




































American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

1 | P a g e  

THE NEXUS OF GOVERNMENT EXPENDITURE AND 
FINANCING MECHANISMS: INSIGHTS FROM GHANA 

 
 

Dr. Esi Ama Anaman 
Department of Banking and Finance, School of Business, University of Education, Winneba, Ghana 

 
Abstract: Government expenditure is a pivotal driver of economic growth globally, facilitating 
infrastructure development and institutional support. However, the financing channels for these 
expenditures play a crucial role in shaping economic trajectories. This paper delves into the intricate 
relationship between government expenditure and its financing modes. While conventional wisdom 
dictates that government spending should primarily rely on direct and indirect taxes, as well as non-
tax revenues, many developing nations resort to borrowing to bridge budgetary shortfalls. This paper 
explores the dynamics of these financing methods and their mutual influences. It sheds light on the 
challenges faced by governments in balancing their budgets and underscores the imperative of 
sustainable fiscal policies. 
Keywords: Government expenditure, Financing channels, Economic growth, Fiscal policies 
Developing nations 
 
 
Introduction  
Throughout the world, government expenditure remains an avenue that provides a strong impetus for 
spurring economic growth and indeed it is the government expenditure outlays in every economy that 
enable the state to create the necessary infrastructure and the relevant institutional mechanisms to 
support the multiplicity of economic activities across the spectrum. Whilst it is recognized that the 
government expenditure is critical in every economy, it must also be noted that such expenditures are 
usually greatly influenced by the financing channels through which the expenditures are derived. Much 
as it is true that the trajectory of government expenditure has riposte on the various financing modes, 
it is also equally an established fact that these financing modes can affect each other .In the literature, 
there is seem to be a general view that government expenditure must as much as possible be financed 
from the conventional sources- direct and indirect tax as well as non-tax revenues. However, in the 
developing world especially, it has become customary to leverage on borrowing modes as a way of 
meeting the government expenditure levels required in the budget plans as the conventional revenue 
raising mechanisms always fall way short of the intended targets sufficient enough for the government 
operations to be pursued seamlessly. 
Within these contexts, there has emerged a strand of empirical research which seeks to examine the 
fiscal behaviours of governments and in particular how the availability of the borrowing modes 
dampens the resolve of the fiscal authorities to be up and doing and maximize revenues. This is well-
articulated in the early studies in fiscal behaviours; Griffin (1970), Heller (1975) and Mosley et al.(1987) 
etc. One important aspect of this discussion centers on the aid effect on the other financing modes and 
government expenditure itself. According to Osei, Morrissey and Lloyd (2005), studies on the effects of 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

2 | P a g e  

aid on fiscal behaviour can generally be categorized into those which direct their attention at 
investigating the effects of aid on the composition of government expenditure and those which in 
addition to examining the effect of aid on the allocation of government expenditures also assess aid 
effect on tax effort and government borrowing.  
In Ghana, just as in a lot of the developing countries, the pressure on successive governments to meet 
the aspirations of the citizenry has meant that government has to go out of the way to find the needed 
resources to ensure that programmes and projects are duly executed even against the backdrop of 
insufficient revenue generated and this situation has persisted for a long time.  There are some who 
believe strongly that this has continued to exist because of the opportunity which is always open for the 
government to look anywhere to fund its activities even though government could be more prudent in 
staying reasonably within its revenues limits or aggressively pursuing the much needed tax reforms 
which could result in enhanced revenue collection. A number of questions thus arise. Does the 
availability of other government expenditure financing modes encourage government to continue to 
increase expenditure? Do aid and borrowing dampen tax revenue generation? Do grants and borrowing 
trigger differential fiscal behaviour by government? Again, how does the availability of the non-tax 
government expenditure financing modes influence the allocation of government expenditure?  
Gleaning the literature, it is obvious that contemporary studies in this arena have moved forward the 
frontiers of knowledge established by the earlier ones, eg Griffin (1970), Heller (1975) and Mosley et al 
(1987) and Khilji and Zampelli (1994).  
The most recent study conducted within the Ethiopian context by Mascagni and Timmis (2014) 
develops a model of fiscal behaviour encompassing tax and non-tax revenues, government expenditure, 
grants and loans which modifies Osei et al (2005) and Lloyd et al (2009) which include government 
(capital and recurrent), total tax revenue and domestic borrowing for the former and foreign financing, 
capital expenditure, recurrent expenditure, tax revenue and domestic borrowing in the case of the 
latter. In these studies, the researchers did not avert their minds to the fact that the dynamics may not 
possibly be the same if the tax financing source is disaggregated into direct and indirect tax channels. 
In other words, in this study apart from categorizing aid as grants and loans, we also include direct and 
indirect tax financing as separate variables. This is because we believe that aid and borrowing may not 
necessarily have the same effects on direct and indirect taxes.  Thus the main difference between the 
present study on one hand and that of Osei et al (2003) and other previous but related studies on the 
other hand is that it we introduce the hypothesis that the responses of direct and indirect taxes 
respectively to borrowing-both external and domestic are different and also have the benefit of current 
data for the analysis to determine whether prevailing circumstances deviates from Osei et al(2003). 
The rest of the paper would be arranged in the following manner; Section II is devoted to examining 
the fiscal policy environment, trends in fiscal management and borrowing by the government of Ghana 
over the years. In section III, we proceed to discuss the theoretical and empirical issues relating to fiscal 
behaviour by government especially focusing on aid and borrowing and their effects on government 
fiscal management. Section IV sets out the econometric approach and a brief description of the data set 
for the empirical analysis whilst Section V reports the results of the data analysis and proceeds to 
discuss them. Finally, section VI covers the synopsis and conclusions from the study.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

3 | P a g e  

Section IIː Trends in Fiscal management in Ghana  
When one does a careful study of fiscal policy in Ghana, one can identify clear, distinct periods of unique 
fiscal behaviours? In the main, the periods the early 1960s to the late 1960s, 1969-1972, 1972-1983, 
1983-1991 and from 1992 to the present can be associated with peculiar fiscal behaviours though in 
some of the periods the fiscal management approaches appear similar. In the sixties, with the 
emergence of the country from colonial rule there was an urgent need for the government to put in 
place structures of state and build critical infrastructure like educational and health institutions while 
also embarking on rapid industrialization and modernization and as such, government committed 
massive public expenditures into achieving these objectives. During this period, a good chunk of the 
expenditures were financed from domestic sources with very little coming by way of aid inflows.   
The succeeding period however saw a modification of fiscal behaviour as government substantially 
disengaged from the previously pervasive role of the government in the economy, in line with the 
philosophy of the people in authority at the time and by virtue of the programme that they entered into 
with the Bretton Woods institutions, government at the time embarked on privatization of a good 
number of the state enterprises. Osei et al(2003)  submit that from the 1960s to early 1970s , aid inflow 
was relatively insignificant and constituted about 2% of GDP and roughly around 12% of all revenues 
available to government. In the middle to the late 1970s, there was a shift in the behaviour of the 
government as government activities were driven essentially by monetary expansion through 
borrowing from the Bank of Ghana as domestic revenues sharply reduced on account of the decline in 
the real side economic activities precipitated by inappropriate policies introduced by the then military 
rulers coupled with adverse economic and external trade climate. The situation was compounded by 
the repudiation of loans which had been contracted by previous governments leading to the virtual 
drying up of the foreign aid inflows.  
In early 1980s, even though the country had returned to constitutionalism, the country continued to 
suffer from the decline in economic activities as result of the deterioration of the macroeconomic 
environment. According Durdonoo’s (2000) calculations, taxes on income and property fell from 2.8% 
of GDP to a mere 0.98% of GDP in  
1983 whilst tax revenue from domestic activities was down to sub one percent in 1983 from 
approximately 5% of GDP. Proceeds from international transactions also dropped from 12% in 1970 to 
2.7% in 1983.The precarious revenue situation in the country is illustrated by Osei et al(2003)  when 
they intimate that overall the tax levels took a nosedive between 1970 and 1983,plummeting from a 
high level of about 700 million USD to 160million USD. The fiscal situation in the country however 
improved dramatically after the Economic Recovery Programme (ERP) was launched. Indeed it is 
estimated that between 1983 and 1998, tax revenue collections shot up in dollar terms to 1.3 billion 
USD representing a more than six fold increase of the 1983 level. Generally speaking, total government 
revenue is measured in some calculations to have increased twenty-six times between 1983 and 1990. 
Osei et al (2003), suggest that since this period was largely marked by good amount of aid inflows, it 
appears that the aid flows did not undermine government's tax revenue mobilization. During the period 
whilst tax revenue and aid inflows were increasing, government expenditure also continued to increase 
though at a slower pace, on account of the some of the ERP measures which had been introduced to 
stem rapidly increasing government expenditure experienced in the period before ERP. The post 1992 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

4 | P a g e  

period has generally been characterized by rapidly expanding government expenditure largely fueled 
by the people across the country making demands on politicians and indeed some actually using the 
provision of certain projects and infrastructure as a tool for cajoling and blackmailing functionaries of 
government. However, revenue mobilization during the period has not kept pace with government 
expenditures and the peoples aspirations for that matter. Indeed for a long time now total government 
revenue is consumed largely by payment of emoluments, statutory obligations and then interests 
payments, which itself is a product of the increasing imperative for borrowing by government. The 
argument that the political structure in Ghana has tended to reinforce fiscal behaviours by successive 
governments in the fourth republic is amplified when one considers Ghana’s fiscal position in election 
years. Election cycles have generally exacerbated the problem and this is evidenced by the fiscal deficits 
which were recorded in the years 2008, 2012 and 2016 respectively. One major development which has 
also to a great extent influenced fiscal behaviours especially post 2012 has been the reclassification of 
Ghana as a middle income country .This has restricted the country's access to concessionary loans and 
grants and compelled governments to syndicate relatively expensive loans from the international 
commercial markets on account of the fact that Ghana's tax to GDP ratio is woefully below the average 
middle income levels. Indeed it is very instructive to note that in the West Africa sub region, Ghana's 
tax collections as a percentage of GDP is the lowest. Against this background, the issuances of Euro 
bonds have become an important feature of government strategy for financing projects and 
programmes of government as missing revenue targets have become a constant feature of fiscal 
management in Ghana.  
Section III ːTheoretical, conceptual issues and empirical underpinnings  
Fiscal policy formulation is one of the basic functions of every government in the sense that it primarily 
involves the strategies that governments use to raise income to be able finance government's activities. 
In the main, most governments rely on revenues generated from taxation as the most reliable source of 
income. However, in most parts of the world particularly the developing world because of the demands 
on government to ensure rapid development and the exigencies of the time, they are unable to stick to 
the incomes available to them through taxation and therefore have to resort to other means of financing 
their programmes and projects. These come in the form of foreign aid- loans and grants and domestic 
borrowing.   
According to Njeru (2004), one of the most critical issues which has been  a subject of debate by 
economists in this area of research is whether or not the aid process is undermined by the ability of the 
aid receiving country to alter the their spending patterns to subvert the sectoral distribution of 
expenditure for designated projects. The general contention is that the ability of the recipient country 
to reallocate the aid can usually affect the intended economic performance envisaged under the aid 
structure. This is particularly the case when aid earmarked for developing critical infrastructure in a 
given economy is diverted into financing government consumption like catering for emoluments of 
workers and buying goods and services for government machinery rather than creating the required 
infrastructural overheads which then provide the necessary platform for increasing the level of 
economic activities. This is what economists usually refer to as aid fungibility. This is reinforced by 
Bwire et al (2017) who contend that fungibility arises when aid recipients do not use the aid for purposes 
for which they were given by the donors. Thus in many respects, a lot of the developing countries employ 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

5 | P a g e  

resources from aid to able to be able to deal with the deficits usually associated with their budgetary 
processes (Devarajan et al, 1998, Ali et al , 1999). These views are very replete in the fiscal response 
studies. For example Matins (2007) asserts that one of the most fundamental issues which relate to the 
effectiveness of aid is how aid influences the government fiscal accounts. In particular, Martins (2007) 
stresses that one critical pillar of the fiscal response studies is assessing how the aid itself is allocated 
between the various expenditure channels, the way it affects tax effort and then its effect on fiscal 
balance and debt sustainability. This is view is reinforced by Mavrotas (2002) who stresses that since 
aid is given to a government, its impact on the overall economy is contingent on fiscal behaviour of the 
government. From the perspective of Mascagni and Timmis (2014), aid is usually a more politically 
expedient and convenient source of revenue and therefore has the tendency to discourage tax effort 
which in the literature is characterized as tax displacement. They however stress that this argument is 
stronger in respect of grants than loans because of the obvious fact that loans require future payments 
whereas grants do not. Mascagni and Timmis (2014) put forward another dimension of the aid–revenue 
debate which is that rather undermining the revenue efforts, aid may actually help strengthen tax 
administration and improve tax policies. Again, it is argued that if aid is utilized properly and effectively 
it may promote economic activities, expand the economy and by that increase tax yields from the 
economy.  
In the view of Njeru (2004), aid inflows into the developing countries has tended to create an ominous 
dependency mentality which seem to affect their economic performances and the absence of such funds  
greatly affect their budgets, usually coming with their attendant consequences. This is echoed by 
Feyzioglu et al (1999) who posit that aid dependence is something which has widespread ramifications 
for countries. There is an also another dimension of the aid  debate which is canvassed by 
Martins(2007) .In his estimation apart from  the fact that aid is sometimes used to offset domestic 
debts, it can trigger off extra government expenditures especially in aid funded projects which require 
some maintenance and recurrent expenditure. Again aid programmes and projects which require 
counterpart funding may in reality also further put pressure on government's already overstretched 
finances and thus lead to mounting deficits. This scenario is what McGillivray and Morrissey (2000) 
describe as aid illusion. Having regard to the fact that foreign aid may be associated with some 
challenges; the other viable alternative is borrowing from domestic sources to be able to undertake the 
necessary government activities. However, this avenue also comes with its own problems. One of the 
challenges that this poses is that it leads to a situation in which government enters the credit markets 
to compete with private entities for the available funds, a situation which generally inhibits the growth 
of privately engineered economic growth in an economy. This can in many respects also affect tax 
mobilization. Aside of these issues ,it is often argued that in a lot of the developing countries, excessive 
reliance on borrowing modes to enable governments meet its commitments in terms of delivering the 
required services has invariably led to compounding debt servicing obligations and thereby constricting 
fiscal space as the piling of debts both internally and externally  have tended  to increasingly impose 
severe servicing and payments obligations on the government thereby limiting what the government 
can achieve within its resource envelop. Studies in fiscal response has its origins in the 1970s starting 
with Heller(1975) who used a utility based government fiscal behaviour function  to show that the aid 
process has effects on how governments manage their  fiscal operations . Despite the vast array of 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

6 | P a g e  

research in this area, results from these studies have largely been inconclusive. According to Mascagni 
and Timmis (2014), this situation may be due to the fact that various studies adopted different 
methodologies and contexts. In his study, Njeru (2004) assessed the impact of foreign aid on public 
expenditure in Kenya and based on government welfare utility function specified government 
expenditure related to aggregated government revenues from tax and domestic borrowing sources, 
programme aid and project aid. The dynamic analysis indicated that aid does not affect government 
expenditure whilst it is also established that government is able to divert aid funds into government 
consumption expenditure.   
A similar study by Osei et al (2003) modelled the fiscal effects of aid in Ghana by particularly employing 
a dynamic impulse response functions. Using the government utility maximization approach,  two 
variants of the empirical model were specified; these are aggregate government expenditure ,domestic 
borrowing ,total government tax revenues and aid finance  on hand  and  government capital 
expenditure, government consumption expenditure, domestic borrowing, total tax revenue and  foreign 
aid. In the analysis, it is established that there are co-integrating relationships in both models. Results 
also showed that in both models, aid finance and domestic revenues are in long run negatively related 
to domestic borrowing whilst government expenditure whether aggregated or disaggregated positively 
influences domestic borrowing .Another important finding that issues from Osei et al (2003) is that aid 
in Ghana over the study period has generally been used to replace domestic borrowing as a method of 
financing government projects and programmes.  
The work of Martins (2007) also explores further the aid-fiscal behaviour nexus within the context of 
the Ethiopian economy and actually separates aid into two components-loans and grants based on the 
premise that fiscal response by government to them may be different. The conclusions from the 
estimations are that whilst aid finance positively affects total government expenditure, its effects on 
government consumption expenditure is less pronounced and that external borrowing has a bigger 
impact on public investment than grants. Another important finding from this study is that aid finance 
undermines domestic revenue mobilization.   
In his contribution in the fiscal response and effectiveness of aid studies, Mavrotas (2002) introduced 
a categorization of foreign aid into project aid, programme aid, technical assistance and food aid and 
based on the popular utility maximization approach obtained results which affirm that aid may be 
fungible.  
The study by Mascagni and Timmis (2014) also dealt with the fiscal effects of aid in Ethiopia employing 
the co integrated vector autoregressive model based on the conventional Heller utility maximization 
function. Their model encompassed total government expenditure, tax and non-tax revenues, grants 
and loans .In the long run , government expenditure was established to be related to  domestic revenue 
and  foreign aid; there is a positive  relationship between tax revenue ,grants and loans. In the short run 
too, government expenditure is established to be influenced positively by both grants and loans whilst 
the equation for tax shows that non-tax revenue, grants and loans are all positive determinants. The, 
loans variable is also impacted positively by non-tax revenue but negatively by tax revenues. The most 
recent study in this area, authored by Bwire et al (2017) also sought to examine fiscal reforms and the 
effects of aid in Uganda employing a dynamic analysis and to test whether aid flows lead to a full or less 
than a full change in government expenditure, determine if aid displaces tax effort as well as ascertain 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

7 | P a g e  

whether aid and domestic borrowing are substitutes in fiscal management in Uganda. Their analysis 
uncovered three co integrating  equations for government expenditure, revenue and aid and found that 
in the long run aid leads to increased tax effort  and public spending  but a reduced domestic borrowing. 
Section IVː Empirical model 
According to Osei et al (2003), there are two broad approaches adopted in the literature to examine the 
fiscal effects of aid. The first approach is the fungibility studies which attempt to assess aid effects on 
the spending patterns of the government and the other method which seeks to integrate revenue 
variables into a government utility function to determine the overall impact of the aid process on the 
fiscal behaviour of the government which they call the fiscal response models (FRMs). Since the latter 
is more comprehensive in its outlook, it is more popular in the literature and has been adopted in most 
of the recent studies. This approach is based on the seminal work of Heller (1975) which posits 
government allocating revenue among the different expenditure streams but subject to some budget 
constraints. In the model, government expenditure is usually categorized into government 
consumption and capital expenditure whilst government derives its income endogenously from 
conventional taxation sources and domestic borrowing. However, in these models, foreign aid is 
defined as an exogenous source of revenue which modifies the government budget constraints; even 
though it is assumed not to be relevant in the utility function of the government since it is not defined 
as one of the variables for which targets are set.  Against this background, Osei et al (2003) set the 
maximum unconstrained value of the utility function represented by α0 as a quadratic expression 
defining a loss in the form below;   
U=α0–α1/2(GK-GK*)2-α2/2(GC-GC*)2-α3/2(R-R*)2-α4/2(D-D*)2                                                              (1),  
where GK*, GC*,R* and D* are exogenous target values of government capital expenditure ,government 
consumption expenditure ,total government revenue and government borrowing from domestic 
sources. The above equation is thus maximized subject to the following budget constraints,           
 GK = (1-ρ1)R + (1-ρ2)F +D                                                                                                                                    (2)  
and   
         GC=ρ1R+ ρ2F                                                                                                                                               (3)  
Where equations (2) and (3) are disaggregated equations derived from the total government 
expenditure constraints, of the form,   
GK +GC = R+F+D                                                                                                                                ( 4)  
From the above equations, it is taken that ρ2 represents the fraction of aid which is diverted into 
financing government consumption ;in other words the extent of the fungibility of aid .The implicit 
argument underlining this formulation is that when foreign aid is received , it is meant for capital 
investment .However , as the aid comes into the economy, a part of it is channeled into financing 
recurrent expenditure which means that mathematically, ρ2=0 ex ante but this according to Osei et al 
(2003) is not in the real world realistic because aside of directing resources into investments in the 
economy, foreign aid sometimes finances certain components of  government consumption 
,particularly in the social sectors especially education and health, hence ρ2≠ 0 is an unrealistic 
assumption but ρ2> 0 at most times is the most realistic assumption to make. This situation occurs 
especially when aid comes in in the form of budgetary support or even strictly as aid funded project in 
an economy. With the inherent limitations of this approach, Franco-Rodriguez et al (1998) modified 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

8 | P a g e  

the approach by defining a utility function such that foreign aid is interacted and integrated directly 
into the function thereby making aid endogenous. The argument put forward is that generally 
governments define targets for aid flows and this tends to influence their fiscal behaviour. As a result, 
the quadratic utility loss function expressed in (1) becomes   
     U =α0– α1/2(GK-GK*)2- α2/2(GC- GC*)2- α3/2(R-R*)2 –α4/2(F-F*)2  -α5/2(D-D*)2                                          
(5) whilst the constraining function becomes  
GC ρ1R + ρ2F + ρ3D                                                                                                                                               (6) 
since external flow of funds tend to influence how resources are allocated among competing needs.  
In this current paper, we further redefine (5) as  
   U =α0– α1/2(GK-GK*)2- α2/2(GC- GC*)2- α3/2(DT-DT*)2 – α4/2(IT-IT*)2 -α5/2(F-F*)2-α6/2(D-D*)2–
α7/2(Gr- 
Gr*)2                                                                                                                                                                         (7)  
Subject to    
GC ρ1DT + ρ2IT + ρ3F+ρ4D +ρ5Gr                                                                                                                     (8)   
The implicit meaning of the above is that both direct and indirect sources of revenue are endogenously 
determined in addition to the other sources of revenue. In this formulation, we separate F, external 
borrowing from Gr, grants because in the literature, it is argued that most governments treat loans 
differently from grants which are not to be paid back. Though Franco-Rodriguez et al (1998) provided 
an improvement of the earlier fiscal response models (FRMs), they did not address the methodological 
challenges that most of the earlier studies were fraught with. Osei et al (2003) therefore in their study 
changed direction to the new vector autoregressive (VAR) approach which in their view provided the 
means to go round the existing problematic methodological frameworks whilst making it easier to 
define the dynamic linkages between the various components of the budget. Building upon Osei et 
al(2003), M'Amanja et al (2005),Martins (2010),Bwire et al(2017) and Mascagni and Timmis(2014), 
we specify two variants of the  
VAR model involving aggregate and disaggregated government expenditure models belowː  
                             (GE, DT, IT, Db, Fb, Gr)   and   
                              (GK, GC, DT,IT, Db, Fb, Gr) respectively where GK is government capital expenditure, 
GC is the government consumption expenditure, IT is indirect tax revenue ,DT defines direct tax 
revenue, Db represents domestic borrowing ,Fb  is used for external borrowing whilst Gr is grants 
obtained from various external sources and finally GE defines aggregate  government expenditure. In 
these models above, the application of the VAR allows us to determine whether the variables are in the 
long run are dynamically related whilst at the same time providing useful information about the short 
run properties of the models. Generally an orthodox VAR model is defined as a dynamic system in 
which all the variables are endogenously determined and each of them is represented as a function of 
its own lags and the lags of the other endogenous variables. The advantage from this, according 
Blanchard and Peroti (1999) is that the system assumes  a priori there is no direction of causation 
among the variables of interest.   
Mathematically, we define our VAR (k) as  
                                 Xt=φ1Xt-1 + φ2Xt-2 +φ3Xt-3 +…..+φkXt-k+ πRt + t, t=1,2,----,n.                                                (9) 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

9 | P a g e  

In the above, Xt is defined as a (m × 1) vector of non-stationary variables whose order of integration is 
one and  which are jointly determined, whilst the Rt is also a vector of deterministic variables of 
dimension, (p× 1 ). The coefficients π and φ which are to be estimated are matrices of the dimension 
(m  p) and (m  m) respectively whereas the disturbance term t   is a vector of dimension (m 1)  and 
k is the lag length of the system.   
Using the Johansson (1991) approach, the general VAR can be transformed into an error correction 
model,  
usually referred as a restricted VAR of the form                       
∆Xt=α+ψXt+τ∆Xt-1+……..+τk-1∆Xt-k+1+ ∈t,t=1,2,-----,n.                                                                                       (10)                                                             
In this expression, we use the τs to define the short run characteristics of the variables.  Specifically the 
coefficients of the lagged dependent variable represent the feedback in the system whilst the coefficients 
of the other endogenous variables in the system define the pass through effects of these variables on the 
dependent variable. The matrix of coefficients ψ represents the long run equilibrium relationships 
among the variables of interest in the system. We start the analysis by examining the stationarity 
properties of the variables. This is important because in empirical analysis, most macroeconomic 
variables have been found to be non-stationary as result of their time dimensions and as a result 
prejudice and distort estimations. This thus makes it imperative for the non-stationary properties to be 
dealt with. In the words of Thomas (1993),if a variable is stationary, it means that the time path traced 
by the variable is stable. In other words,  a series is said to be stationary when it has a spectrum which 
is finite but non-zero at all frequencies.   
Mathematically determining the stationarity of a series Yt involves finding whether the equation   
                            Yt = α0+α1t +α + ut                                                                                      (11) 

follows an AR process.  
Typically, assessing the stationarity properties of variables involves testing the following hypotheses;                  
H0ː The series has unit roots   ,H1 ːThe series has no unit roots. 
In the conventional VAR system, the order of integration of the variables allowed is one meaning that 
the each variable in the system must attain stationarity after first differencing.  
Beyond examining the stationarity status of the variables, we employ the Johansson approach to test 
for co integration; that is to ascertain whether there exists a linear combination of the variables which 
is also stationary. According to Anaman et al (2017),co integration is the statistical implication of the 
existence of  a long run equilibrium relationship between economic variables. Soli et al (2008) also 
characterize co integration as representing the tendency of variables to drift together over time. The 
obvious advantage in the Johansson approach over the other methods of determining long run 
equilibrium relationships is that it makes it possible to uncover more one co integrating vector at a time. 
To proceed with this, we test,  
H0ːthere is no long co integrating vector in the system,  as opposed to 
H1ː At least one co integrating vector exists in the system.  
In a VAR system, a major requirement is that for variables to be co integrated they must have the same 
order of integration. This is underscored by Enders (1995) who emphasizes that for variables to be co 
integrated they must be integrated of the same order and have a linear   combination of residual 
sequence which is stationary .  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

10 | P a g e  

In order to extract more information from the system and sufficiently understand the dynamic 
relationships among the variables in our system, we undertake impulse response analysis. The 
importance of the impulse response analysis is buttressed by Osei et al (2003).According to them, when 
the interrelationships that characterize economic systems are considered, it is always more informative 
to undertake an impulse response analysis especially when the analysis involves uncovering short and 
long run relationships within a given system. Osei et al (2003) assert that the advantage that the 
impulse response analysis has is that it captures the net effect of both the direct and indirect impact of 
a shock, not only in the long run but also at all periods after the shock has been transmitted. Johnston 
and DiNardo (1997) underline the relevance of the impulse response function by intimating that it 
traces the chain reaction or the knock-on effects arising from one standard deviation perturbation in 
one innovation in the system over time on the other variables in the system granted that no other shock 
affects the system afterwards. Impulse response functions can thus measure both the current and future 
values of the given endogenous variable to one standard deviation shock in one of the innovations. 
Lutkepohl and Rimmers (1992) also reinforce the importance and suitability of the impulse response 
in a dynamic analysis.    
Generally, the impulse response function can be defined as the moving average representation of our 
equation (9)  
,expressed as  
Xt  πRt-1                                                                                                                                              (12), 

where, the As are of dimension (m  m)  
Apart from the impulse response analysis, we employ the forecast error variance decomposition from 
our VAR model to ascertain and predict the most important innovation for each endogenous variable 
along the entire time horizon. This will enable us to identify which variable is most relevant in achieving 
a given objective.   
According to Bhasin (2004), in a VAR model, variance decomposition is usually  employed to isolate 
the innovations of the endogenous variables into the portions which can be attributed to own 
innovations and that which are due to innovations of other variables in the system and in doing so we 
recourse to the Cholesky method based on Sim's recursive approach.   
Data Set   
For the purposes of this study, we employ annual series for all the variables from 1978 to 2017 .The 
variables were largely extracted from the World Bank Databases and supported with data from Ghana 
Statistical Service(GSS) and the Bank of Ghana.  
Results of Data Analysis  
Test for stationary (Unit roots tests)  
In the tables below, we report the results of the stationary tests. 
 
 
 
 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

11 | P a g e  

 Table1ːUnit root tests of log levels of variables 

Variable  
(log levels)  

ADF test 
Statistic  

Prob level  Phillip Perron 
test statistic  

Prob level  

ldb  -2.298031  0.1778  -1.917340  0.3212  

ldt  -1.682815  0.4318  -1.597301  0.4744  

lfb  -0.413932  0.8968  -0.404310  0.8985  

lgc  -0.141300  0.9375  -0.072070  0.9455  

lge  -1.130286  0.6941  -1.120580  0.6980  

lgk  -0.724807  0.8286  -0.710912  0.8322  

lgr  -1.946252  0.3085  -1.946252  0.3085  

lit  -2.194319  0.2115  -2.057338  0.2623  
Sourceː Author’s calculations using E Views  
 Table 2 ː Unit root tests of first differences of variables 

Variable  
(first 
differences)  

ADF statistic  Prob. level  Phillips Perron  
Statistic  

Prob. level  

dldb  -4.411600  0.0012  -4.425672  0.0011  

dldt  -5.736767  0.0000  -9.443833  0.0000  

dlfb  -6.196007  0.0000  -6.195848  0.0000  

dlgc  -5.475900  0.0001  -5.466520  0.0001  

dlge  -5.076895  0.0002  -4.944660  0.0003  

dlgk  -5.107552  0.0002  -5.032008  0.0002  

dlgr  -6.775472  0.0000  -6.799061  0.0000  

dlit  -7.863574  0.0000  -9.818331  0.0000  
Sourceː Generated from E Views estimations  
From tables 1 and 2, we infer that all variables are non-stationary at log levels but are stationary at first 
differences .This means that the order of integration of all variables is one. We proceed to determine 
the optimal lag for the disaggregated and the aggregated models respectively. For the disaggregated 
model, we determine whether or not there is first or higher order serial correlation in the initial model 
by performing the autocorrelation LM test. The test results are presented below 
Table 3  
  

Included observations: 37  

Lags  LM-Stat  Prob  

1  39.13263  0.8424  

2  48.47734  0.4942  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

12 | P a g e  

From the results generated , we  fail to reject the null hypothesis that there is no serial correlation in 
the model  meaning that the model is can be correctly specified by using the first lags of all variables.   
We proceed to corroborate the above finding by determining the optimal lag structure of the model 
using various criteria. The table below shows the selected optimal lag structure using various criteria 
for the disaggregated model.  
Table 4  
  

Included observations: 36    

 Lag  LogL  LR  FPE  AIC  SC  HQ  

       

              

  
0  

  
 39.11941  

  
NA  

  
  3.96e-
10  

  
 -
1.784412*  

  
  -
1.476505  

  
 -1.676944  

1   
70.04318  

 
48.10365*  

 1.14e-
09*  

-
0.780177  

 
1.683075*  

 
0.079564*  

2   114.0449   51.33531   2.08e-
09  

-
0.502493  

 4.116104   1.109521  

3   170.9768   44.28037   3.70e-
09  

-
0.943154  

 5.830788   1.421133  

       
  * indicates lag order selected by the criterion   

       
    

 LR: sequential modified LR test statistic (each test at 
5% level)  

    

  
The results confirm that appropriate lag to be used in the analysis is one considering that four out of 
the five criteria settle on lag one. In the case of the aggregated model, the test for 1st and 2nd order serial 
correlation is presented in the table below.  
Table 5  
  

Null Hypothesis: no serial 
correlation at lag order h  

Sample: 1978 2017    

Included observations: 38  

Lags  LM-Stat  Prob  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

13 | P a g e  

   

      

  
1  

  
 47.48022  

  
 0.0955  

2   50.66403   0.0533  
  
From these statistics, we conclude therefore that in the aggregated model, there is evidence of serial 
correlation in the residuals at lags one and two so we proceed to determine the optimal structure for 
the model. The following table shows the selection analysis.  
Table 6  
  

Included observations: 36    

 Lag  LogL  LR  FPE  AIC  SC  HQ  

0   26.83531  NA    1.27e-08  -1.157517   -
0.893598*  

 -
1.065402*  

1   59.96044   53.36827   1.53e-08  -
0.997802  

 0.849636  -0.352997  

2   101.9825   53.69481   1.30e-08  -1.332359   2.098598  -0.134863  

3   156.1070    
51.11764*  

  7.91e-
09*  

 -
2.339279*  

 2.675197  -
0.589093  

 * indicates lag order selected by the 
criterion  

      

 LR: sequential modified LR test statistic (each test at 
5% level)  

    

  
From the results shown in the table above, we firmly conclude that the optimal lag for the aggregated 
model is three based on the different criteria. Having completed the tests for stationarity and optimal 
lag structures for the two models we then enter the log levels of the variables in the two models into the 
Johansson test for co integration, the results of which are presented in the tables below. 
Table 7 Johanssen test for co integration for the disaggregated model Series: lgc lgk ldt 
lit lfb ldb lgr  

Unrestricted Cointegration Rank Test (Trace)      
Hypothesized    Trace  0.05    
No. of CE(s)  Eigenvalue  Statistic  Critical 

Value  
Prob.**  

None *   0.716509   146.2219   134.6780   
0.0087  

At most 1   0.623665   98.31997   103.8473   0.1098  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

14 | P a g e  

At most 2   0.424277   61.18352   76.97277   0.4275  
At most 3   0.318226   

40.20262  
 54.07904   

0.4603  
At most 4   0.256078   

25.64646  
 35.19275   

0.3620  
At most 5   0.187746   

14.40532  
 20.26184   

0.2625  
At most 6   0.157301   

6.503534  
 9.164546   0.1553  

 Trace test indicates 1 cointegratingeqn(s) at 
the 0.05 level  

        

 * denotes rejection of the hypothesis at the 
0.05 level  

        

 **MacKinnon-Haug-Michelis (1999) p-values      
Sourceː Generated from E-Views.  
  
From the results presented above, we reject the hypothesis that there is no co integration in our series 
in favour of the alternative hypothesis that there is one co integrating equation in our model. Using the 
un-normalized coefficients, we derive the long run equation for government consumption expenditure 
below by normalizing on government consumption expenditure.   
Table 8 Long run equation for government consumption expenditure  

LGC  LGK  LDT  LIT  LFB  LDB  LGR  C  

 
1.000000  

 
0.265024  

 0.945707   4.311506  -
2.071946  

 0.967689  -1.537559   
38.95029  

   
(0.32764)  

 
(0.58921)  

 
(0.72563)  

 
(0.50301)  

 
(0.43209)  

 
(0.30392)  

 
(6.55926)  

Sourceː Generated from E-Views. 
From the long run equation, we observe that government capital expenditure, direct taxes, indirect 
taxes as well as domestic borrowing have negative effect on government consumption expenditure with 
about 27%,95%, 431%and 97%  impacts respectively on government consumption expenditure with  a 
100%  increase in  each of the variables. However, in the long run , borrowing from abroad and grants 
are financing sources which have  positive impact on government consumption expenditure. 
Specifically, a 100% increase in external borrowing in the long run triggers a little over 207% increase 
in government consumption expenditure whilst a 100% increase in grants also leads to a 154% upswing 
in government consumption expenditure. In the table below, we present the results of the tests for co 
integration in the aggregated model.   
 
 
 
 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

15 | P a g e  

Table 9: Test for Co-Integration in the Aggregated Model    
Series: LGE LDT LIT LFB LDB LGR  

Hypothesized    Trace  0.05    
No. of CE(s)  Eigenvalue  Statistic  Critical Value  Prob.**  
None *   0.642654   127.6208   103.8473   0.0006  
At most 1 *   0.606512   88.51696   76.97277   0.0050  
At most 2   0.493073   53.07422   54.07904   0.0613  
At most 3   0.275454   27.25744   35.19275   0.2761  
At most 4   0.199710   15.01346   20.26184   0.2256  
At most 5   0.158281   6.547753   9.164546   0.1525  
 Trace test indicates 2 cointegrating eqn(s) at the 0.05 level    
 * denotes rejection of the hypothesis at the 0.05 level    

Sourceː Output generated by author from E-Views using data  
In the table above, we test the null hypothesis that there is no co integrating relationship in our series 
against the alternative hypothesis that there is at least one co integrating relationship. From the table 
above we fail to accept the hypothesis that there is at most one co integrating relationship but fail to 
reject the null hypothesis that there are most two co integrating vectors in our model. This thus means 
that in our series, we can uncover two co integrating relationships. In the table below, we present the 
first co integrating equation from our model.  
Table10. Long run function for aggregated government expenditure 

  

LGE  LDT  LIT  LFB  LDB  LGR  C   
 1.000000   7.033729   18.66572  -8.250564  -0.279982  -6.581200   259.1778  
   (2.36778)   (3.53366)   (1.53188)   (1.71301)   (1.37989)   (30.0200)  

Sourceː Generated by author using E-Views estimation.   
From the results, we determine that in the long run, direct and indirect taxes negatively impact on 
government expenditure whereas external borrowing, domestic borrowing and grants   exert a positive 
effect on government expenditure. The estimated negative long run impacts of direct and indirect taxes 
on government expenditure are respectively 7.03 and 18.67 units as each of these increases by a unit. 
On the other hand, a unit increase of each of external borrowing, domestic borrowing and grants leads 
to about 8.3, 0.28 and 6.58 units increase in government expenditure. We derive the second co 
integrating equation from the un–normalized co integrating coefficients by normalizing on external 
borrowing .We thus derive the long run equation for external borrowing in the form below;  
Table 11.Long run equation for external borrowing  
  
  LGE                 LDT                LIT                 LFB                   LDB                     LGR                   C  
-0.346960            0.850657       0.668850         1.000000           -1.277610               0.027437            -
13.132143  
Sourceː Output generated by author based on E-Views estimations.  
From the table, we define the long run equilibrium relationship between external borrowing and the 
endogenous variables. In this relationship, we observe that government expenditure and domestic 
borrowing exert positive effects on external borrowing meaning that in the long run an increase in both 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

16 | P a g e  

government expenditure and domestic borrowing lead to an increased external borrowing. Generally 
from the estimation results, a 100% increase in government expenditure leads to about 35% increase 
in external borrowing whilst a 100% increase in domestic borrowing calls forth a whopping 128% 
increase in external borrowing. On the other hand, direct taxes, indirect taxes as well as grants 
expectedly all impact negatively on external borrowing. More specifically, in the long run a 100% 
increase in each of direct taxes, indirect taxes and grants precipitates about 85%,67% and 3% decline 
in external borrowing. Using tables 8,10 and 11 ,  we derive the error correction terms ect1,ect2 and ect3 
respectively which are entered into the short run models  to determine the short run effects of each of 
the endogenous variables on the other endogenous variables in tour system.   
Short Run Relationships  
Proceeding with our analysis, we estimate the short run/error correction models for the government 
consumption expenditure, aggregated government expenditure and external borrowing (the estimates 
are provided in the appendices). In these models, we observe the signs of the error correction terms are 
all negative meaning that the behaviours of the short run equations are in line with the theory that once 
these are co integrated then there is a tendency for each of them to be moved towards the desired 
equilibrium position.; that is each system is eventually drawn towards the equilibrium time path when 
there is a deviation from their expected long run position. Of the three models, the equation for 
aggregated government expenditure is estimated to have the fastest return to its equilibrium time path 
after a deviation with a speed of adjustment of about 81% per period. This followed by the equation for 
external borrowing with a speed of adjustment of about 43% per period when it deviates from the 
equilibrium .The government consumption equation however has about 20% of  its deviation from the 
long run corrected in  each period. In the general government consumption equation,   our short run 
estimates show that the government consumption expenditure is significantly impacted by a feedback, 
growth in government capital expenditure, direct taxes, domestic borrowing and grants. Their 
contemporaneous effects are estimated at -0.701589, 0.564860, 0.3202703,0.223057 and 0.203551 
respectively. This shows that previous period government consumption expenditure tends to have a 
negative impact on current government spending on consumption.   
It is also noticed from the estimation that the previous government capital expenditure has a positive 
effect on current government consumption expenditure. This result contrasts with Osei et al (2003). In 
actual terms, from the results, a 1 unit increase in the previous period government consumption 
expenditure triggers about 0.56 unit increase in current government consumption expenditure.    
Lastly a previous increase in grants precipitates an increase in current government consumption 
expenditure with a 100% previous increase in grants leading to a 20% increase in the current values of 
general government consumption expenditure. In the aggregated government expenditure function, 
just as is witnessed in the consumption expenditure equation registers a negative feedback with a 
magnitude of 0.134650 per unit increase in government expenditure. The only difference is that in the 
case of the aggregated government expenditure function, the feedback comes from the third period. 
Also, growth in direct taxes and indirect taxes respectively exert positive and negative effects on 
government expenditure with contemporaneous impacts of about 1.15 and 0.40 when there is a unit 
increase in each of them. Thus the dynamic effects of domestic revenue from these results appear 
mixed, and therefore do not fall wholly in tandem with the finding of Njeru (2004).   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

17 | P a g e  

In relation to external borrowing, the short run behaviour is explained by growth in aggregated 
government expenditure, direct taxes, indirect taxes as well as grants. In the estimates, it is seen that a 
100% increase in government expenditure in the first period expectedly leads to about 35% increase in 
current levels of external borrowing. From the short-run equation, it is also realized that in the third 
period, increased direct taxes leads to a decline in external borrowing. From the results, a 100% increase 
in direct taxes tends to lead to about 92% decline in external borrowing which falls in line with 
expectation that increased domestic revenue mobilization leads to reduction in dependence on external 
sources of financing projects and programmes. The indirect tax variable, in the period also elicits a 
negative response from external borrowing. The measured effect, significant in the first period is even 
bigger in magnitude than that of direct taxes. In real terms, a 100% growth in indirect taxes precipitates 
over 138% decline in external borrowing. The effect of growth in grants on external borrowing is felt in 
two periods-the first and third periods and in both periods their impacts are positive. In the first period, 
a 100% growth in grants tends to increase external borrowing by about 105% whereas in the third 
period, the effect is smaller at 0.31 unit’s growth in external borrowing with respect to a unit increase 
in grants. One major position which is dominant in the literature that we wanted to verify was whether 
or not the availability of other sources of financing government activities dampens tax effort.  
In the disaggregated government expenditure model, we are unable to substantiate the hypothesis that 
external financing tends to stunt domestic mobilization of revenue. Our regression results indicate that 
the impact of external borrowing and grants are positive and negative respectively. Thus for 100% 
increase in external borrowing, we experience about 33% increase in direct taxes but the same amount 
of increase in grants precipitates a 4% decline in direct taxes. For the aggregated government 
expenditure, the story is similar that is positive and negative in respect of external borrowing and grants 
respectively. The impacts of external borrowing and grants on indirect taxes are mostly insignificant 
except in the aggregated expenditure model in which growth in external borrowing triggers a decline in 
indirect tax yield. These findings are partly consistent with Mascagni and Timmis (2014) who 
discovered positive but significant impacts of grants and loans on the tax revenue variable.In the 
aggregated model, growth in external borrowing rather than leading to a decline in growth in direct tax 
mobilization actually triggers an increase. This finding coincides with Osei et al (2003). From the 
estimated equation, a 100% growth in external borrowing in the second period precipitates about 42% 
growth in direct taxes in the current period. However its estimated effect on indirect taxes is negative 
.The estimates indicate that a 100% increase leads to about 24% decline in indirect tax. The effect of 
domestic borrowing variable on the revenue variables- direct and indirect tax is very interesting. In the 
aggregated government expenditure models, we notice a negative impact of domestic borrowing on 
both direct and indirect tax variables. However whilst its effect is negative and significant with respect 
to growth in direct taxes, the measured impact is not significant in the case of indirect tax. From the 
estimated restricted VAR, a 100% growth in domestic borrowing elicits about 42% decline in direct 
taxes. On the other hand, for the disaggregated government models, the effect of domestic borrowing 
on both direct and indirect taxes is in line with the results for the aggregated model  , meaning that  an 
increased growth in domestic borrowing also impacts negatively on both direct and indirect taxes. 
Finally the effects of grants on the revenue channels-both direct and indirect taxes are estimated to be 
negative .Whilst its effect on direct taxes are significant that on indirect tax is insignificant. 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

18 | P a g e  

Results of Forecast Variance Decomposition 
In line with conventional dynamic analysis, we proceed to do a variance decomposition of the residuals 
of the variables and the results can be gleaned from the appendices of the paper. In dynamic analysis, 
variance decomposition is particularly very relevant in determining how much of the variation in a 
given variable can be traced to own innovations and innovations from other variables. The 
decompositions are performed on the basis of the aggregated and disaggregated government 
expenditure and in consonance with the Cholesky approach which ensures that the decomposition is 
carried out maintaining the ordering of the variables just as pertains in the co integration test as well 
as the error correction estimations.  
In the aggregated government model, we determine the most important innovations for attaining a 
particular objective for the various variables the aggregate government expenditure, direct taxes, 
indirect taxes, external borrowing, domestic borrowing and grants. From the results generated, it is 
clear that in respect of government expenditure, from the second period, growth in grants assumes a 
very important position in accounting for over 50% of the behaviour of the government expenditure 
variable. In the long run, it accounts for close to 70% of the movements of the government expenditure 
variable. For direct taxes, own innovations are largely responsible for its variations in the short to the 
long term accounting for over 85% of its movements. The next most variables are indirect taxes and 
growth in government expenditure which between them from the short to long term explain more than 
30% of the movements in the direct tax variable. From the variance decomposition of the indirect tax 
variable, its movements in the short term are dominated by own innovations and that from direct taxes. 
However in the medium to the long term the most important variable that influences movements in 
indirect tax is growth in grants. In respect of growth in external borrowing under the aggregated model, 
in the short to the medium term, own innovations are largely responsible for its behaviour though in 
the long run, growth in grants assumes the most dominant position accounting for just over 39% of 
variations in external borrowing. For domestic borrowing, in the short to the medium term, its 
variations are explained mainly by own innovations and that from external borrowing accounting for 
over 90% to about 30% between them. In the long term, however, growth in grants becomes the most 
dominant as it caters for over 54% of variations in the domestic borrowing variable.  
For grants, its own innovations are most dominant in explaining its movements from about 51% in the 
first period to over 60% in the tenth period. In the short to the medium term, however, the growth in 
external borrowing is second most important innovation which affects movements in grants. We now 
consider the forecast error decomposition in the disaggregated government expenditure model. From 
the derived results, we observe that from the short to the long term, the important variable that explains 
the behaviour of government consumption expenditure is own innovations which constitutes 100% to 
60% of its movements from the short to the long term. It is followed in terms of significance by the 
innovations due to government capital expenditure. In respect of government capital expenditure, the 
movements are mostly explained by own innovations and that emanating from government 
consumption expenditure.  The movements in direct taxes are dominated by own innovations from the 
short to the long term accounting for  over 99% to  about 78% whilst that due to domestic borrowing  
takes about 12% of the innovations. The contributions of the various innovations to the movements in 
the indirect tax variable are mainly due to own movements and those that coming from direct taxes. 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

19 | P a g e  

Specifically, own innovations account for over 63% to about 58% from the short to the long period 
whereas proportion of innovations from direct taxes range from 34% to 30%. Again movements in 
external borrowing are dominated by own innovations and predictably followed by innovations due to 
grants.  
Impulse Response Functions  
Finally in our analysis, we attempt to trace the effects of shocks emanating from the other variables in 
the system on each endogenous variable. (Estimations are found in the appendices) We first consider 
the aggregated government expenditure model. In respect of total government expenditure we realize 
that its time path around equilibrium is not very much affected by own shocks and that emanating from 
the other variables. However shocks coming from own innovations and from other variables cause more 
instability in the trajectory of direct tax variable around the equilibrium path. The instability as 
witnessed from the graphs is more pronounced especially in response to own shocks and the shocks 
which originate from total government expenditure and indirect taxes. The greatest effect of any shock 
in the system on indirect taxes comes from grants. However, the trajectory of external borrowing is 
affected much more by shocks from grants and then by own shocks than shocks coming from any other 
variable in the system.   
For domestic borrowing, apart from shocks triggered from grants the other shocks appear not to have 
any significant drift in its time path. Finally movements in the grants are largely unaffected by shocks 
which are transmitted from other variables. It is only own shocks which appear to drift the trajectory 
of grants from the equilibrium position. In the disaggregated government expenditure model, the story 
is different from that which is experienced in the aggregated expenditure model. From the graphs, we 
observe that shocks from government capital expenditure aside of own shocks are those which have 
more impact on the movement of government consumption expenditure. The time path of government 
capital expenditure is affected more in the early periods by shocks from government consumption 
expenditure and own shocks. The shocks from the other variables do not cause as much trepidation. In 
respect of  direct taxes own shocks are the most prominent among all the shocks which are transmitted 
from the various  variables whilst indirect taxes react to own shocks and that which emanates from 
direct taxes. It is also observed that the effects of shocks from government capital expenditure are 
noticeable only in the early period of the time horizon.  
Conclusions and Policy Implications  
In this study, our major preoccupation has been to establish the nexus between total government 
expenditure and disaggregated government expenditures and their corresponding financing modes, 
particularly focusing on the effects of foreign aid well as the response from domestic borrowing. Its 
import has been to verify whether the theoretical precepts established in the fiscal response models 
found in the literature still hold true for the Ghanaian economy using current data available. In our 
analysis we have generally found that whether government expenditure is aggregated or disaggregated, 
there exists one or other long run equilibrium relationship between expenditure and other variables in 
the model. More specifically, in the disaggregated government expenditure model, we have found that 
there only one co integrating equation exists between government consumption expenditure and other 
variables – government capital expenditure, direct taxes, indirect taxes, external borrowing, domestic 
borrowing and grants whereas in the model involving aggregate government expenditure, we 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

20 | P a g e  

discovered two co integrating equations-one for government expenditure and the other for external 
borrowing. In the long run, we find that external borrowing and grants lead to increased government 
consumption expenditure but government capital expenditure negatively impacts on government 
consumption expenditure. The positive effect of external borrowing and grants on government 
expenditure may point to aid fungibility though that conclusion may be erroneous or flawed on the 
grounds that some aid and grants come in the form of budgetary support and are therefore legitimately 
channeled into those areas of government spending which are important in the government's scheme 
of things.  
In respect of the aggregate model, external borrowing, domestic borrowing and grants all in the long 
run lead to increase in government expenditure which confirms concept of aid illusion but surprisingly 
the domestic revenue streams –direct and indirect taxes trigger a negative response from government 
expenditure. The estimated long run equation for external borrowing also shows that increased 
government expenditure precipitates increased external borrowing. Domestic borrowing also has the 
same effect but direct taxes, indirect taxes and grants all exert a negative effect on external borrowing. 
The positive effect of domestic borrowing on external borrowing probably gives the indication that 
because of the inadequacy of the domestically mobilized revenues, external and domestic borrowings 
have become an important but constant feature of financing government activities. Thus in the long 
run, in the disaggregated model we were able to adduce evidence of domestic revenues being used to 
replace external borrowing as a financing avenue. In the literature there is an opinion which articulates 
the view that external borrowing leads to a lax attitude towards domestic revenue mobilization, usually 
characterized as the displacement hypothesis. This is partially affirmed by our results in the short run. 
This is because whilst the effect of external borrowing on direct taxes is positive in both aggregate and 
disaggregated expenditure models it leads to a decline in indirect taxes in the aggregate model and has 
an insignificant impact on indirect taxes in the disaggregated model. From the short run results, the 
external sources of government financing impact positively on the government capital expenditures 
and this implies these resources are going into areas of the economy which may be reproductive and 
thus helping to expand economic activities in the long run. In long run it is established that an increased 
external borrowing and grants lead to more than proportionate growth in aggregate government 
expenditure which suggests that these external financing channels come with local or counterpart 
funding components which also exert more pressure on government finances. To ease pressure on 
government, government would have to enter into external funding agreements which do not require 
too much of counterpart funding. One other view proffered by some economists in the literature is that 
governments in developing countries have a preference for grants than loans for financing projects and 
programmes.   
In our analysis, it is obvious that the effect of grants undermines direct tax collection and it does appear 
because grants are normally free, its increased flow into the Ghanaian economy dampens the direct tax 
collections. Policy makers are encouraged to continue design tax policies and mechanisms which would 
in spite of increased flow of grant enable the government to rake in the desired revenues.  
Another significant and illuminating finding is the fact that short run effect of domestic borrowing on 
both direct and indirect taxes is negative in the Ghanaian economy which signals that domestic 
borrowing may be inhibiting economic  activities and thus may ultimately be having a distortionary 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

21 | P a g e  

impact on tax collections through its effect on economic activities. The government would therefore do 
well to scale down on its appetite for borrowing from domestic sources which particularly has a 
strangulating effect on private sector activities and ultimately impacts negatively on economic activities 
that generate the revenue needed by government.  
One of the objectives of the study is to determine whether the borrowing modes have differential effects 
on the domestic tax channels and our estimated equations suggest that the tax channels do not response 
in the same way to borrowing. This therefore allows policy makers to design the relevant but right 
mechanisms to ensure continuous increased tax yields from both direct and indirect sources by creating 
unique mechanisms which work for each tax channel. Finally we also find that the short-run equation 
for domestic borrowing in the disaggregated government expenditure model shows that external 
borrowing is used to substitute domestic borrowing to certain extent and this has a huge implication 
for the Ghana's debt sustainability which has become a source of worry to international agencies and 
economic think-tanks within Ghana even against the background of a re-based economy. To conclude 
we would say that though we have through this study unearthed some important facts relating to the 
nexus among the fiscal variables and the borrowing modes in Ghana, we would have wished that we 
were able to segregate aid into the various other forms project, programme or even technical by which 
they come, which in our view would have enriched the analysis .It is therefore our hope that future 
studies would tackle this aspect to further add to the existing stock of knowledge in this area. Another 
area which may be interesting to examine in the future is the effects of these borrowing modes on 
private investments and economic growth.  
References  

Ali, A.A., G.,Malwanda,C., & Sliman, Y. (1999). Official development assistance to Africa: an overview. 
Journal of African Economies, 8 (4), 504-527.  

Anaman, E., A., Gadzo, S., G., Gatsi, J., G., & Pobbi, M. (2017). Fiscal aggregates government borrowing 
and economic growth in Ghana: an ever correction approach. Advances in Management and 
Applied Economics, 7 (2), 83-104.  

Bhasin, V.K. (2004). Dynamic interlinks among exchange rate price level and terms of trade ina 
managed floating exchange system: the case of Ghana. AERC Research Paper, 141. African 
Economic Consortium, Nairobi, Kenya.  

Blanchard, O., & Perrotti,R. (1999). An empirical characterization of dynamic effects of changes in 
government spending and taxes on output. NBER Working Paper, 7269.  

Bwire,T.,Lloyd, T., &Morrissey, O. (2017). Fiscal reforms and the fiscal effects of aid in Uganda. The 
Journal of Development Studies, 53 (7), 1019-1036.  

Deverajan, S., Rajkumar, A., S., & Swaroop, V. (1998). What does aid do to African finance? AERC/ODC 
Sponsorship on Managing a Smooth Transition from Aid Dependence in Africa, Washington 
DC.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

22 | P a g e  

Durdonoo, C. (2000). Fiscal trends: 1970-1995 in E.A. Aryeetey et al (eds.). Economic Reforms in 
Ghana: the miracle and the mirage. Trenton, USA: Africa Word Press.  

Enders,W. (1995). Applied econometric time series. New York: Wiley Press.  

Feyzioglu, T.,Swaroop, V., & Zhu, M. (1998). A panel data analysis of the fungibility of foreign aid. 
World Bank Economic Review, 65, 429-445.  

Griffin, K. (1970). Foreign capital domestic savings and economic development. Oxford Bulletin of 
Economics and Statistics, 55, 99-112.  

Heller,S., P. (1975). A model of public fiscal behaviour in developing countries: Aid, investment and 
taxation. American Economic Review, 65 (3), 429-445.  

Johassen.(1991). Estimation and hypothesis testing of co integration vectors in Gaussian vector                
autoregressive models. Econometrica Vol59ːpp 1551-1580  

Johnston, J.& Di-Nardo, J.(1997). Econometric methods. (4thed.). Singapore: McGraw Hill.  

Khilji,N.M.,& Zampelli, E., M. (1994). The fungibility of US military and non-military assistance and 
the impacts on expenditures of major aid recipients. Journal of Development Economics, 43, 
345-362.  

Lloyd, T.,McGillivray, M., Morrissey, O., & Opoku-Afari,M. (2009). The fiscal effects of aid in 
developing countries: a comparative dynamic analysis. Studies in Development Economics and 
Policy, 158-179.  

Lutkepohl, H., & Rimmers, H., E. (1992). Impulse response analysis of co integrated systems. Journal 
of Economic Dynamics and Control, 16, 53-78.  

M’Amanja, D.,Lloyd, T., &Morrissey, O. (2005). Fiscal aggregates, aid and growth in Kenya: a vector 
autoregressive (VAR) analysis. CREDIT Research Paper, 5 (7).  

Martins, P.M.G. (2010). Fiscal dynamics in Ethiopia: the co integrated VAR model with quarterly data. 
University of Nottingham CREDIT Research Paper, 10 (5).  

Mascagni,G., &Timmis, E. (2014). Fiscal effects of aid in Ethiopia: evidence from CVAR application. 
University of Nottingham CREDIT Research Paper, 14 (6).  

Mavrotas, G. (2002). Foreign aid and fiscal response: does aid disaggregation matter? 
Weltwirtschaftliches Archive, 138, 534-559.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

23 | P a g e  

McGillivray,M& Morrissey,O. (2000). Aid fungibility in assessing aid: red herring or true concern? 
Journal of International Development, 12 (3), 413-428.  

Mosley, P., J., Hudson, J., & Horrell, S. (1987). Aid and public sector and the market in less developed 
countries. Economic Journal, 97 (9), 616-641.  

Njeru, J. (2004). The impact of foreign aid on public expenditure: the case of Kenya. AERC Research 
Paper, 135. African Economic Research Consortium, Nairobi.  

Osei, R., Morrissey, O., & Lloyd,T. (2003). Modelling the fiscal effects of aid: an impulse response 
analysis for Ghana. University of Nottingham CREDIT Research Paper, 3 (10).  

Thomas, R.L. (1993). Introductory econometrics: theory of applications. (2nded.). London: Longman 
Press.  

Appendix A 1a  
Unnormalized Co Integrating Coefficients For Disaggregated Government Expenditure 
Model  

               
     LGC       LGK       LDT        LIT       LFB       LDB       LGR         C  
 1.294453   0.343061   1.224174   5.581041  -2.682037   1.252628  -1.990298   
50.41932  
 2.473469 -4.926923  3.323305  0.203476  0.760543  1.084518  1.902956 -9.314356 -2.734475  
2.252014  1.876378 -0.711541 -1.777021 -1.152836 -0.274655  58.36492  
 0.626448  0.708572 -0.757148 -3.873066 -2.603299  4.320096  1.164597 -2.573859 -3.721753  
2.645566  2.914102 -2.369617  1.228391 -1.890964 -1.548250  31.02504  
 1.690610  -1.846113   4.246725  -3.162973   0.280160  -0.497808   0.356670  -
6.716738  
 2.546556  -2.991196  -0.573203   2.166093   1.917874  -0.844341  -0.672078  -
18.44867  
                
  1b Unnormalized Co Integrating Coefficients For Aggregated Government 
Expenditure Model                
              
 LGE       LDT      LIT   LFB      LDB     LGR       C  
 0.331685   2.332985   6.191146  -2.736592  -0.092866  -2.182888   85.96550  
-1.536605  5.185054 -5.442559  1.073965  2.058536  3.170925 -61.33498  1.272844 -3.814032 -
0.244547 -3.236886  0.620246 -0.427752  60.13527  
-0.924257   2.266703   1.782250   2.664649  -3.404383   0.070304  -34.99253  
-0.416011  -2.639002   3.722309   1.874406   0.312871   0.073111  -38.92193  
 0.081428  -2.878044   0.715399  -0.521188   0.300898   0.115664   8.388109  

     
Appendix B  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

24 | P a g e  

Short-Run /Error Correction Estimates  
Short Run Estimates For The Disaggregated Government Expenditure Model  

  
Error 
Correction:   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  
ECT1(-1)   

  
-
0.019809   

  
-0.024172   

  
-0.001566   

  
-
0.018902   

  
0.001415   

  
-
0.005638   

  
0.022009   

  (0.00701)   (0.00619)   (0.00604)   (0.00679)   (0.00492)   (0.00699)   (0.02305)   
  [-

2.82725]   
[-
3.90412]   

[-
0.25948]   

[-
2.78290]   

[ 
0.28745]   

[-
0.80695]   

[ 
0.95466]   

  
DLGC(-1)   

  
-0.701589   

  
0.016095   

  
0.155862   

  
0.284686   

  
-
0.044776   

  
0.180555   

  
0.416556   

  (0.15231)   (0.13459)   (0.07841)   (0.14765)   (0.10699)   (0.07136)   (0.14713)   
  [-

4.60624]   
[ 0.11959]   [ 1.98767]   [ 1.92805]   [-0.41851]   [ 

2.53034]   
[ 2.83116]   

  
DLGK(-1)   

  
0.564860   

  
-
0.296333   

  
-0.125927   

  
-0.074017   

  
0.071090   

  
-0.217699   

  
-
0.206533   

  (0.17444)   (0.15415)   (0.15030)   (0.16911)   (0.12254)   (0.17396)   (0.57399)   
  [ 

3.23805]   
[-
1.92237]   

[-
0.83782]   

[-
0.43769]   

[ 0.58015]   [-1.25140]   [-
0.35982]   

  
DLDT(-1)   

  
0.320703   

  
0.606177   

  
-0.546285   

  
0.319727   

  
-
0.302274   

  
0.032411   

  
0.414471   

  (0.15278)   (0.28556)   (0.18010)   (0.16165)   (0.14998)   (0.20845)   (0.68779)   
  [ 

2.09905]   
[ 2.12277]   [-

3.03324]   
[ 1.97785]   [-

2.01549]   
[ 0.15549]   [ 

0.60261]   
  
DLIT(-1)   

  
0.186466   

  
-0.566537   

  
-0.001965   

  
-
0.346928   

  
-0.189941   

  
0.065452   

  
-
1.080978   

  (0.26140)   (0.22288)   (0.22522)   (0.17619)   (0.18362)   (0.26068)   (0.86010)   
  [ 0.71335]   [ -

2.54186]   
[-
0.00873]   

[-
1.96908]   

[-
1.03445]   

[ 0.25109]   [-
1.25680]   

  
DLFB(-1)   

  
0.098782   

  
0.050804   

  
0.331934   

  
-0.319583   

  
-
0.268403   

  
-
0.132600   

  
1.249801   

  (0.30483)   (0.02321)   (0.16903)   (0.29551)   (0.11911)   (0.05443)   (1.00302)   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

25 | P a g e  

  [ 
0.32405]   

[ 2.18861]   [1.96381]   [-
1.08146]   

[-
2.25347]   

[-
2.43620]   

[ 
2.24603]   

  
DLDB(-1)   

  
0.223057   

  
0.364305   

  
-0.307923   

  
-
0.430906   

  
0.076035   

  
-
0.093922   

  
-1.103910   

  (0.11371)   (0.20497)   (0.11830)   (0.21693)   (0.16293)   (0.03904)   (0.45124)   
  [1.96165]   [ 1.77738]   [ -

2.60286]   
[-
1.98634]   

[ 
0.46667]   

[-
2.40604]   

[ -
2.44639]   

  
DLGR(-1)   

  
0.203551   

  
-0.100615   

  
-
0.040873   

  
-0.100401   

  
0.120268   

  
0.041354   

  
-0.145643   

  (0.08956)   (0.07914)   (0.01616)   (0.08682)   (0.06099)   (0.08931)   (0.29467)   
  [ 2.27291]   [-1.27141]   [-2.52971]   [-1.15647]   [ 1.97183]   [ 

0.46305]   
[-
0.49425]   

  
C   

  
-0.001451   

  
0.002197   

  
0.001929   

  
0.000621   

  
0.001887   

  
-
0.000925   

  
-
0.063636   

  (0.04241)   (0.03747)   (0.03654)   (0.04111)   (0.02979)   (0.04229)   (0.13954)   
  
  

[-
0.03421]   
  

[ 
0.05863]   
  

[ 0.05279]   
  

[ 0.01511]   
  

[ 
0.06336]   
  

[-
0.02187]   
  

[-
0.45605]   
  

  
R-squared   

  
0.551149   

  
0.549121   

  
0.504434   

  
0.583416   

  
0.516661   

  
0.212850   

  
0.526910   

Adj. R-
squared   

0.422906   0.420298   0.362844   0.464393   0.378565   -
0.012050   

0.391742   

Sum sq. 
resids   

1.859653   1.452124   1.380543   1.747652   0.917605   1.849415   20.13418   

S.E. 
equation   

0.257713   0.227731   0.222047   0.249832   0.181029   0.257003   0.847985   

F-statistic   4.297695   4.262614   3.562631   4.901676   3.741300   0.946419   3.898175   
Log 
likelihood   

2.824038   7.400244   8.335439   3.973200   15.89205   2.926176   -41.24349   

Akaike AIC   0.333836   0.086473   0.035922   0.271719   -0.372543   0.328315   2.715864   
Schwarz 
SC   

0.725681   0.478318   0.427767   0.663564   0.019302   0.720160   3.107709   

Mean 
dependent   

-
0.003087   

0.001756   0.003245   -
0.001704   

0.000154   0.000452   -0.066715   

S.D. dependent   0.339245   0.299103   0.278178   0.341370   0.229642  
 0.255468   1.087286   
                

                  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

26 | P a g e  

  
Appendix C  
Short Run Estimates For The Aggregated Government Expenditure Model  
Error Correction:  DLGE  DLDT  DLIT  DLFB  DLDB DLGR  
              
              
ECT2(-1) -0.814967 -0.559246 -0.693817  0.972253  0.828267 -1.359340  
  (3.79677)  (0.44945)  (0.46205)  (0.52819)  (0.73398)  (2.44790)  
 [-3.09437] [-1.24428] [-1.50162] [ 1.84071] [ 1.12847] [-0.55531]  
              
ECT3(-1)  0.232244  0.012426  0.353103 -0.425366 -0.111076  0.397426  
  (0.17271)  (0.21543)  (0.22146)  (0.20943)  (0.35180)  (1.17329)  
 [ 1.34474] [ 0.05768] [ 1.59442] [-2.03106] [-0.31574] [ 0.33873]  
              
DLGE(-1)  0.054739  0.261523  0.573447  0.354344 -0.082371  0.867620  
  (0.17577)  (0.21924)  (0.22539)  (0.11625)  (0.35804)  (1.19409)  
 [ 0.31143] [ 1.19284] [ 2.54427] [3..04821] [-0.23006] [ 0.72660]  
              
DLGE(-2) -0.090568  0.205657  0.458363 -0.061591 -0.333092  0.248939  
  (0.16896)  (0.21075)  (0.21665)  (0.24767)  (0.34416)  (1.14782)  
 [-0.53604] [ 0.97584] [ 2.11564] [-0.24868] [-0.96783] [ 0.21688]  
              
DLGE(-3) -0.134650  0.035054  0.129480  0.087230  0.271188  0.231454  
  (0.06922)  (0.15817)  (0.16260)  (0.18588)  (0.13594)  (0.86147)  
 [-1.97187] [ 0.22162] [ 0.79629] [ 0.46927] [ 1.99489] [ 0.26867]  
              
DLDT(-1)  0.172514  0.050142  0.517520  0.256857  0.033832 -2.202289  
  (0.23221)  (0.28965)  (0.18901)  (0.34039)  (0.47301)  (1.10333)  
 [ 0.74293] [ 0.17311] [ 2.73802] [ 0.75459] [ 0.07153] [-1.99603]  
              
DLDT(-2)  1.148476  0.025518 -0.138434 -0.536571 -0.845334  2.062187  
  (0.25936)  (0.32352)  (0.33258)  (0.38020)  (0.52832)  (1.76201)  
 [ 4.42808] [ 0.07888] [-0.41624] [-1.41130] [-1.60004] [ 1.17036]  
              
DLDT(-3)  0.349772  0.069256  0.305151 -0.916470 -1.702434  3.984505  
  (0.35723)  (0.03213)  (0.45809)  (0.46590)  (0.72769)  (1.50826)  
 [ 0.97911] [ 2.15542] [ 0.66614] [-1.96709] [-2.33951] [ 2.64179]  
              
DLIT(-1)  0.553068  0.253030 -0.338736 -1.382197 -1.282164  2.899246  
  (0.37157)  (0.46348)  (0.47647)  (0.54468)  (0.64301)  (2.52431)  
 [ 1.48846] [ 0.54593] [-0.71093] [-2.53762] [-1.99399] [ 1.14853]  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

27 | P a g e  

              
DLIT(-2) -0.395099  0.330036  0.236805 -0.224872 -0.091448 -0.446796  
  (0.17676)  (0.16527)  (0.12177)  (0.25911)  (0.36005)  (1.20082)  
 [-2.23527] [ 1.99690] [ 1.94477] [-0.86787] [-0.25398] [-0.37208]  
              
DLIT(-3) -0.234890 -0.094442 -0.243058  0.104862  0.414603 -1.457794  
  (0.18283)  (0.22806)  (0.23445)  (0.26802)  (0.37243)  (1.24211)  
 [-1.28471] [-0.41411] [-1.03671] [ 0.39125] [ 1.11323] [-1.17364]  
              
DLFB(-1)  0.115739 -0.055537 -0.337443  0.523889  0.179581  0.007247  
  (0.36633)  (0.45694)  (0.46975)  (0.26518)  (0.74621)  (2.48870)  
  [ 0.31594]  [-0.12154]  [-0.71835]  [ 1.97559]  [ 0.24066]  [ 0.00291]        
       

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

28 | P a g e  

DLFB(-2)   0.224531   0.424402  -0.240696  -0.437316  -0.682534  
   (0.09086)   (0.16517)   (0.08399)   (0.31780)   (0.44162)  
  [ 2.47116]  [ 2.56939]  [-2.86581]  [-1.37606]  [-1.54553]  
            
DLFB(-3)   0.150271   0.119247  -0.043971   0.000501  -0.094436  
   (0.20228)   (0.25232)   (0.25939)   (0.29653)   (0.41205)  
  [ 0.74287]  [ 0.47260]  [-0.16952]  [ 0.00169]  [-0.22919]  
            
DLDB(-1)  0.337644  -0.223752   0.276325   0.308478   0.447370  
   (0.13910)   (0.29508)   (0.30334)   (0.34677)   (0.48187)  
  [2.42731]  [-0.75829]  [ 0.91093]  [ 0.88957]  [ 0.92840]  
            
DLDB(-2)  0.259235  -0.424442  -0.196132   0.022124  -0.162952  
   (0.18529)   (0.21368)   (0.23760)   (0.27162)   (0.37744)  
  [1.39905]  [-1.98639]  [-0.82545]  [ 0.08145]  [-0.43173]  
            
DLDB(-3)  0.087809  -0.126465   0.133979   0.311050   0.399951  
   (0.16787)   (0.20940)   (0.21527)   (0.24609)   (0.18434)  
   [ 0.52306]  [-0.60394]  [ 0.62238]  [ 1.26399]  [ 2.16959]  
            
DLGR(-1)  0.969935  -0.641608  -0.430083   1.049242   1.055592  
   (0.36425)   (0.45435)   (0.46708)   (0.53395)   (0.74197)  
  [2.66285]  [-1.41215]  [-0.92079]  [ 1.96507]  [ 1.42269]  
            
DLGR(-2)  0.505786  -0.464436  -0.363172   0.510144   0.410434  
   (0.23730)   (0.23587)   (0.30429)   (0.34786)   (0.48338)  
  [2.13141]  [-1.96904]  [-1.19349]  [ 1.46653]  [ 0.84909]  
            
DLGR(-3)  0.070551  -0.231593  -0.088270   0.314585   0.177093  
   (0.12391)   (0.15457)   (0.15890)   (0.11515)   (0.25241)  
  [0.56935]  [-1.49833]  [-0.55551]  [ 2.73185]  [ 0.70159]  
            
C   0.004849   0.012664   0.020949  -0.005055  -0.014179  
   (0.02025)   (0.02526)   (0.02597)   (0.02969)   (0.04125)  
  [ 0.23941]  [ 0.50132]  [ 0.80665]  [-0.17026]  [-0.34369]  
            

 
            
 R-squared   0.927544   0.778822   0.895330   0.783659   0.654236  
 Adj. R-squared   0.824036   0.462853   0.745801   0.474601   
0.160288  

 1.368240  
 (1.47284)  
[ 0.92898]  
  
 2.359789  
 (1.18157)  
[ 1.99717]  
  
 0.735352  
 (1.60710)  
[ 0.45756]  
  
 0.261685  
 (1.25881)  
[ 0.20788]  
  
-0.663780  
 (1.14047)  
[-0.58202]  
  
-2.475217  
 (2.47456)  
[-1.00027]  
  
-1.610878  
 (1.61213)  
[-0.99922]  
  
-0.333660  
 (0.84183)  
[-0.39635]  
  
-0.076360  
 (0.13759)  
[-0.55499]  
  

 
  
 0.790353  
0.490856  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

29 | P a g e  

 Sum sq. resids   0.189327   0.294576   0.311316   0.406836   
0.785588  
 S.E. equation   0.116290   0.145056   0.149120   0.170469   
0.236883  
 F-statistic   8.961041   2.464869   5.987662   2.535638   1.324503  
 Log likelihood   41.68061   33.94453   32.97730   28.29429   
16.77889  
 Akaike AIC  -1.181749  -0.739688  -0.684417  -0.416817   0.241207  
 Schwarz SC  -0.248541   0.193521   0.248792   0.516392   1.174415  
 Mean dependent   0.002435   0.012086   0.015653   0.003679  -
0.002856  
 S.D. dependent   0.277223   0.197919   0.295767   0.235180   
0.258505  
            

8.738089  
0.790031  
 2.638939  
-25.37886  
 2.650220  
 3.583429  
-0.076870  
 1.107195  
  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

30 | P a g e  

  
  
  
  
  
                
Variance Decomposition of DLGE:  
 Period  S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR  
                

 
                
1 0.111813 100.0000 0.000000 0.000000 0.000000 0.000000 0.000000 
2 0.265269 23.43584 20.37757 3.577401 0.422971 0.413212 51.77301 
3 0.385462 15.65435 12.85675 2.279425 1.161694 0.582181 67.46560 
4 0.444459 15.27655 11.13926 1.777144 0.909225 4.016351 66.88147 
5 0.480041 18.49762 9.680806 2.009450 0.842194 3.446230 65.52370 
6 0.509495 18.37302 9.024340 2.928824 2.621776 3.661696 63.39034 
7 0.530525 18.10874 8.592325 2.945605 2.653703 3.634360 64.06527 
8 0.592717 16.41770 9.398415 2.439467 2.725268 3.237613 65.78154 
9 0.645311 16.74329 8.484315 2.063404 2.578772 3.102590 67.02763 
10 0.707590 17.26560 7.260408 2.026676 2.782769 2.649450 68.01510 
                

 
                
 Variance Decomposition of DLDT:               
 Period   S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR   
                

  1                 
 0.135609 14.67399 85.32601 0.000000 0.000000 
0.000000 0.000000 

 2    0.168781 18.62678 57.94360 15.88752 2.083583 
5.320575 0.137940 

 3    0.171894 18.51302 56.19375 15.58305 2.923035 
6.649722 0.137425 

 4    0.187967 15.56514 56.71454 16.41705 2.446050 
6.768987 2.088241 

 5    0.196661 15.84631 52.83267 19.01200 2.663066 
6.243143 3.402820 

 6    0.201969 15.12285 50.18036 18.64150 3.122657 
5.935793 6.996830 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

31 | P a g e  

 7    0.202326 15.12402 50.07573 18.69772 3.124804 
5.957755 7.019973 

 8    0.206023 14.68438 49.98696 18.16153 3.086644 
7.243165 6.837325 

 9    0.207504 15.14699 49.39309 18.35866 3.175843 
7.150039 6.775385 

 10   
  

 0.210232 15.37626 48.35545 17.94419 3.096044 
6.967732 8.260324 
              

  
 Variance Decomposition of 
DLIT:   

              
              

 Period   
  

S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR   
              

  1                 
 0.172417 1.586740 48.92190 49.49136 0.000000 
0.000000 0.000000 

 2    0.222251 6.532833 39.76821 36.29187 2.542720 
4.247651 10.61671 

 3    0.240533 5.666080 42.98798 32.38567 2.495584 
3.642069 12.82262 

 4    0.282419 7.145570 32.56612 24.49860 1.925561 
2.648976 31.21517 

 5    0.305613 7.189801 30.23615 20.92503 1.708072 
3.511665 36.42929 

 6    0.324604 9.538451 27.59595 18.58652 1.582996 
3.176390 39.51969 

 7    0.340375 11.50168 25.84900 19.30786 1.613194 
2.956258 38.77200 

 8    0.349941 12.00197 24.45653 18.30123 1.805046 
3.433497 40.00173 

 9    0.372219 12.16112 23.06242 16.47424 2.106031 
3.147840 43.04836 

 10   0.392478 12.20646 21.59352 14.82943 2.053484 3.118234 46.19888 
                

                    
Variance Decomposition of DLFB:  
 Period  S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR  
                

 
                

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

32 | P a g e  

1 0.158397 0.044068 1.797966 0.832714 97.32525 0.000000 0.000000 
2 0.177824 0.283293 10.59105 4.897140 78.39354 0.021085 5.813892 
3 0.231639 1.864778 7.092211 2.953018 60.72698 1.889828 25.47319 
4 0.245491 1.801467 6.655475 7.144054 56.78387 3.317991 24.29714 
5 0.256219 3.597538 6.326243 7.361394 52.13185 3.170792 27.41218 
6 0.265092 5.934772 5.950920 7.822540 49.15187 3.312355 27.82755 
7 0.276468 6.407326 6.361810 7.229981 45.20437 3.128504 31.66801 
8 0.292901 7.634311 6.580809 6.595101 40.28310 3.123580 35.78309 
9 0.302819 8.628087 6.568402 6.291867 37.78864 3.198804 37.52420 
10 0.311030 9.505549 6.228335 6.097690 36.03080 3.074052 39.06357 
                

 
                
 Variance Decomposition of DLDB:               
 Period  S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR   
                

 
                
1 0.206871 2.265834 0.810045 2.773664 22.89687 71.25359 0.000000 
2 0.223675 2.783841 6.202074 4.149667 19.80236 64.21923 2.842822 
3 0.288792 3.716311 3.763684 2.544012 22.32057 38.78066 28.87476 
4 0.338815 12.26334 2.735667 1.852837 16.64807 31.38268 35.11741 
5 0.401912 19.12568 2.211283 6.381911 11.92865 23.45652 36.89596 
6 0.441166 19.68044 2.240487 6.544079 11.51591 19.74382 40.27526 
7 0.461121 19.85563 3.003299 5.993447 10.54970 18.09470 42.50323 
8 0.502054 17.95271 5.140998 5.121903 9.264051 15.39093 47.12941 
9 0.537271 16.97087 5.065356 4.512318 9.047540 13.47473 50.92919 
10 0.574863 16.64304 4.665101 3.945109 8.396938 11.83882 54.51099 
                

 
                
 Variance Decomposition of DLGR:               
 Period  S.E.  DLGE DLDT  DLIT  DLFB  DLDB DLGR   
                

 
                
1 0.755211 0.207621 2.294906 4.433442 39.47587 2.841255 50.74691 
2 0.892897 5.268719 2.287596 3.424055 40.66342 10.57656 37.77965 
3 1.118448 6.441865 2.333939 4.617557 28.37545 6.795336 51.43585 
4 1.162192 11.41789 2.963043 4.304557 26.28069 6.318548 48.71527 
5 1.250068 12.21659 4.194314 4.814527 22.72084 5.506416 50.54731 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

33 | P a g e  

6 1.335245 12.91167 4.548363 4.353771 20.51086 4.894152 52.78119 
7 1.416154 13.08263 5.335398 3.877774 18.46247 4.949913 54.29181 
8 1.525871 13.21428 4.937466 3.401963 16.37939 4.283290 57.78361 
9 1.618501 13.31405 4.747701 3.062418 15.41920 3.826015 59.63062 
10 1.693009 14.17167 4.431628 2.800191 14.16598 3.500881 60.92966 
                

 
             
Variance Decomposition of Disaggregated Model  

  

                  

    
 Variance  
Decomposition of 
DLGC:     

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  
  

S.E.   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  1    
 
0.21708
5   

  
 
100.000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

 2   
0.26964
9   

 
65.0425
8   

 23.12333    1.721511    
5.952523   

 1.337863    
2.821443   

 
0.00074
6   

 3   
0.27782
6   

 
61.27276   

 21.81767    
1.692895   

 
7.846924   

 1.277528    
3.468516   

 
2.623707   

 4   
0.27882
7   

 
60.9569
0   

 
21.66242   

 1.705474    
7.793662   

 1.791860    
3.467854   

 
2.621829   

 5   
0.279151   

 
60.8222
2   

 21.63173    
1.708604   

 7.792124    1.787999    
3.499983   

 
2.757338   

 6   
0.27924
8   

 
60.78512   

 21.62981    1.708247    
7.786927   

 1.810778    3.523712    2.755411   

 7   
0.27926
2   

 
60.7797
8   

 
21.62889   

 
1.708456   

 7.786194    1.810644    3.527721    2.758316   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

34 | P a g e  

 8   
0.27926
6   

 
60.77815   

 
21.62924   

 
1.708446   

 
7.786028   

 1.811031    
3.528838   

 
2.758263   

 9   
0.27926
6   

 
60.7780
3   

 21.62921    
1.708450   

 
7.786039   

 1.811035    
3.528958   

 
2.758272   

 10  
  

 
0.27926
6   
  

 
60.7780
0   
  

 
21.62923   
  

 
1.708450   
  

 
7.786037   
  

 1.811035   
  

 
3.528974   
  

 
2.758273   
  

  
 Variance  
Decompositio
n of DLGK:   

  
  

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  
  

S.E.   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  1    
 
0.22126
0   

  
 
28.6533
7   

  
 71.34663   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

 2   
0.25624
1   

 
22.7309
3   

 
54.92869   

 2.311531    17.06961    
0.387709   

 
0.28668
8   

 
2.284841   

 3   
0.25960
9   

 
22.6632
9   

 
53.98672   

 
2.253508   

 16.75220    
1.325450   

 
0.502436   

 2.516391   

 4   
0.26065
2   

 
22.49917   

 
53.67423   

 
2.237255   

 16.72579    1.333753    
0.538337   

 
2.991464   

 5   
0.26092
9   

 
22.4750
5   

 
53.57948   

 
2.234198   

 16.69179    1.418543    0.615461    
2.985473   

 6   
0.26098
4   

 
22.4669
2   

 53.56351    
2.234675   

 
16.68504   

 
1.418000   

 
0.630732   

 3.001125   

 7   
0.26100
0   

 
22.46491   

 
53.56006   

 
2.234580   

 
16.68338   

 1.420510    
0.635649   

 
3.000918   

 8   
0.26100
1   

 
22.4647
0   

 
53.55959   

 
2.234606   

 
16.68323   

 1.420555    
0.636214   

 
3.001109   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

35 | P a g e  

 9   
0.26100
2   

 
22.4646
3   

 
53.55954   

 
2.234606   

 16.68319    1.420579    
0.636324   

 3.001121   

 10  
  

 
0.26100
2   
  

 
22.4646
3   
  

 
53.55953   
  

 
2.234607   
  

 
16.68320   
  

 
1.420583   
  

 
0.63633
0   
  

 3.001121   
  

  
 Variance 
Decompositio
n of DLDT:   

  
  

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  S.E.   DLGC   DLGK   DLDT   DLIT   DLFB   DLDB  
 DLGR   
                  

                  
1 0.177084  0.109107  0.077847  99.81305  0.000000  0.000000  0.000000  0.000000  
2 0.200855  1.745636  2.080662  80.25215  0.882415  0.001371  12.52711  2.510653  
3 0.202226  1.821304  2.507557  79.17233  1.231177  0.137470  12.52452  2.605643  
4 0.202403  1.852392  2.556090  79.03863  1.231784  0.193493  12.52652  2.601091  
5 0.202479  1.851029  2.572860  78.98032  1.242394  0.193999  12.52149  2.637904  

6 0.202498  1.852722  2.572986  78.96633  1.243655  0.201673  12.52373  2.638906  
7 0.202501  1.852663  2.573234  78.96395  1.243682  0.202162  12.52383  2.640481  
8 0.202502  1.852679  2.573349  78.96325  1.243719  0.202460  12.52393  2.640620  
9 0.202502  1.852677  2.573347  78.96319  1.243719  0.202496  12.52392  2.640650  
10 0.202502  1.852677  2.573353  78.96317  1.243719  0.202502  12.52392  2.640656  
                  

    
 Variance  
Decomposition of DLIT:  
   

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  
  

S.E.   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  1    
 
0.21264
6   

  
 
1.939034   

  
 
0.55468
2   

  
 
34.35595   

  
 
63.15033   

  
 
0.00000
0   

  
 
0.00000
0   

  
 
0.00000
0   

 2   
0.23307
0   

 
5.509995   

 
0.50597
8   

 
30.8045
7   

 
59.4287
0   

 0.027701    
3.722568   

 
0.00049
2   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

36 | P a g e  

 3   
0.23484
3   

 
5.627085   

 
1.376441   

 
30.35335   

 
58.5369
3   

 0.222132    
3.827006   

 
0.057054   

 4   
0.23528
6   

 
5.608014   

 
1.415256   

 
30.2403
0   

 
58.3827
9   

 
0.250664   

 
3.825924   

 
0.277053   

 5   
0.23538
7   

 
5.60929
0   

 
1.414108   

 
30.21875   

 
58.3383
8   

 
0.295249   

 
3.830620   

 
0.293600   

 6   
0.23540
8   

 
5.608271   

 
1.414022   

 
30.21479   

 
58.3287
0   

 
0.299418   

 
3.830326   

 
0.304476   

 7   
0.235414   

 
5.608157   

 
1.414510   

 
30.21367   

 
58.3260
5   

 0.301472    
3.830780   

 
0.305368   

 8   
0.235415   

 
5.608128   

 
1.414509   

 
30.21357   

 
58.3257
2   

 
0.301660   

 
3.830780   

 
0.305632   

 9   
0.235415   

 5.608121    
1.414552   

 
30.21354   

 
58.3256
4   

 0.301701    
3.830795   

 
0.305656   

 10  
  

 
0.235415   
  

 
5.608120   
  

 
1.414554   
  

 
30.21353   
  

 
58.3256
3   
  

 0.301705   
  

 
3.830795   
  

 
0.305658   
  

  
 Variance  
Decompositio
n of DLFB:   

  
  

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  
  

S.E.   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  1    
 
0.156790   

  
 
0.00078
0   

  
 
7.993319   

  
 
4.30490
4   

  
 6.65E-
05   

  
 
87.70093   

  
 
0.00000
0   

  
 
0.00000
0   

 2   
0.178720   

 
0.612045   

 
9.62814
8   

 
4.875545   

 
2.49460
7   

 67.50957    
0.08868
2   

 14.79141   

 3   
0.184871   

 
0.773861   

 
10.6418
0   

 
4.59492
6   

 
2.99994
6   

 
65.36298   

 1.283818    14.34266   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

37 | P a g e  

 4   
0.185743   

 
0.95058
4   

 
10.5569
4   

 
4.56668
4   

 
2.98130
2   

 
64.84869   

 1.658247    14.43755   

 5   
0.18608
0   

 
0.955757   

 
10.6652
2   

 
4.550185   

 
2.97332
0   

 64.63541    1.794920    14.42519   

 6   
0.186132   

 
0.961358   

 
10.6650
8   

 
4.547643   

 
2.97704
5   

 
64.60594   

 
1.825808   

 14.41713   

 7   
0.186140   

 
0.961336   

 
10.6676
4   

 
4.547317   

 
2.97678
9   

 
64.6000
8   

 1.829933    14.41690   

 8   
0.186142   

 
0.961392   

 
10.6676
9   

 
4.54726
0   

 
2.97703
0   

 64.59911    
1.830600   

 14.41691   

 9   
0.186142   

 0.961391    
10.6676
9   

 
4.547263   

 
2.97703
2   

 
64.59909   

 
1.830626   

 14.41691   

 10  
  

 
0.186142   
  

 0.961391   
  

 
10.6676
9   
  

 
4.547263   
  

 
2.97703
7   
  

 
64.59906   
  

 1.830627   
  

 14.41692   
  

  
 Variance 
Decompositio
n of DLDB:   

  
  

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period  S.E.   DLGC   DLGK   DLDT   DLIT   DLFB   DLDB  
 DLGR   
                  

                  
1 0.215948  4.175000  4.859923  0.489323  0.605885  22.17905  67.69082  0.000000  
2 0.234199  3.952763  7.392635  1.278903  0.638402  18.86131  65.22346  2.652535  

3 0.237645  4.077868  7.873997  1.298188  0.911934  18.73193  64.46117  2.644914  
4 0.237988  4.068213  7.914957  1.301817  0.927342  18.70947  64.41447  2.663731  
5 0.238062  4.065940  7.924473  1.302481  0.934178  18.70143  64.39396  2.677537  
6 0.238067  4.065806  7.924171  1.302719  0.934944  18.70362  64.39132  2.677422  
7 0.238068  4.065833  7.924185  1.302764  0.934989  18.70346  64.39078  2.677982  
8 0.238069  4.065830  7.924253  1.302763  0.935003  18.70350  64.39066  2.677997  
9 0.238069  4.065836  7.924250  1.302763  0.935004  18.70349  64.39065  2.678003  
10 0.238069  4.065835  7.924255  1.302763  0.935004  18.70349  64.39065  2.678004  
                  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

38 | P a g e  

    
  
 Variance  
Decomposition of 
DLGR:     

  
  

  
  

  
  

  
  

  
  

  
  

  
  

 Period   
  

S.E.   
  

DLGC   
  

DLGK   
  

DLDT   
  

DLIT   
  

DLFB   
  

DLDB   
  

DLGR   
  

  1     
 0.755153   

  
 
0.002803   

  
 
11.06231   

  
 1.502395   

  
 
2.097732   

  
 
16.00708   

  
 
9.021366   

  
 
60.30632   

 2    
0.822087   

 
0.060881   

 
12.26010   

 1.790907    
3.884525   

 21.34494    
9.482869   

 51.17578   

 3    
0.829529   

 0.293374    
12.52555   

 
2.002059   

 
3.815918   

 
20.96703   

 
9.393659   

 51.00241   

 4    
0.832194   

 
0.295626   

 12.77577    1.997428    
3.841262   

 20.91620    
9.436867   

 
50.73685   

 5    
0.832515   

 0.311644    
12.76592   

 
1.996002   

 
3.850621   

 20.91331    9.464671    
50.69783   

 6    
0.832614   

 0.311572    
12.77297   

 1.995730    
3.850479   

 
20.90860   

 
9.469079   

 
50.69158   

 7    
0.832635   

 0.311966    
12.77253   

 1.995715    
3.851079   

 20.90871    
9.470205   

 
50.68979   

 8    
0.832637   

 0.311967    
12.77253   

 1.995745    
3.851055   

 
20.90879   

 
9.470231   

 
50.68968   

 9    
0.832638   

 0.311969    
12.77252   

 1.995752    
3.851068   

 
20.90879   

 
9.470237   

 
50.68966   

 10   
  

 
0.832638   
  

 0.311969   
  

 
12.77252   
  

 1.995754   
  

 
3.851067   
  

 
20.90880   
  

 
9.470235   
  

 
50.68965   
  

  
Cholesky  
Ordering  
DLGC   
DLGK 
DLDT  
DLIT 
DLFB  
DLDB 
DLGR   
  

  
  
  

  
  
  

  
  
  

  
  
  

  
  
  

  
  
  

  
  
  

  
  
  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 11 Issue 2, April-June 2023 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 
 

American Research Journal of Economics, Finance and Management 
https://americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

39 | P a g e  

                   
 
IMPULSE RESPONSE FUNCTIONS FOR AGGREGATED GOVERNMENT EXPENDITURE 
MODEL  
  
Response to Cholesky One S.D. Innovations ± 2 S.E. 

 
MODEL FUNCTIONS  
Response to Cholesky One S.D. Innovations ± 2 S.E. 

 

mailto:contact@americaserial.com
mailto:contact@americaserial.com

