debt position by permanently increasing tax rates by 5% point. However, an increase
of consumption tax rates (Value Added Tax (VAT) and Excise) by this magnitude to meet debt reduction is found
to be relatively more distortionary affecting consumption, especially for the poor households, in both the short and
long run leading to large temporary reductions in the Gross Domestic Product (GDP).
Keywords: Equity, Inequality, Debt, Taxes, Fiscal Consolidation, Public Finance, DSGE, Macro-Fiscal Policy
Introduction
Uganda’s public debt has increased significantly on average from 21% Gross Domestic Product (GDP) in
2010 to about 38% in 2016—the highest rate since debt forgiveness through the Heavily Indebted Poor Country
(HIPC) and the Multi-lateral Debt Relief Initiative (MDRI) in 2006 (Lakuma et al., 2017). Public debt has
increased in large part due to expenditure on the National Development Plan (NDP), stagnation of the ratio of
tax revenues to GDP, large and ever growing informal sector that is hard to tax, a fall of global commodity
prices and the resultant lower export revenues, the lagged return to investments by capital project during the
NDP, and reduction in donor financing.1 In the absence of significant fiscal consolidation measures,
debt-to-GDP ratio is likely to remain high over the medium term. Against this backdrop, the government of
Uganda is expected to undertake policies to reduce debt through tax revenue mobilization and recurrent
expenditure restraint. These policies are well spelled out in the various Medium Term Debt Strategies (MDTS),
Corti Paul Lakuma, research fellow, Macroeconomics Department, Economic Policy Research Centre, Makerere University,
Kampala, Uganda.
Joseph Mawejje, economist, Macro Fiscal Management Practice, World Bank Group, Kampala, Uganda.
Musa Mayanja Lwanga, head of research, Uganda Bankers Association, Kampala, Uganda.
Ezra Munyambonera, senior research fellow, Macroeconomics Department, Economic Policy Research Centre, Makerere
University, Kampala, Uganda.
1 Uganda’s long term development strategies are articulated in the Vision 2040, which is implemented through five-year NDP.
Presently, Uganda is implementing its second NDP.
DAVID PUBLISHIN G
D

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
439
the 2013 Uganda Public Debt Management Framework and the 2016 Charter of Fiscal Discipline.2 The
primary objective of fiscal consolidation is to maintain Uganda’s low risk of debt distress. Other coincidental
factors motivating the planned fiscal consolidation are Uganda’s attempt to align it real sector performance to
other East African Community (EAC) member states in respect of the East Africa Monetary Union (EAMU)
convergence criteria (EAC, 2013).3
However, in a developing country like Uganda with no broad social protection program and a small public
pension system, a cut back on social spending on education and health programs can have dire consequences
among them increased inequality. In this regard, this paper employs the Dynamic Stochastic General
Equilibrium (DSGE) model to conduct an ex-ante regulatory impact assessment of Uganda’s fiscal
consolidation efforts from 2016 to 2040. The model examines the impact of a 5% point increase in the income
tax, corporate tax, and consumption taxes (Value Added Tax (VAT) and excise tax) on household welfare, the
real sector and in achieving a sustained reduction in public debt burdens and fiscal deficits.4
The model is calibrated so as to reflect key characteristics of Uganda’s economy including debt, poverty
level, and output. However, we must note that this paper examines the impact of a tax rate increase fiscal
consolidation policy stance only. In this regard, an extensive evaluation of the expenditure cuts policy stance is
beyond the scope of this study. The focus on taxation stance is motivated by the urgency to increase Uganda’s
domestic tax effort and the need to identify a tax rate reforms that strengthen the fiscal position while being, if
not growth-promoting, at least minimally distortionary and growth-retarding, while respecting equity concerns.5
The rest of the paper is organized as follows. Section 2 presents an overview of Uganda’s macro and fiscal
position. Section 3 presents a review of related literature on fiscal consolidation. Section 4 offers the empirical
methodology. Section 5 illustrates how the model was calibrated. Section 6 discusses the results prior to
conclusions in Section 7.
Overview of Uganda’s Macro and Fiscal Position
In the 1990’s, Uganda’s debt had peaked to unsustainable levels such that the economy did not have the
capacity to meet its debt obligations. Fortunately, Uganda benefited from two waves of, earlier mentioned, debt
relief that eased Uganda’s debt service obligations. The first was the HIPC Initiative in 1998 and 2000; the
second was MDRI in 2006 (MoFPED, 2013).6
Uganda’s debt remains sustainable; the present value of public debt-to-GDP ratio is projected to peak at
about 41% in 2021, well below the benchmark level of 56% associated with heightened public debt
vulnerabilities for medium performers (IMF, 2015). However, the relatively short average maturity of domestic
debt combined with a low revenue base continues to be a matter of concern for Uganda’s fiscal policy. Interest
payments on debt are rising fast recorded at 12.6% in the 2017/2018 (MoFPED, 2016b). Forecasts suggest that
2 Charter for Budget Responsibility (“the Charter”) guides the government in the formulation and implementation of fiscal policy
and policy for the management of the national debt. The charter also specifies indicators that will enable Uganda to attain the
convergence criteria in the year 2021.
3 The EAMU Convergence criteria requires that all EAC member states maintain inflation at 5%, fiscal deficit ceiling of 6% of
GDP, Debt-to-GDP ratio of 50% and tax-to-GDP ratio of 25%.
4 The value of 5% point is influenced by the highest the current Uganda’s tax rate can increase without exacerbating the excess
burden and the loss in consumer surplus rate.
5 See Trabandt, M., & Uhlig, H. (2011), “The Laffer Curve Revisited”, Journal of Monetary Economics, 58(4), 305-327.
doi:10.1016/j.jmoneco.2011.07.00 for a discussion on tax revenue maximizing tax rate. Otherwise known as the “laffer curve”.
6 The savings from debt forgiveness where largely expended on the Poverty Action Fund (see Lakuma & Lwanga, 2017).

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
440
interest payments will account for 16% of domestic revenue in 2020 (MoFPED, 2016b). This is above the
thresholds (15%) set in the Medium Term Debt Strategy.
The Pace of Fiscal Consolidation
The over-arching need is for substantial fiscal consolidation, as the planned scaling up of public
investment approaches completion, to both reduce levels of public debt and provide space to address social
expenditure and domestic arrears. Figure 1 sets out an adjustment path for Uganda and illustrates the scope of
the debt reduction challenge. The negative fiscal impulse suggests a tighter fiscal policy from 2017. This policy
stance is consistent with the Charter of Fiscal Responsibility and criteria for the EAMU by 2021 (MoFPED,
2016a).
Figure 1. Fiscal consolidation in Uganda (Source: Authors’ own calculation).
Note. Dotted line represents forecast.
Composition of Fiscal Consolidation
Table 1 sets out how reductions in borrowing between 2017/2018 and 2021/2022 are to be achieved. The
deficit is to be reduced through a combination of increase in taxes, cuts to social sector spending and a squeeze
on development spending. For example, the government plans to raise an extra UGX 2,112 billion in taxes,
UGX 1,232 billion squeeze in development expenditure, and UGX 527.8 billion cut in agriculture and social
sector expenditure in 2018/2019 to reduce deficits and debt. While the planned fiscal consolidation may
increase the tax burden on households and have an impact on the efficiency of the productive sectors, it will
-2
-1
0
1
2
3
4
Primary Balance
Cyclically-Adjusted
Primary Balance
Fiscal Impulse

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
441
reduce the budget deficit and debt significantly. MoFPED (2016b) suggests that changes in the fiscal position
will be largely driven by reforms in the tax system.7
Table 1
Consolidation Plan: Change in Deficit From 2017/2018 (UGX Billions)
2017/2018
2018/2019
2019/2020
2020/2021
2021/2022
Taxes
-1,948.0
-2,112.0
-2,081.0
-2,376.0
-2,860.0
Non tax revenue
-73.0
-25.0
-69.0
-74.0
-93.9
Development expenditure
2,208.0
-1,232.0
-104.0
288.0
3,602.0
Agriculture and social sector spending*
-606.3
-527.8
277.0
560.4
2,304.9
Net lending and investment
266.0
-92.0
-118.0
319.0
212.0
Notes. *Social sectors are education, health, water and environment and social development and programs earmarked under the
Poverty Action Fund. Source: Authors’ own calculation with data MoFPED (2016b).
Medium Term Risk
Fostering underlying growth in the second NDP (2015-2020) remains imperative to manage the debt
dynamics as suggested in Figure 2. Growth is critical to debt sustainability, reducing the relative scale of
taxation needed (De Mooij & Keen, 2013). In addition, tax burden on households and firms tend to change
much less as cash receipts tend to move more in line with the size of the economy. However, Figure 2 shows an
underlying poor growth performance during the first NDP (2010-2015). This is likely a reflection of the less
than potential demand, particularly for Uganda’s exports as made evident in Uganda’s sustained trade deficits
(IMF, 2016).
Figure 2. Uganda’s GDP growth (actual and potential billions shillings). Source: Authors’ own calculation.
7 Other strategies are efficiency in tax administration, which will require investments on technology and human resources.
0
1
2
3
4
5
6
7
8
9
10
Actual Real GDP (Billions)
Potential GDP (HP Filter)

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
442
Figure 3 reveals that real interest rate is high due to uncertainty about the future outlook and inflationary
effect of fiscal policy. This has often necessitated Bank of Uganda to tighten monetary policy, with negative
effect on private sector credit. The uncertainty has also presented a challenge to the coordination of monetary
and fiscal policy.
Figure 3. Real interest rate (computed using expected inflation rate). Source: Authors’ own calculations.
A Review of Related Literature
After the financial crises of 2008, many countries, particularly Organization for Economic Co-operation
and Development (OECD) members, have adopted and are implementing fiscal consolidation through increase
in tax rates and expenditure cuts to reduce debt (OECD, 2012). This is largely because high debt weighs
negatively on output growth, limits room for accommodation of future negative shocks and adds to fiscal
challenges resulting from future government expenditure (Rawdanowicz, 2014).
In addition, a high debt-to-GDP ratio plays an important role in the determination of government bonds
and treasury bills rates. Typically, the returns to bonds and treasury bills increase with risk perception (De
Grauwe & Ji, 2013). In this case, the perception of risk by the financial market participants and behavioural
factors such as perception biases and extrapolative expectations may influence the bond rates (Rawdanowicz,
2014). Moreover, high bond rates have the potential to pass through to borrowing costs increasing credit cost
for the private sector and heightening the risk of sovereign default (Rawdanowicz, 2014). This phenomenon
-2
0
2
3
5
6
8
9
11
12
14
15
17
18
20
21
23
24
26
27
29
30
Jan-10
Jun-10
Nov-10
Apr-11
Sep-11
Feb-12
Jul-12
Dec-12
May-13
Oct-13
Mar-14
Aug-14
Jan-15
Jun-15
Nov-15
Apr-16
Sep-16
Real Policy Rate
Real Bond Rate

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
443
can lead to sudden escalation of debt as was evidenced during the Greece economic crisis (Arghyrou &
Kontonikas, 2011).
Also, carrying out a fiscal consolidation to reduce debt is associated with reduction in fixed capital. For
example, while the South African government reduced its debt/GDP ratio from almost 50% to 27% between
1994 and 2008, this reduction was accompanied by a significant decrease in government’s fixed capital/GDP
ratio from 90% to 55% (Burger, Siebrits, & Calitz, 2016). This suggests that fiscal sustainability may not
necessarily improve the government’s balance sheet. For this reason, Auerbach and Gorodnichenko (2012)
argue for postponing consolidation as a large frontloaded adjustment can reduce GDP growth with negative
fallout for the fiscal situation. Such effects are more likely when output and unemployment gaps are large and
credit constraints are binding (Lakuma, Marty, & Kuteesa, 2016).
Moreover, fiscal consolidation has been associated with an increase in poverty (Smeeding, 2000) and an
increase in income inequality (Bova, Kinda, Woo, & Zhang, 2013). In some OECD countries, increased
taxation and expenditure cuts have led to record unemployment, economic stagnation, collapsing financial
sector (OECD, 2012). Some governments have lost elections due to implementing fiscal consolidation (OECD,
2012).
This brings to fore the capacity gaps in developing countries, such as Uganda, in carrying out fiscal
consolidation. The question then is what impact will increase taxation on incomes and consumption, and
expenditure cuts have on a developing country. It should be noted that, even without fiscal consolidation,
income inequality has risen in Uganda and is higher today than it was 25 years ago (Ssewanyana & Kasirye,
2012). Nearly 10% of households continue to live in chronic poverty (Ssewanyana & Kasirye, 2012).
However, Uganda has also achieved a sustained decline in poverty to 19.7% over the years, which
suggests that a rise in income inequality can be avoided (World Bank, 2016). As such, the rise in inequality is
relevant to this study granted its’ negative consequences on growth, efficiency, and welfare. Nevertheless, the
impact of inequality on growth and of growth on inequality is unclear since there are equalizing
and un-equalizing effects during the economic cycle (Hoeller & Pisu, 2014). Indeed, Kuznets’ (1955)
pioneering findings suggest that a country goes through an inverted U-curve of economic growth and economic
inequality.
The last set of issues concern the increment in the tax rates itself as a policy tool for achieving reduction in
debt and a substantial shift in the tax effort. Hutton et al. (2014) argue that there is no room for tax rate
increment in developing countries with low collection efficiency and, as such, much of tax mobilization efforts
should concentrate on expanding the number of tax payers. However, the cost and the benefits of a tax rate
increment have not been empirically examined in the developing world, in particular Uganda. The aim of
Section 6 of the paper is therefore to assess the merits of income and consumption taxes in achieving
consolidation, growth, and fairness.
The Model
As earlier mentioned, this paper models the impact of a 5% point increase in the tax rate on a closed
economy with sectors for households (non-poor (Ricardian) and poor (non-Ricardian)), firms and government
(fiscal authority, social sector (education, pension, and health)), and the monetary authority. The modelling
implemented in this work is a variant of the work by Costa Junior and Sampaio (2014). The model includes

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
444
poor household to reflect the inability of households to quickly adjust their consumption and investments in
event of a shock.8
Households
The household sector is composed of two types of representative agents: non-poor (Ricardian) and poor
(non-Ricardian). The non-poor household represents households living above the poverty line. Typically, these
kinds of households are economically active and pay a variant of taxes other than consumption taxes. These
households form (1-ω) of the total population, while the poor households form the remaining proportion of the
population. The non-poor household is able to maximise its intertemporal utility by choosing consumption,
savings, investment, and leisure. For saving, the household can choose between two different savings
instruments—physical capital and government bonds. Also, with the disposable income after payment of taxes,
the non-poor household can purchase consumer goods, capital goods, and/or, government bonds. On the other
hand, the poor household just allocates its income in acquisition of consumer goods.
Non-poor (Ricardian) households (R): Taxpayers. Relying on the behaviour described about the
households, the non-poor household chooses how much to consume, how to work, and how to acquire financial
and physical assets to maximize the discounted stream of the expected utility as expressed in Equation (1):9
 ,
1
1
1+
1
∞
=0
(1)
Subject to their budget constraint as expressed in Equation (2):
󰇛1 + Τ󰇜,+++1
=󰇧1Τ
Τ󰇨+󰇧1Τ
󰇨+
(2)
And in relation to the law of motion of capital as expressed in Equation (3):
+1 =󰇛1󰇜+
(3)
Where  is the expectations operator,  󰇛0,1󰇜 is the intertemporal discount factor,  is the
consumption of non-poor household, L is the labour,  is the intertemporal consumption shock,  is the
shock on the labour supply,  is the marginal disutilty of labour, and  is the coefficient of relative risk
aversion.
In the budget constraint,  is the general price level,  is the investment,  is the government bond
maturing in one period,  is the rate of return on government bond (basic interest rate),  is the wage , 
8 Besides the inclusion of poor households, this model has two other frictions: monopolistic competition and staggered pricing a
la Calvo. The latter friction aims to avoid the model to have a very fast adjustment in relation to shocks, a factor noticed in
empirical evidence.
9 The most common utility function to represent the choices of Family Representative is the utility function of constant of
Constant Relative Risk Aversion (CRRA) (Gali, 2008; Lim & McNelis, 2008; Clarida et al., 2008; Gali & Monaceli, 2005;
Christoffel & Kuester, 2000; Christoffel et al., 2009; Ravenna & Walsh, 2006). The other common parameterizations for the
utility function in the literature, examples logarithmic utility function, 󰇛,󰇜=ln +
ln󰇛10󰇜 (Hansen, 1985); and
utility function that would be a combination of logarithmic and of the CRRA, 󰇛,󰇜=ln󰇛󰇜
1+
1+ (Gertler & Karadi,
2011). A utility function must have certain characteristics: > 0 and < 0. This means that consumption and labor have a
positive and negative effect, respectively, over the happiness of the households. On the other hand,  < 0 and  < 0,
indicating the utility function is concave. This represents that if the consumption increases the utility level also increases, but in a
smaller and smaller proportion. Another assumption regarding the utility function says that this function is additionally separable
in time. This assumption allows speaking of an instantaneous utility function, wherein the agent receives utility solely from
consumption that performs at a given moment in time.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
445
is the return to capital, K is the stock of capital,  and  are the stochastic components of labour income
and capital/corporation income respectively, while Τ,Τ,Τ,Τ represent the statistic components of the tax
on consumption, labour, capita/corporate, and income respectively. This paper adopts the convention that  is
the nominal bond issued in (t-1) and matured in t. Then, +1 and +1 are decided in t.
The non-poor household purchases of consumer goods 󰇛󰇜 and investment goods 󰇛󰇜 at the price level
󰇛󰇜, also buys or sells government bonds (B) maturing in one period. These bonds pay a risk-free , which is
also controlled by the monetary authority.
This kind of household pays three types of taxes (consumption tax, income tax on labour, and income tax
on capital/corporate tax) and also contributes to social security.10 The household income comes from three
sources: labour income, which depends on the level of nominal wages 󰇛󰇜; return on capital rental to firms,
which is a function of the rate of return to capital 󰇛󰇜; and income from government bonds acquired in the
previous period.
To solve the problem of the Ricardian household, a Lagrangian function is used:
= 
∞
=0 ,
1
1
1+
1 + 
󰇩󰇛1 + Τ󰇜,++1 󰇛1󰇜++1
󰇧1Τ
Τ󰇨
󰇧1Τ
󰇨󰇪
(4)
The first order conditions associated with the choices of ,,,+1 and +1 are respectively:

,=,
󰇛1 + Τ󰇜= 0
(5)

=
+󰇧1Τ
Τ󰇨= 0
(6)

+1
=󰇛1 + Τ󰇜++1 󰇩󰇛1󰇜+1 󰇛1 + Τ󰇜++1 󰇧1Τ
󰇨󰇪= 0
(7)

+1
=
++1 = 0
(8)
From Equation (5):
=,

󰇛1 + Τ󰇜
(9)
Substituting the Equation (9) into (6), it results in the equation of labour supply:

,
󰇯󰇛1 + Τ󰇜
󰇡1Τ
Τ󰇢󰇰=

(10)
Substituting Equation (9) in Equations (7) and (8), we obtain the Euler equations:
10 Consumption tax is composed of VAT and Excise tax.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
446
,
 =+1
,+1

+1󰇛1 + Τ󰇜󰇩󰇛1󰇜+1󰇛1 + Τ󰇜++1 󰇧1
+1
󰇨󰇪
(11)
,

=+1
,+1

+1
(12)
Non-Ricardian households (NR): Poor household. Poor Households have a simpler behavior.11
Because they do not maximize their intertemporal utility, their consumption is limited to the value government
social expenditure ( ..). Under this hypothesis:
󰇛1 + Τ󰇜,= ..
(13)
Aggregate consumption. The aggregate consumption of this work follows the functional form 󰇛=
󰇛1󰇜+󰇜 very common in this type of literature (Coenen & Straub, 2004).12 13
Thus, aggregate consumption of the individual non-poor households and poor-households is performed as
follows:
=󰇛1󰇜,+,
(14)
Shocks to related households. There are two shocks related to non-poor household behaviour: the shock
in intertemporal preferences 󰇛󰇜 and the shock on labor supply 󰇛󰇜, while the first affects the choice of
intertemporal consumption and the second affects labour supply and determination of nominal wages. The
shock  was included to capture changes in valuation between the present and future which the literature on
intertemporal behavior suggested as key to the understanding of aggregate fluctuations (Primiceri, Schaumburg,
& Tambalotti, 2006). Additionally the shock  was added to model changes in labor supply that Hall (1997)
and Chari, Kehoe, and McGrattan (2007) identified as responsible for major changes in employment over the
business cycle. There are two other shocks in the stochastic components of the taxes on labour income 󰇛󰇜
and on capital income 󰇛󰇜. These shocks were included to characterize the stochastic component related to
these three types of taxes, which are the objects of this paper. Thus, the movement rules of such shocks are
presented in Equations (15) to (18):
log =󰇛1󰇜log 
+1
+,
(15)
log =󰇛1󰇜log 
+1
+,
(16)
log =󰇛1󰇜log 
+1
+,
(17)
log =󰇛1󰇜log 
+1
+,
(18)
Where ,,,,,,, are exogenous shocks, and ,,, are autoregressive components, of
the intertemporal consumption shock, of the shock on labor supply, of the shock of taxes on consumption, of
the shock of taxes on labor income, and of the shock of taxes on capital income, respectively.
Firms
The productive sector of the economy in this paper is divided into two subsectors: firm producers of
finished goods (retail); and firm producers of intermediate goods (wholesale). The wholesale sector is formed
by great number of firms, each producing a different good according to the structure of monopoly competition.
11 Generally, the DSGE literature treats the non-Ricardian agent/poor household as an individual without capacity to maximize
the intertemporal utility due to liquidity conditions. In this work, the assumption is that this type of agent does not maximize its
utility due to poverty.
12 =,
1
0=󰇛1󰇜,+, given that agents belonging to the same group are identical.
13 Also see Bosca et al., 2010; Gali et al., 2007; Itawa, 2009; Fulanetto, 2007; Dallari, 2012; Mayer et al., 2010; Stahler &
Thomas, 2011; Swarbrick, 2012; Motta & Trelli, 2010; Diaz, 2012; Colciago, 2011; Mayer & Stahler, 2009; Forni et al., 2009.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
447
In the retail industry, there is a single firm that aggregates intermediate goods in a single good (Y) that will be
consumed by economic agents. Besides these features, it should be mentioned that the markets for productive
factors follow a structure of perfect competition.
Firm producers of finished goods (retail). First, it is necessary to define the aggregator behaviour of the
production function. The finished good is produced by a single firm that operates in perfect competition. For
this purpose, the firm combines a continuum of intermediate goods and aggregates them into a single finished
good using the following technology:
=󰇧 ,
1

1
0󰇨
1
(19)
Y is aggregate output,  is the intermediate product j, and  is the elasticity of substitution between
intermediate goods. The form adopted to aggregate the assets is called a Dixit-Stiglitz aggregator (Dixit &
Stiglitz, 1977).
As mentioned, the finished goods producer is in perfect competition and maximizes its profit by using the
technology of Equation (19), whereas the prices of intermediate goods are given. Therefore, the problem of the
retail firm is:
max
, ,,
1
0
(20)
Substituting Equation (19) into (20):
max
,󰇧 ,
1

1
0󰇨
1,,
1
0
The first order condition for each intermediate good j is:
󰇧 ,
1

1
0󰇨
11,
1
1,= 0
,=󰇧
,󰇨
(21)
Equation (21) demonstrates that the demand for intermediate good j is a decreasing function of its relative
price and increasing in relation to the aggregate output of the economy.
The general price level is obtained by substituting Equation (21) in (19):
=󰇱 󰇩󰇧
, 󰇨󰇪1

1
0󰇲
1
=󰇧 ,
1

1
0󰇨
1
(22)
Firm producers of intermediate goods (wholesalers). The wholesaler firms solve the problem in two
steps. In the first step, firm take as given the prices of production factors: wages (W) and return to capital (R).
They determine the quantities of those inputs that will minimize their costs. In the second stage, firms

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
448
determine the optimal price of good j and they determine the quantity that will be produced in accordance with
this price.
First step. The objective of the first step is to minimize the cost of production in Equation (23):
min
,,,+,
(23)
Subject to the following technology14
,=,
,
1
(24)
Where  the share of capital in output is,  is the productivity, whose law of motion is:
log =󰇛1󰇜log +log 1+,
(25)
Where , is an exogenous shock and  is autoregressive components of the productivity shock. Using
the Lagrangian function to solve the previous problem of wholesaler firm:
=,+,,
,
1
(26)
The first order conditions are:

,󰇛1󰇜,
,
1 = 0
(27)

,=,
1,
1 = 0
(28)
From Equations (27) and (28), we arrive at:
=󰇛1󰇜,
,
(29)
=,
,
(30)
And from Equations (29) and (30),

󰇩󰇛1󰇜
󰇪,
,
(31)
Second step. In the second step, the wholesale firm maximizes its profit by choosing the price of its good
j:
max
,,,,,
(32)
Substituting Equations (21), (29), and (30) in (32):
max
,,󰇧
,󰇨󰇧
,󰇨
It lies in the following first order condition,
󰇛1󰇜󰇧
,󰇨+󰇧
,󰇨,
1= 0
14 As in the case of the utility of the households, the production function must have some properties: to be strictly increasing
(> 0 and > 0); to be strictly concave ( < 0 and  < 0); and to twice differentiable. It is also assumed that the
production function has constant returns to scale, 󰇛,󰇜=. Still, this function must fulfill the calls Inada conditions:
lim0=; lim = 0; and lim = 0.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
449
=1
,
(33)
Substituting Equation (33) into (29) and (30), and knowing that these firms have the same
technology—,=  ,=—the results for prices of the factors of production are:

=1
󰇛1󰇜

(34)

=1


(35)
Pricing a la Calvo. The wholesale firms choose how much to produce in each period, but following a rule
ala Calvo (Calvo, 1983) that says they fail to choose the price of their good in all periods. At each period t, a
fraction 0 < 1 < 1 of firms is randomly selected and allowed to choose the price of their good for period
t, ,
. The remaining firms (the ratio  of firms) keep the price of the previous period ,=,1 for the
product.
Thus, solving Equation (31) to ,: ,=󰇩󰇛1󰇜
󰇪,

And substituting this result in the production function (Equation (24)):
,=,󰇫󰇩󰇛1󰇜
󰇪
󰇬1
Getting:
,=,
 
󰇛1󰇜
1
(36)
And,
,=,
 
󰇛1󰇜

(37)
The wholesale firm has a probability  to keep the price of the previous period for the good and the
probability 󰇛1󰇜 to choose the price optimally. Once fixing the price in period t, there is the probability 
that this price will remain fixed in period t+1, a probability 2 that this price will remain fixed in period t+2,
and so on. This firm should take into account these probabilities when choosing the price of its own good in it
capacity to perform this adjustment.
Thus, the problem of the firm able to adjust the price of the good is:
max
,
󰇛󰇜
∞
=0 ,
,+++,+++,+
(38)
Where  is the factor of rigidity in the adjustment of prices and ,
 is the optimal price set by the firm
with the ability to adjust the price of your product. Equation (38) is the discounted profit of the firm during the
period which the price ,
 is in progress.
Substituting Equations (21), (36), and (37) in (38):
max
,
󰇛󰇜
∞
=0 +1 󰇧+
,
󰇨󰇫,
+
++
󰇛1󰇜1
+
+󰇬
Arriving at the following first order condition:

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
450
0 = 󰇧 󰇛󰇜
∞
=0 ,+󰇫1+++
,
+󰇛1󰇜1
+
+󰇬󰇨
,
=1
 󰇛󰇜,+++
+󰇛1󰇜󰇣󰇡1
󰇢+
+󰇤
∞
=0  󰇛󰇜
∞
=0 ,+
(39)
Combining the pricing rule of Equation (22), and the assumption that all firms with the ability to adjust
define equal value and that firms without this ability retain the same price, the overall price level is obtained by
the equation:
=1
1+󰇛1󰇜11
1
(40)
Government
The government sector in this paper is divided into three subsectors: fiscal authority, social sector, and the
monetary authority.
Fiscal authority. The government collects taxes and issues bonds to finance its spending on goods and
services. The result of the social sector is transferred to the rest of the government. So if social spending shows
a deficit (or surplus), this is financed (or appropriated) for the remainder of the government. Therefore, the
change in public debt is given by the rule as expressed in Equation (41):
+1
=
(41)
As could not be otherwise, the expense of the government is sensitive to the size of the public debt
(current debt 󰇛󰇜 relative to its steady-state level),  :
 =󰇛󰇜
(42)
Where  is the sensitivity of the government spending relative to the size of the public debt.
And tax revenue is obtained by the following equation:
=󰇛+󰇜+
+

(43)
Social expenditure. The social expenditure balance, , is the difference between the total collected
with the social security contributions and taxes of active workers, , and the Social Expenditure,
...15
Thus,
=+   ..
(44)
Monetary authority. The Central Bank of Uganda appears in this work following a simple Taylor rule
(1993) with the dual goal of output growth and maintenance of price stability:
=󰇛󰇜+󰇛󰇜+

(45)
Where 󰇛󰇜 and 󰇛󰇜 are the sensitivities of the basic interest rate in relation to the product and to the
inflation rate, respectively. The inflation rate is defined as:
=
11
(46)
15 Social expenditure include: Pension transfers, education and health expenditure.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
451
Equilibrium Condition of Goods Market
To complete the model it is necessary to use the equilibrium condition in the goods market. Wherein
aggregate production  is demanded by households ( and ) and Government 󰇛󰇜:
=++
(47)
Calibration
The model parameters are calibrated using past economic literature on Uganda’s monetary and fiscal
assumption. The model equilibrium is a set of 21 equations representing the behaviour of 21 endogenous
variables (See Table 2): ( ,,,,,,,,,,,,,,,,,,, ).
Consequently, it is necessary to assign values somehow for the structural parameters of the model
,,,,, ,,, σ, , ,,,,,.. , ,,,.
The main calibration procedure adopted here is to obtain the values of parameters from other relevant
works in the literature. Doshi et al. (2016) analysed the dynamic properties of a DSGE model for Uganda under
alternative parameterizations and identified “allowable ranges” of values for some of the key parameters in the
literature. This study retains some of the parameters used by Doshi et al. (2016) such as the discount factor (β);
the share of capital in output (α); the rate of capital depreciation (δ); consumption tax (Τ); corporate tax (Τκ);
labour tax (Τι); and the elasticity of substitution between intermediate good ().
The coefficient of relative risk aversion (σ) was obtained from Ostry and Reinhart (1992). The sensitivity
of the basic interest rate on the product (a) and on the inflation rate (b) were obtained from Taylor (1993). The
sensitivity of government spending relative to public debt (χ) was obtained from Costa Junior and Sampaio
(2014).
Table 2
Model Parameters Calibrated
Parameters
Definition of parameter
Value
Source
β
Discount factor
0.938
Doshi et al. (2016)
δ
Rate of public capital depreciation
0.05
Doshi et al. (2016)
σ
Coefficient of risk aversion
1/0.4551
Ostry & Reinhart (1992)
ψ
Marginal disutility of labour
1.5
Cavalcanti & Vereda (2010)
Τ

Consumption taxes
0.1123
Doshi et al. (2016)
Τκ
Corporate taxes
0.0366
Doshi et al. (2016)
Τι
Labour taxes
0.0366
Doshi et al. (2016)
Τp
Rate of contribution of pension
0.0169
MoFPED (2016b) & URBRA (2016)
W
Proportion of poor household
0.197
World Bank (2016)
Soc. Exp.
Total social expenditure (domestic and external)
0.789
MoFPED (2016b)
α
Share of capital in output
0.333
Doshi et al. (2016)
θ
Index of price stickiness
0.666667
Alidou (2014)
a
Central Bank response to output
0.5
Taylor (1993)
b
Central Bank response to inflation
1.5
Taylor (1993)
Elasticity of substitution between intermediate
goods
30
Doshi et al. (2016)
χ
Sensitivity of government spending relative to the
size of debt
0.1
Costa Junior & Sampaio (2014)
Source: Authors’ own calculation.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
452
Results
In this section, we analyse the effects of increase of tax rates on labour income, corporate profits, and
consumption. Note that the transition dynamics of selected variables, following a permanent increase in tax rate,
shows percentage deviations from initial steady state. The impulse response functions are illustrated in the
appendix.
Impact of Income Tax Increase
The immediate effect of an income tax increase is a reduction of overall household consumption driven by
a reduction of wages. In this case, the short term consumption of both the poor and the non-poor households’
decreases (Figure A1). However, the decrease among the poor households is stronger, as labour income is the
main determinant of their consumption. The cumulative decrease in consumption due to income taxation is 4.48%
point larger for poor households than that of non-poor household. This is largely because the non-poor
households are able to smooth the effect of income tax increase on consumption over a longer time
horizon—due to positive long-run wealth effect. The positive wealth effect moderates the impact of an increase
in income tax on total consumption.
Also an increase in the income tax leads to reallocation of production inputs from labour to capital, leading
to an increase of a capital demand and a decrease of labour utilisation (Figure A2). Inflation rises as a result of
a higher marginal cost of labour (Figure A3). Given that the Bank of Uganda places more weight on inflation
than on GDP, the nominal interest rate rises (Figure A3). The increase in interest rate initially decreases
investment (Figure A2).16 In addition, the fall in GDP implies a fall in employment (hence, an increase in
unemployment) (Figure A2). Unemployment suggests a reduced wage claims by workers (Figure A2).
Nevertheless, the increase in the income tax results in lower debt and a higher long run government expenditure
(figure A4).
Impact of Corporate Tax Increase
The immediate effect of an increase in the corporate tax rate is the reallocation of production inputs from
capital to labour, which results in a higher labour demand/employment and a lower capital utilisation (Figure
A2). The higher demand for labour leads to an increase on wages, which more than compensates for the
decrease in consumption of the non-poor households (Figure A1).1718 The marginal cost increases as a result of
the rise in rental rate (Figure A3). This is followed by an increase in inflation and the nominal interest rate
(Figure A3). The interest-rate-sensitive consumption of non-poor households’ decreases (Figure A1). The
consumption of poor households decreases slightly due to loss of labour income (Figure A1). Investment
initially decreases then increases significantly in the medium term as a result of decrease in the discounted
rental rates (Figure A2).19 On the fiscal side, government revenue increases and public debt decreases in the
same proportion as the increase in government spending (Figure A4).
Impact of Consumption (VAT and Excise) Tax Increase
An increase in consumption tax rate results in a rise in consumer prices lasting approximately four years
(Figure A3). Consequently, as goods and services become more expensive, poor-households will reduce
16 The decrease last for one year only.
17 Cumulative wage growth of 0.9% point.
18 Cumulative consumption growth of 0.03% point.
19 The decrease in rental rates takes two years.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
453
consumption, while non-poor households initially decrease consumption but eventually increase it two years
due to wealth effects (Figure A1). However, the total effect is a decrease in consumption of both poor and
non-poor households (Figure A1). Lower demand for goods, implied by the consumption tax increase, results
in a decrease of the demand for labour, a lower capital utilisation, a temporary decrease in output, and a fall in
wages. Inflation and interest rates initially rise as a result of a higher marginal cost, but eventually fall in the
medium term. The short term increase in interest rate decreases investment sharply in the short-run.
Nevertheless, the increase in the consumption tax results in higher government revenue, lower debt and a
higher government expenditure (Figure A4). Overall, the effects of the increase in consumption taxation are
largely similar although larger than those following a rise in income taxation.
Conclusions
This paper develops a closed economy DSGE model calibrated for Uganda with a comprehensive fiscal
block. The model is motivated by recent fiscal actions and announcements in Uganda and is primarily aimed at
stimulating the impact of fiscal consolidation, through a tax increase, on households and the real sector.
Specifically, the paper examines the impact of income tax, corporate tax, and consumption tax (VAT and
Excise tax) on households and the real macroeconomic aggregates.
The paper finds significant distributional effects of increasing the tax rate in Uganda to reduce debt. In
particular, the paper finds that an increase in the tax rate is associated with an increase in inequality, declines in
wage income and in the wage share of income, and increases in long-term unemployment. The result also
suggests that consumption taxes are relatively efficient in reducing the short to long term debt ratio when
compared to income and corporate tax rate. However, consumption tax also wipes the fiscal gains out through a
decline in output.
The adverse effects of consumption tax increase is typical if fiscal authorities engage in repeated rounds of
tax rate hikes in an effort to get the debt ratio to converge to the official target. For Uganda, rapid fiscal
consolidation may exacerbate poverty especially in the lagging regions of eastern and northern Uganda
(Ssewanyana & Kasirye 2013). This problem could be addressed by setting and monitoring debt targets in
cyclically-adjusted terms.
However, whether fiscal consolidation will achieve its objectives of reducing debt, maintaining the same
level of welfare will depend on how the burdens of taxation and the benefits of social spending are distributed.
One way is to differentiate the consumption taxes by retaining the current rate for durable and non-durables
consumed by the poor and adjusting the rate upwards for those consumed by the non-poor.20 This ensures that
the poor are shielded from the adverse effect of a tax increase as the government reduces it debt ratio.
It should be noted, however, that the effectiveness of such a policy will also depend on the allocative and
technical efficiency of the government budget apparatus, for it is possible to allocate larger shares of social
spending and tax exemption with limited welfare gains. Also, the drive to minimise the impact of
inequality-increasing tax should be rational and evidence driven. For instance, it is possible to reduce inequality,
while increasing poverty.
Another recurrent policy suggestion is to earmark the savings of an increased consumption tax to some
valued purpose such as provision of universal primary/secondary education. However, policy makers should
20 Examples of non-durables consumed by the poor are: rent, electricity, water, paraffin, charcoal, firewood, matches, and
toothpaste among others. Durables consumed by the poor: clothes, shoes, and blanket among others.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
454
note that earmarking may constrains spending on the prioritised item, in which case it impedes efficient
resource allocation. In this case, there is need to consider the nature of public expenditure driven by an increase
in revenue emanating from consumption taxes. Indeed, accompanying consumption tax reform with targeted
protection of the poorest consumers will automatically limit the impact on those likely to have the highest
marginal propensity to consume.
Overall, the results from this paper also concur with Lakuma and Lwanga (2017) who suggest that there is
limited space to increase tax rates in the medium term and consolidation should largely depend on expanding
the tax base, particularly by improving tax administration, reducing exemption, and reducing the size of the
informal sector.
References
Arghyrou, M. G., & Kontonikas, A. (2011). The EMU sovereign-debt crisis: Fundamentals, expectations and contagion.
Economic Papers 436. European Commission, DG Economic and Financial Affairs.
Alidou, S. (2014). Degree of price rigidity in LICs and implication for monetary policy. MPRA Paper No. 59844.
Auerbach, A. J., and Gorodnichenko. Y. (2012). Measuring the Output Responses to Fiscal Policy. American Economic Journal:
Economic Policy 4 (2): 1–27
Bova, E., Kinda, T., Woo, J., & Zhang, S. (2013). Distributional consequences of fiscal consolidation and the role of fiscal policy:
What do the data say? (Unpublished, Washington: International Monetary Fund).
Burger, P., Siebrits, K., & Calitz, E. (2016). Fiscal consolidation and the public sector balance sheet in South Africa. Stellenbosch
Economic Working Papers 11/15.
Calvo, G. (1983). Staggered prices in a utility-maximizing framework. Journal of Monetary Economics, 12, 383-398.
Cavalcanti, M. A. F. H., & Silva, N. L. (2010). Impactos de Políticas de Desoneração do Setor Produtivo: Uma Avaliação a Partir
de um Modelo de Gerações Superpostas. Estudos Econômicos, 40(4), 943‐966.
Chari, V. V., Kehoe, P., & McGrattan, E. R. (2007). Business cycle accounting. Econometrica, 75, 781-836.
Coenen, G., Straub, R., & Trabandt, M. (2013). Gauging the effects of fiscal stimulus packages in the euro area. Journal of
Economic Dynamics and Control, 37(2), 367-386.
Coenen, G., and R. Straub (2004). Non-Ricardian Households and Fiscal Policy in an Estimated DSGE Model of the Euro Area.
Manuscript, European Central Bank.
Costa Junior, J. C., & Sampaio, A. V. (2014). Tax reduction policies of the productive sector and its impacts on Brazilian
economy. Dynare Working Papers Series, Working Paper #36.
De Grauwe, P., & Ji, J. (2013). Self-fulfilling crises in the Eurozone: An empirical test. Journal of International Money and
Finance, 34, 15-36.
De Mooij, R., & Keen, M. (2013). Fiscal devaluation and fiscal consolidation: The VAT in trouble times. In A. Alesina and F.
Giavazzi (Eds.), Fiscal policy after the financial crisis (pp. 443-485). Chicago: University of Chicago Press.
Dixit, A. K., & Stiglitz, J. E. (1977). Monopolistic competition and optimum product diversity. The American Economic Review,
67, 297-308.
Doshi, T., Joutz, F., Manzano, B., Lakuma, P., & Lwanga, M. (2016). The challenges of macroeconomic management of natural
resource revenues in developing countries: The case of Uganda. EPRC Working Paper #124.
East Africa Community (EAC). (2013). Protocol on the establishing of the East African community monetary union.
Hall, R. (1997). Macroeconomic fluctuations and the allocation of time. Journal of Labour Economics, 15(1), 223-250.
Hoeller, P., & Pisu, M. (2014). Introduction. In P. Hoeller (Ed.), Income inequality in OECD countries: What are the drivers and
policy options? OECD.
Hutton, E., Thackray, M., & Wingender, P. (2014). Uganda revenue administration gap analysis program—the value-added tax
gap. International Monetary Fund (IMF).
International Monetary Fund (IMF). (2015). Fifth review under the policy support instrument and request for waiver of an
assessment criterion and modification of assessment criteria—debt sustainability analysis.
International Monetary Fund (IMF). (2016). Sixth review under the policy support instrument and request for a one-year
extension.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
455
Kuznets, S. (1955). Economic growth and income inequality. The American Economic Review, 45, 1-28.
Lakuma, C. P., Katunze, M., Nagawa, M., Mawejje, J., Lwanga, M., Mbowa, S., & Shinyekwa, I. (2017). Country reviews of
capacity development: The case of Uganda. EPRC Occasional Paper #41.
Lakuma, C., & Lwanga, M. (2017). Linking budgets to plans in a constrained institutional and resources and environment: The
case of Uganda. EPRC Research Series No. 131.
Lakuma, C. P., Marty, R., & Kuteesa, A. (2016). Survival analysis of regional unemployment in Uganda: Evidence from the
Uganda National Panel Survey (UNPS). African Development Review, 28(1), 140-154.
Ministry of Finance, Planning and Economic Development (MoFPED). (2013). Public debt management framework.
Ministry of Finance, Planning and Economic Development (MoFPED). (2016a). Charter of fiscal responsibility. Kampala.
Ministry of Finance, Planning and Economic Development (MoFPED). (2016b). The national budget framework paper 2017/2018.
Kampala.
Ostry, J. D., & Reinhart, C. M. (1992). Private savings and terms of trade shocks. Evidence from developing countries. IMF Staff
Papers, 39(3), 495-517.
Organisation for Economic Co-operation and Development (OECD). (2012). Fiscal consolidation targets, plans and measures in
OECD countries. Restoring Public Finances, 2012 update.
Primiceri, G. E., Schaumburg, E., & Tambalotti, A. (2006). Intertemporal disturbances. NBER Working Paper #12243.
Rawdanowicz, L. (2014). Choosing the pace of fiscal consolidation. OECD Journal: Economic Studies, 2013(1), 91-119.
Retrieved from http://dx.doi.org/10.1787/eco_studies-2013-5k3tq96mbr44
Smeeding, T. (2000). Changing income inequality in OECD countries: Updated results from the Luxembourg Income Study (LIS).
Working Paper #252. Luxembourg Income Study.
Ssewanyana, S. N., & Kasirye, I. (2012). The dynamics of income poverty in Uganda: Insights from the Uganda national panel
surveys of 2009/2010 and 2010/2011. EPRC Occasional Paper #35.
Ssewanyana, S. N., & Kasirye, I. (2013). Poverty and inequality dynamics in Uganda: Insights from the Uganda national panel
surveys 2005/2006 and 2009/2010. EPRC Research Series # 94.
Taylor, J. B. (1993). Discretion versus policy rules in practice. Carnegie Rochester Conferences Series on Public Policy, 39,
195-214. North-Holland.
Uganda Retirement Benefits Regulatory Authority (URBRA). (2016). Annual pension industry report 2015.
World Bank. (2016). The Uganda poverty assessment report 2016. Report No. ACS18391. World Bank, Washington DC.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
456
Appendix A
Figure A1. Impulse-response functions for tax shocks on households, consumption, and income.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
457
Figure A2. Impulse-response functions for tax shocks on employment, wage, investments, and capital.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
458
Figure A3. Impulse-response functions for tax shocks on interest rates, prices, rent, and inflation.

THE DISTRIBUTIONAL IMPACTS OF FISCAL CONSOLIDATION IN UGANDA
459
Figure A4. Impulse-response functions for tax shocks on Govt. Exp., tax revenue, and capital income.