AN ANALYSIS OF CROWDING-OUT OF PRIVATE SECTOR BY GOVERNMENT

BORROWING IN KENYA

KINYUA LUKE WAHOME

A RESEARCH PROJECT SUBMITTED TO THE DEPARTMENT OF ECONOMIC

THEORY IN THE SCHOOL OF ECONOMICS IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE AWARD OF THE DEGREE OF MASTERS OF
ECONOMICS OF KENYATTA UNIVERSITY.

NOVEMBER, 2022
 DECLARATION
This project is my original work and has not been presented for a degree in any other
university or any other award.

Signature………………… Date………………
Kinyua Luke Wahome…………
B. Economic and Sociology (Hons)
K102/30687/2015

I confirm that the work reported in this project was carried out by the candidate under my
supervision.

Signature…………………………..Date………………
Dr. Charles Nzai (PHD)
Department of Applied Economics
School of Economics
Kenyatta University

ii
 DEDICATION
To my dear parents and my beloved family

iii
 ACKNOWLEDGEMENT
I would like to thank Almighty God for providing me life and well-being during my studies. I
also appreciate my supervisor Dr. Charles Nzai for guidance and dedication to my
supervision. Special gratitude also goes to my workmates, fellow students and friends for
their moral support.

iv
 TABLE OF CONTENTS

DECLARATION ...................................................................................................................... ii
DEDICATION ........................................................................................................................ iii
ACKNOWLEDGEMENT ...................................................................................................... iv
TABLE OF CONTENTS ..........................................................................................................v
LIST OF TABLES ................................................................................................................ viii
LIST OF FIGURES ................................................................................................................ ix
ABBREVIATIONS AND ACRONYMS ..................................................................................x
OPERATIONAL DEFINITION OF TERMS ........................................................................ xi
ABSTRACT............................................................................................................................ xii
CHAPTER ONE .......................................................................................................................1
INTRODUCTION .....................................................................................................................1
1.1 Background of the Study ..................................................................................................1
1.1.1 The Crowding-Out Effect ............................................................................................1
1.1.2 Government Borrowing and Domestic Debt in Kenya .................................................2
1.1.3 Private Sector Financing in Kenya ...............................................................................7
1.2 Statement of the Problem ......................................................................................................8
1.3 Research Questions ...............................................................................................................9
1.4 Objectives of the Study .........................................................................................................9
1.5 Significance of the Study ..................................................................................................... 10
1.6 Scope of the Study............................................................................................................... 10
1.7 Organization of the Project .................................................................................................. 10
CHAPTER TWO .................................................................................................................... 12
LITERATURE REVIEW ....................................................................................................... 12
2.1. Introduction. ....................................................................................................................... 12
2.2 Theoretical Literature .......................................................................................................... 12
2.2.1 Crowding-out Theory ................................................................................................ 12
2.2.2 The Debt Overhang Theory ....................................................................................... 14
2.2.3 Credit Creation Theory .............................................................................................. 15
2.3 Empirical Literature ............................................................................................................ 16
2.4 Summary of the Literature ................................................................................................... 21
CHAPTER THREE ................................................................................................................ 22

v
METHODOLOGY.................................................................................................................. 22
3.1. Introduction ........................................................................................................................ 22
3.2 Research Design .................................................................................................................. 22
3.3 Theoretical Model ............................................................................................................... 22
3.3.1 Determinants of Government Borrowing in Kenya .................................................... 22
3.3.2 Effects of Government Borrowing on Private Sector Capital in Kenya ...................... 23
3.4 Model Specification ............................................................................................................ 24
3.5 Definition and Measurement of Variables ............................................................................ 24
3.6 Data Collection Procedure ................................................................................................... 25
3.7 Time series Analysis ........................................................................................................... 25
3.7.1 Testing for stationarity .............................................................................................. 25
3.7.2 Testing for Co-integration ......................................................................................... 26
3.8 Diagnostic Tests .................................................................................................................. 26
3.8.1 Multicollinearity Test ................................................................................................ 26
3.8.2 Heteroscedasticity Test .............................................................................................. 26
3.8.3 Normality Test .......................................................................................................... 27
3.8.4 Serial Correlation Test ............................................................................................... 27
3.9 Data Analysis and Presentation............................................................................................ 27
CHAPTER FOUR ................................................................................................................... 28
EMPIRICAL FINDINGS ....................................................................................................... 28
4.1 Introduction ......................................................................................................................... 28
4.2 Descriptive Statistics ........................................................................................................... 28
4.3 Time Series Test Results ..................................................................................................... 29
4.3.1 Unit Root Test ........................................................................................................... 29
4.3.2 Co-integration Test Analysis ..................................................................................... 30
4.3.3 Correlation Test Analysis .......................................................................................... 32
4.4 Diagnostic Tests .................................................................................................................. 33
4.4.1 Heteroscedasticity Test Analysis ............................................................................... 33
4.4.2 Normality Test .......................................................................................................... 34
4.4.3 Serial Correlation (LM) Test ..................................................................................... 35
4.4.4 Stability Test ............................................................................................................. 36
4.5 Regression Results Findings ................................................................................................ 37
4.5.1 The determinants of government borrowing in Kenya. ............................................... 37
4.5.2 Effect of government borrowing on private sector capital in Kenya ........................... 41
vi
CHAPTER FIVE..................................................................................................................... 48
SUMMARY, CONCLUSION AND RECOMMENDATIONS ............................................. 48
5.1 Introduction ......................................................................................................................... 48
5.2 Study Summary ................................................................................................................... 48
5.3 Conclusion .......................................................................................................................... 50
5.4 Policy Implications .............................................................................................................. 53
5.5 Contribution to Knowledge ................................................................................................. 55
5.6 Areas for Further Research .................................................................................................. 56
REFERENCES ........................................................................................................................ 57

vii
 LIST OF TABLES
Table 1.1: Percentage distribution of Domestic Credit in Kenya ……………………………. 3
Table 4. 1: Descriptive Statistics Results………………..…………………...………………28
Table 4. 2:Unit Root Tests Results ............................................................................................ 30
Table 4. 3:Co-integration Test Results....................................................................................... 31
Table 4. 4:Correlation Test Results............................................................................................ 32
Table 4. 5:Heteroscedasticity Test Results-Equation 3.3 ............................................................ 33
Table 4. 6:Heteroscedasticity Test results-Equation 3.4 ............................................................. 33
Table 4. 7:Serial Correlation (LM) test results-Equation 3.3 ...................................................... 35
Table 4. 8:Serial Correlation (LM) test results-Equation 3.4 ...................................................... 35
Table 4. 9:Ramsey RESET Test Results .................................................................................... 36
Table 4. 10:Ramsey RESET Test Results .................................................................................. 36
Table 4. 11:Determinant of government borrowing in Kenya Regression Results ...................... 38
Table 4. 12:Effect of government borrowing on private sector capital results ............................ 42

viii
 LIST OF FIGURES
Figure 1.1: Debt to GDP Ratio in Kenya ………………………………….………………...5
Figure 1.2: Budget Deficit in Kenya ………………………………………………………… 6
Figure 4.1: Histogram-Normality Test Results-Equation 3.3 ……………………………….34
Figure 4.2: Histogram-Normality Test Results-Equation 3.3 ……………………………….34

ix
 ABBREVIATIONS AND ACRONYMS
ADF Augmented Dickey Fuller
ARCH Autoregressive Conditional Heteroscedasticity
ARDL Autoregressive Distributed Lag Model
CBK Central Bank of Kenya
ESF Exogenous Shock Facility
IMF International Monetary Fund
GDP Gross Domestic Product
OLS Ordinary Least Squares
VAR Vector Auto Regression
VECM Vector Error Correction Model

x
 OPERATIONAL DEFINITION OF TERMS
Crowding It is a situation in which government actively borrows internally leading to
Out increase in real interest rate resulting a decline in savings available to
private investors
Credit Contractual accord where a borrower gets something of value now and
agrees to pay the lender at a later time, normally with interest.

xi
 ABSTRACT
The government of Kenya has been running budget deficits every year of its national
accounting. The government has therefore been forced to borrow either from the domestic or
external financial markets to bridge those deficits to offer its citizens services. Government
borrowing creates other macroeconomic problems in the economy. The objectives of the
research were to evaluate the determinants of government borrowing and the effects of that
borrowing on private sector credit, the crowding out impact of state loans. Two models for
each objective with clear variables based on economic theory were constructed for estimation
where variables included government borrowing or domestic debt, budget deficits, lending
rates, efficiency of tax agency, private sector credit and political factors. Data was collected
from reports published by government agencies for period 1990-2021 and data analysis
techniques included the Auto Regressive Distributed Lag model. Co-integration and unit root
tests were done prior to analysis. Domestic government borrowing is determined by the level
of budget deficit, domestic savings, inflation rate in economy and lending rates. Budget
deficit and inflation rate were both found to positively and significantly determine
government borrowing in the long-run, while else domestic savings and lending rate were
found to be negative and significant in determining domestic government borrowing. The
research findings on the impact of loans acquired by government on free enterprise economy
capital in Kenya showed that domestic government borrowing negatively and significantly
affect the level of private sector capital. The crowding out effect is huge as a percentage rise
in loans acquired by government causes six percent of private sector investors crowded out of
investment. Interest and lending rates were found to negatively and significantly affect
private sector capital level. Based on the study findings, budget deficit should be minimized
as much as possible by reducing unnecessary government expenditure. Similarly, domestic
savings should be encouraged as this stimulates domestic investment due to availability of
stock of capital for investment. Inflation rate should be maintained as low as possible below 5
percent in order to stimulate government borrowing. Lending rates should be kept low by
central banks to increase lending by commercial banks as this ensures enough money is in
circulation to facilitate borrowing and investment by the government in key sectors.
Domestic government borrowing should be discouraged as much as possible, other sources of
funds should be sort to finance expenditure such as health, infrastructural development,
education and manufacturing. The study has demonstrated that state loans from local sources
causes crowding out of the private sector investment and has also recommended ways
through which state loans from local sources can be reduced to ensure private sector
investment and spur economic growth of the economy

xii
 CHAPTER ONE

INTRODUCTION

1.1 Background of the Study

1.1.1 The Crowding-Out Effect

Crowding-Out Effect describes the cutback in investments in the private sector brought about

by an increment in expenditure in the public sector. Whenever the states expenditure goes

above the usual amount, it can prompt a rise in state loans (Ahmed & Miller, 2000). This

consequently raises the demand for credit reserve. Such an addition brings about an increase

in lending rate and a reduction in the quantity of money at hand to meet the requirements for

investment of the private sector. Consequently, such actions by the government crowds out

the private investment in the country, that is, individuals and businesses of all sizes are forced

out, or “crowded-out” of the market (Cruz & Teixeira, 1999).

The link between private investment and state borrowing is a perpetual matter in economic

development and growth based on many empirical and theoretical academic documents that

have been authored to form a concept if state borrowing brings about crowding in or

crowding out Macfarlane, et al., (2017). Proper use of loans could bring about enhanced

growth in the economy and thus, improved living standards, Kendren (2009). To enhance the

effectiveness of state loans, there is need for comprehensive reforms in the administration of

public finances. Nonetheless, in many instances, loans have not been adequately put into

service, for example, projects funded by loans from international bodies have, due to

insufficient planning, failed to produce enough resources to pay back the borrowed loan.

Whittaker (2007) states that funds should be borrowed with the sole purpose of “assisting

reduce poverty by having confidence in the norms of work and entrepreneurial potential of

the poorest people in the world”. Nonetheless, sometimes the locally acquired loans are used
1
for recurrent expenditure and not investment and manufacturing. As a result of these

opposing views, crowding out effect has been discussed for over a hundred years in various

forms and will continue to be debated.

1.1.2 Government Borrowing and Domestic Debt in Kenya

Governments normally get loans from local financial institutions or foreign lenders in

situations when public expenditure is way above current revenue (Fayed, 2012). Domestic

Public Debt is the accounts collectible owed to bearers of Government securities such as

Treasury Bonds and Bills. Domestic debt is issued on behalf of the Government by floating

Treasury bonds and bills through the central bank of Kenya. Domestic debt is also made up

of securities against small savings and market stabilization schemes. In Kenya, loans acquired

locally is used for different purposes including funding the recurrent budget deficits when the

government is unable to meet its budget obligations with the locally collected revenue and

grants sourced externally and loans. Domestic debt is also used for execution of monetary

policy via open market activities and also the state can use debt tools for financial markets

enhancement (Hasnat & Ashraf, 2018). The reason for getting loans is also to impact demand

for keeping up a stable economy (Ouattara, 2004). Table 1.1 presents a breakdown of

Domestic Debt in Kenya.

2
Table 1.1: Percentage distribution of Domestic Credit in Kenya
Years 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Total
Domestic
Credit 1,271. 1,532. 1,727. 1,957. 2,329. 2,793. 2,973. 3,279. 3,450. 3,660. 4,340.
(Millions) 6 1 7 5 0 9 2 3 2 5 9
National
Governme
nt (%) 27.3 21.4 24.5 21.0 8.8 18.8 19.9 22.8 24.9 24.6 31.3
Other
Public
Bodies
(%) 1.7 2.0 2.9 2.0 2.1 3.0 3.5 3.4 2.9 2.5 2.1
Private
Sector (%) 70.7 76.6 75.0 79.5 80.9 78.3 76.5 73.7 72.2 72.9 66.6
Source: KNBS, 2021
Table 1.1 shows the distribution of credit among the national government, government

agencies and the private sector. Total domestic credit has been increasing over the years

however; the proportion available to private sector is declining over the same period. Access

to loans by the National Government and other public bodies has been increasing from about

23 per cent in 2013 to 33.4 per cent in 2020. However, loans to private sector declined from

79.5% in 2013 to 66.6% in 2020. Slowing down loans to the private sector has been

associated with many elements along with the enactment of the interest rate cap that became

effective in September 2016; rise in non-performing loans, effect of winding up of three

commercial banks; impact of liquidity shock in 2015/2016 and inability of minor banks to

acquire credit in the interbank market. In spite of the borrowing rate ceiling, credits advanced

to the Government have gone up tremendously leading to crowding-out of vital sectors of the

economy as well as manufacturing and production. Moreover, Table 1.1 above indicates that

domestic debt has been rising throughout the period 2010 to 2020. The national government’s

borrowing also indicates that it was crowding out private sector borrowing between the

periods because as national government borrowing increased it can be observed that the

private sector received less credit from the financial institutions.

3
Kenya’s economy for many decades has been marked by inadequate revenue driving the

government to rely on getting loans from internal and external sources to fund deficiency in

the budget. Moreover, more public sector entities such as the County Governments and State

Corporations also get loans from separate sources (Republic of Kenya, 2013). Because of

recurrent expenditure, the Kenya government has been weighed down by public debt adding

up to Ksh 5 trillion in 2018 which is equivalent to 58% of GDP (Republic of Kenya, 2018)

Those in authority in government have always been in the spotlight for engaging in

imprudent taking of loans from internal sources and thus suppressing growth by denying

private sector access to credit within the financial market which is still relatively not adequate

enough to meet the demands of both government and private sector (Lau, Tan & Liew, 2019).

The assertion warrants significant observation in the context of the nation’s urgent need to an

improved economic expansion. The growth standard in the country in the past has been quite

reasonable at about 5% annually. However, halving poverty by 2030 which is a crucial

millennium development objective requires that the growth rate improve significantly. The

country’s blue print, the Vision 2030 expected GDP growth rate of 10% plus annually from

the year 2012. Furthermore, the unemployment rate at about 40% is yet to be reduced. It is

clear that in the current state, the economy should strive at achieving growth impetus. Figure

1.1 shows the trends of Debt to GDP Ratio for the years 1990 to 2021. Figure 1.2 shows the

budget deficit for the financial years 2009/2010 and 2020/2021.

4
 140

120

100
Percentage

80

60

40

20

0
1992

1999

2001

2008

2010

2017
1990
1991

1993
1994
1995
1996
1997
1998

2000

2002
2003
2004
2005
2006
2007

2009

2011
2012
2013
2014
2015
2016

2018
2019
2020
2021
Years

Figure 1.1: Debt to GDP Ratio in Kenya

Source: IMF, 2022

Based on Figure 1.1, the debt to GDP ratio was at its lowest of about 22.24 per cent in 2010

and has been on a steady increase of up to 69.73% in 2021. The figure also shows that in

1990s debt to GDP ratio was relatively high of over 90 percent indicating that a larger

proportion of Kenya GDP was going for debt repayment at the expense of service provision

in key sectors of the economy. However, post 2000 the ratio is on the declining trend leading

to the growth of the private sector. From 2010, debt to GDP again starts to increase

constraining credit availability to the private sector leading to crowding out of the key sector

that provides employment opportunities, production of goods and services as well as revenue

generation to the government. This can further be elaborated by Figure 1.2 which shows that

the budget deficit relatively decreases between 2009/10 and 2010/11 financial years and

sharply increases to -9.11 per cent in 2015/16 financial years, then further to -9.22 per cent in

2016/17. The deficit then started to decline to -7.06 per cent in 2017/18 financial year

(Republic of Kenya, 2021).

5
 0
-1 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21
-2
Deficit as % of GDP

-3
-4
-5
-6
-7
-8
-9
-10
Financial Years

Figure 1.2: Budget Deficit in Kenya

Source: National Treasury, 2021

Figure 1.2 shows that budget deficit in the country was at its lowest in 2010/11 standing at -

4.5 per cent of GDP but started rising hence forth to a record of above -9 per cent of GDP in

2016/17. Generally, other than the year 2010/11 where budget deficit showed an

improvement where it dropped all the other years 2009 to 2017 the country’s budget deficit

rose. In relation to Table 1.1, it may be presumed that rising budget deficit led to increased

loans acquired by the government thus crowding out the private sector in the country. The

budget deficit forces the government to borrow more from both the external and internal

sources to bridge the gap. Government can resort to borrow internally and with the limited

credit available, the amount left for the private sector diminishes and if the process continues

eventually crowds out the private sector investment (Švaljek, 2009). This indicates the reason

as to why government should not resort to internal sources of credit but seek from the

external sources such as International Monetary Fund (IMF), World Bank and other corporate

lenders in the world market.

6
1.1.3 Private Sector Financing in Kenya

Kenya has an active private sector, primarily consisting of small informal businesses with

less than five employees. There are 7.4 million micro, small scale and medium businesses in

the nation, most of which need additional finance to recover from the detrimental effects of

the COVID-19 pandemic. A private sector that is prosperous is key to attaining the goals of

Vision 2030 in the country. Moreover, the accomplishment of Vision 2030 is hugely based on

the part played by the private sector in attaining the nation’s output targets and, in doing so,

generating substantial job opportunities and wealth. Kenya greatly depends on funding

support from donors and has traditionally paid attention to capacity building than projects

that generate revenue (Were, 2011). In the infrastructural sector for instance, Kenya is faced

with a huge infrastructure funding deficit approximated at $2.1 billion per year, which

inhibits development and output. Continuous spending of about $4 billion annually is

essential in meeting the nation’s infrastructure requirements. With public debt currently at

about 57 percent of GDP, this shortfall cannot be covered by resources from public. The

nation ought to implement the public-private partnership framework to address the financial

deficit in the infrastructural sector. The World Bank Group approximates that enhancing

infrastructure funding could increase Kenya’s per capita output rate by three percent

(Murungi & Okiro, 2018).

Private sector credit is a key path for private investment in emergent countries like Kenya.

Loans acquired locally by private sector as capital are advanced by financial institutions via

purchase of non-equity securities, loans, trade credits and other invoices that start off an

appeal for payment (McLeay, Radia & Thomas, 2014). The financial institutions comprise

deposit money banks, monetary authorities, and other financial institution’s where

information is obtainable. Private sector development and investment are key in lowering
7
poverty level. In comparison with public sector endeavors, private investment mostly in

competitive markets has massive prospects to support growth. Private markets are the main

mechanism of producing fruitful jobs and higher earnings (Ramírez, 1994). An empirical

examination of the assertion of the importance of private sector credit to stimulate economic

expansion and the adverse effects of government borrowing to frustrate growth of private

credit was therefore necessary (King’wara, 2015). This study hypothesizes that if there is

enough cash in the financial system, loans by the government may not necessarily influence

private investment and government borrowing in Kenya is majorly a political matter than

economic one.

1.2 Statement of the Problem

Practically, for many countries around the world particularly the developing countries, the

easiest option for governments is to acquire loans from either local or foreign markets to

solve the problem of budget deficits. Government borrowing due to budget deficits then

creates domino effects in the economy such as a rise in borrowing rates and eventually

leading to crowding out of the private sector. Kenya has not been an exception to this

scenario. The government of Kenya has suffered budget deficits for many years and has

resorted to both domestic and external borrowing to bridge budget deficits. As a result, the

country’s private sector has been crowded-out from the domestic capital market resulting in

decline in private sector investments. The country’s economic blue print, the Vision 2030 is

anchored in private investments and the country will not achieve the targets set in the vision

as long as the private sector is starved of capital due to government borrowing from internal

sources (King’wara, 2015).

8
Studies on crowding-out phenomenon in Kenya are still necessary in relation to similar

studies done in the country because, one, the studies by Lidiema (2018); and, Mutuku and

Kinyanjui (2018) did not cover adequately the period when the country adopted a devolved

system of 47 county governments which naturally has direct implications on government

borrowing and consequently crowding-out phenomenon. These studies revealed mixed results

of the crowding-out phenomenon in the country which warrants further research. Mutuku and

Kinyanjui (2018) observed that crowding-out phenomenon in the country is a short-run

outcome while Lidiema (2018) observed that crowding-out of the private sector in the

country is a long term phenomenon. Even though the above mentioned studies focused on the

effect of government borrowing both in short-run and long-run, none of them considered the

analysis of crowding out of private investment by government borrowing in Kenya that the

current study is anchored on.

1.3 Research Questions

i. What are the determinants of government borrowing in Kenya?

ii. What is the magnitude of the crowding-out phenomenon in Kenya?

1.4 Objectives of the Study

The general objective of the research is to conduct an analysis of the crowding-out

phenomenon in Kenya. The specific objectives were:

i. To investigate the determinants of government borrowing in Kenya.

ii. To estimate the magnitude of the crowding-out phenomenon in Kenya.

9
1.5 Significance of the Study

Credit creation by commercial banks in Kenya is inadequate and government borrowing is

equally becoming a chronic problem for the country. The study’s key stakeholders who will

benefit from its findings are therefore government agencies like the National Treasury and

Central Bank and the country’s commercial banks. The civil society who represents the

interests of the ‘common man’ will also get the opportunity to acquire knowledge on

crowding out and government borrowing that facilitates their initiatives in championing the

interests of their constituents. Lastly, by addressing the gaps that were noted in both the

empirical and theoretical literature that was reviewed, the research findings are an addition to

available knowledge.

1.6 Scope of the Study

This study used time series data for the time between 1990 and 2020. The study’s variables

such as budget deficits, loans by commercial banks, interest rates and interest rate regimes

have been in play in the country’s economy during this period of study. This period of study

is equally current and would therefore provide the latest status on the behavior of these

variables.

1.7 Organization of the Project

The project is arranged in five different chapters. The first one covers Study Background,

Problem Statement, Research Questions, Study Objectives, Significance of Study, Study

Scope and Project Organization. The second chapter comprises the Review of Literature and

covers Introduction, Theoretical Literature, Empirical Literature and Summary of Literature.

The third one comprises Methodology of Study and covers Introduction, Research Design,

Definition and Measurement of Variables, Data Collection Procedure and Analytical

10
Techniques, Theoretical Framework and Model Specification. The fourth one comprises

Empirical Findings of the Study. The fifth chapter contains the Summary, Conclusion and

Policy Implications of the research. Lastly the project provides References.

11
 CHAPTER TWO

LITERATURE REVIEW

2.1. Introduction.

This chapter comprises the theoretical and empirical literature reviewed on government

borrowing, crowding out effects and credit creation by commercial banks. The first section

reviewed the theoretical foundations that this study was based on with clear discussions on

the interrelations between the variables at play in them. The second section reviewed

empirical studies carried out on these phenomena while the last section attempted to bring out

the gaps that this reviewed literature had and which this study intended to fill.

2.2 Theoretical Literature

There are many theories on which government borrowing, crowding out effects and credit

creation are anchored on. Some of the theories are discussed below:

2.2.1 Crowding-out Theory

Crowding out is the tendency of reducing private or individual consumption as an outcome of

enhanced government spending or a circumstance in which operations that are private

reduces due to increased operations of the government in the financial sector (Buiter, 1990).

This mainly is when the national government increases spending financed by taxation

implying more taxes has to be collected hence reducing disposable income of individuals and

therefore affecting consumption. On the other hand if the spending has to be increased

without increasing tax revenue, then governments always resort to internal borrowing

reducing funds available for private investment hence crowd out the sector from the financial

space. Murungi & Okiro (2018) the entry of government to secure loans reduces amount of

loan available to high risk individuals or private sector as a whole leading to a decline on the

private investment hence affecting service delivery and limiting employment opportunities.

12
Most countries particularly in developing nations operates a deficit budget where the revenue

generated is far much less than the expenditure in any financial year. In such a case, the

government resorts to external loans to fund budget shortcomings. Although in rare

circumstances this does not influence the internal rate of interest and amount of money

available to the private sector for loans, the likelihood of crowding out the private sector is

quite high (Mlachila, Loko & Beaugrand, 2012). Murungi & Okiro (2018) argued that if the

deficit comes from the expenditure of domestically produced goods, then, external borrowing

resort by the government causes a rise in exchange rate which in turn causes crowding out of

the private exporters and producers. On the other hand, when the government borrows

whether internally or externally to produce services and goods that completes or counteracts

those that the private sector provides would possibly lead to crowding out effect, signaling

bankruptcy in the government that later affect investment opportunities within the private

sector.

The theory is relevant to the research as it shows how public borrowing simultaneously

brings about crowding out of the private investment in economy in which there is increased

government borrowing both internally and externally. The weakness of this theory is that it

does not show the model on how this transmission effect occurs. Nonetheless, those who

make policy policies should take into consideration how public borrowing brings about

crowding-out effect on private sector investment, via whatever medium applied by the

national government, and to what extent to which the action outweighs the benefits accrued

from the use of borrowed money or value for money is realized by the public as is argued by

the classical economists (Chhibber, 1988).

13
Yap (1990) and Borensztein (1990) established that external debt negative effects weakened

private sector investment. Edo (2002) also affirmed that public debts negative affect

investment in developing countries which Kenya is a member, this was due to increase in

public expenditure, balance of payment problems and interest rates leading to increase debt

volume. Measures suggested by the theory to curb debt accumulation were; export

promotion, privatization and expanding capital markets through structural adjustment

programs. Even though the theory articulates determinants of debt, measure to address the

problem, the theory fails to scientifically explain how external borrowing brings about

crowding out of the private sector which is its shortcoming.

2.2.2 The Debt Overhang Theory

It is mostly used in the field of economic growth against borrowing by the public which is

always perceived to have a negative relationship between debt volume and investment level

leading to low capital formation, this negative connection is known as debt overhang

(Krugman,1988). The theory opioned that repayment of the outstanding debt always falls

below the real value, implying that the current value of the allocated resources do not match

the value of the outstanding loan.

Sachs (1989) together with Bulow & Rogoff (1990) argued that economies that are heavily

debt distressed normally results to distortions and eventually slows down the economic

growth of both private and public sectors. Further, servicing of these debts exhausts a larger

percentage of a country’s revenue such that possibilities of a country getting back to normal

growth path is almost close to impossibility. Levy-Livermore & Chowdhury (1998)

suggested that the effect will still be felt on investment even if structural programs are

implemented to check on government borrowing, the theory states that debt overhang does

14
not occur due to accumulation of capital stock but also through changes in debt repayment

environment which delay repayment hence increasing the debt stock in the economy.

Public debt in Kenya has become a burden for the future generations as debts incurred today

will be paid in future with an interest (Were, 2011). It is estimated that the public debt is

likely to reduce consumption level in Kenya over lifetime as the money invested in

development projects available over lifetime consumption may not decline. The theory is

applicable as there is likelihood that public debt volume will outweigh the ability of the

country to repay thereby becoming unsustainable resulting to decline in private investment as

a lot of funds will be used to service debt.

2.2.3 Credit Creation Theory

The theory was first applied in the banking sector during the first in eighteenth century by

John Law Economist Frenchman. John Law was a mercantilist when blood circulation was

discovered and together with other scholars MacLeod, Keynes, Albert Hahn and Hayek

agreed that credit creation was the blood of the society. The theory is mostly discussed in

textbooks on the money multiplier concept of credit creation, little has been talked about in

academic books on the same concept within the banking. The proponents of the theory

believed that banks do not require deposits in order to issue loans as the core mandate of

banks is to create credit as it is useful for economic activities such as production, investment

as well as capital accumulation (MacLeod, 1906).

15
2.3 Empirical Literature

Various studies have been done to ascertain the best theory to explain the functioning of

banks Turner (2012), Werner (2014) & Werner (2016), on financial intermediation theory,

credit creation and fractional reserve theory, the empirical findings were that it is only the

credit creation theory that explains social optimum allocation of resources among different

economic sectors in the economy. Theory holds that money is the most advance and abstract

nature of credit as compared to credit and wealth. It comes from endogenous exchange

process that comprise of credit transfer, obligation and termination where it creates both

claim and repayment obligation. Hahn (1920) maintained MacLeod opinion but diverted a

little bit by believing that it is not only savings that creates credit as commercial banks can

create credit without necessarily savings from the members but from capital formation.

Through technological progress and labour in the economy, credit boost capital formation.

A study done by Lidiema (2018), on the effect of loans taken by the government on

investment by private sector in Kenya found that internal borrowing has significant negative

relationship with overall gross fixed capital accumulation which diminishes in future or in the

long term. Further, the study opined that continuous loans that the government takes

domestically have a deleterious effect on investment which in the end affects the economy.

The research proposed that proper blue print should be put down to guard domestic

borrowing and internal interest rates but stimulate the growth of small and micro enterprises

that encourage the economic development of the nation. The research adopted time-series

data for a period of 39 years collected from World Bank on study variables such as gross

fixed capital formation as dependent variable with gross domestic per capita, domestic debt,

lending rates, domestic credit to private sector, external debt as independent variables. Auto

Regressive Distributed Lag (ARDL) was used in testing the co-integration while cumulative
16
sum squares and cumulative sum tests were conducted to find out short and long run

stableness of the variables coefficients.

According to Ado & Ibrahim (2019), while investigating the connection between government

borrowing and the growth of public sector in Nigeria, found that there is a higher chance or

capacity by the government to take loans than the private sector hence in the end crowd–out

the private sector which negatively affect the advancement of the sector. The research

suggested that both monetary and fiscal policies that enhance the growth of private sector

should be adopted to improve the growth of the private sector. The Vector Auto-regression

model was adopted in the examination of time-series data gathered on the study using

variables like credit growth to government sector, credit to private sector, interest rates and

domestic debt. Multivariate co-integration was adopted in testing co-integration between the

variables.

A study done by Lau, Tan & Liew (2019) to find out the impact of public loans on private

section investment in Malaysia found that public loans crowds out private section investment

in both long and short run hence lowering the output of the private sector which is key in the

provision of the employment opportunities in the nation. The research recommended that

government ought to maintain a healthy debt levels that encourage the growth of the public

sector rather than crowding out of the sector.

According to Cooray (2019), while studying the influence of public loans on the private

section in Sri Lanka found that domestic borrowing by the government does not lead to

crowding out of the private but rather leads to inflation in and/or indebtedness of the

economy. This implies that government can actually finance deficits in the budget without
17
necessary affecting the investment in the private sector, this can be achieved through

employment of monetary policies to curb the negative effects on the private investment. The

study used time-series data for a period of 54 years from 1960 to 2014 on study variables

such as private investment as dependent variable while interest, gross domestic product and

public borrowing as independent variables. Vector Auto-regression model was adopted to

analyze the research after undertaking time-series pretest analysis on the study variables. The

study opined that internal borrowing by the government has no crowding out effect but only

affect inflation rate in the economy as well as debt distress an area that the current study

determined to obtain the right position of government borrowing on private investment.

Hasnat & Ashraf (2018) carried out a study to investigate long run crowding out in private

sector as an outcome of government internal loans interest rates sensitivity and fiscal deficit

induced interest rates significantly crowd out private investment in India. Auto-regressive

distributed lag approach was adopted to analyze the impact of borrowing rate sensitivity on

private investment while granger causality was used to fiscal deficit induced interest rates on

private investment as well as two-tier undertaking to investigate the existence of financial

crowding out impact on the private section investment. Time-series data was gathered on

research variables like market capitalization as dependent variable used as a proxy for private

investment while interest rates, government bond capitalization and commercial bank credit

as independent variables.

Ayturk (2017), examined the impact of government loans on corporate investment in 15

advanced countries in European countries between 1980 and 2014 found that government

borrowing significantly determine corporate investment in developed nations. Further, the

study opined that there is a negative link connecting private investment and government
18
borrowing levels in any given countries under study. Additionally, the study found that

increased government borrowing crow-out private sector investment, at the same time, large

investment firms which are more credit worth were found to be more sensitive to government

borrowing than those which are financially fixed. The study used panel data from the 15

developed countries in European nations on study variables such as government net debt as

dependent variable while inflation rate, price levels, gross domestic product, return on equity

and yield on bonds as independent variables.

Zaheer, Khaliq & Rafiq (2017), did a study to investigate the influence of government loans

got from domestic commercial banks on private sector investment. The outcome was that of a

significant negative influence of government loans affecting credit worthiness of the private

sector as one percent increase in government loans causes an 8 percent crowding out of the

private section in Pakistan because of reduction in loanable funds. However, banks’ lending

capacity was established as positively significant to private section investment while discount

rate had a negative coefficient which was statistically insignificant due to low variability in

concession rate during the duration of the study. On the other hand, industrial production

index had a minimal significant positive impact on the private section credit while consumer

price index was statistically insignificant. The study applied vector correction model (VECM)

with monthly data from 1998 June to 2015 December on study variables such as private

sector loans by commercial banks as the dependent variable while government borrowing,

discount rate, lending capacity growth, industrial production index growth, and consumer

price index as the independent variables.

A study done by Ehalaiye, Botica-Redmayne & Laswad (2017) on determinants of

government borrowing in Zealand found that income level is a major determinant of

19
government borrowing rather infrastructural spending. Similarly, spending on other assets

such as education, health leads to increased government borrowing. Further, interest rates in

the country and natural disaster and calamities leads to increased borrowing in order to

address these problems affecting economic growth as well as the well-being of the citizens.

Panel data from 17 cities in Zealand was used to analyze the study using variables such as

government debt as the dependent variable with capital expenditure on infrastructure, revenue

rates, other income other than revenue, other assets apart from infrastructural assets, funds for

own investment as the independent variables. The results showed that revenue rates, other

assets apart from infrastructure and other income apart from revenue were found to be

positive and significantly influence with government loans. In addition, finances for own

government investment negatively affected government borrowing in Zealand over the period

of study.

According to Tkačevs & Vilerts (2019) while studying on impact of government borrowing

on fiscal discipline and the subsequent influence on the magnitude of investment in the

private section established that borrowing locally by the government negatively and

significantly affect private sector investment implying that increase in government borrowing

lead to decline or deterioration of the fiscal policy in the country which eventually negatively

affect private sector investment. The exploration research methodology with panel data

obtained from member states on fiscal policy as dependent variable, debt to gross domestic

product, output level, government borrowing costs and political stability as independent

variables from 1985 to 2015. Based on the study findings, the research suggested that interest

rates should be moderate as low interest rates encourage government borrowing which

eventually leading to crowding out of the private sector.

20
2.4 Summary of the Literature

The discussion above proposes that there are no irrefutable empirical outcomes on whether

extra loans acquired by government can lead to crowding-out or not. Some arguments favor

crowding-in but others support the crowding-out impact. The arguments above have also

differed on the rationale of government borrowing as well and have not consented on what

exactly amounts to sustainable government borrowing. By and large, it is considered that the

impact of loans acquired by government on private investment varies with situations

depending on the current socio-economic circumstances of a country under analysis

especially the nature and level of development of its financial market. Nevertheless, it is

unanimous that private sector requires capital for it to grow and therefore this study chose the

school of thought that government borrowing crowds out private investments particularly

given the fact that Kenya’s economy is dominated by the private sector.

Unlike the empirical studies that have been reviewed above, this study was unique in that it

had adopted the model that incorporates political influence as one key variable in determining

government borrowing. In analyzing the effect of government borrowing on private credit,

the study while supported by theory had used unique variables in its model than those used by

the studies reviewed. For example, this study considered the variables used by the latest study

done on the subject matter by Kenya Bankers Association such as External Debt, Domestic

Debt and Gross Domestic Product per capita Rate as unnecessary and misplaced and included

budget deficit as a variable which in this study was more appropriate in explaining

government borrowing than external debt because external debt would be reflected as part of

the budget deficit if any.

21
 CHAPTER THREE

METHODOLOGY

3.1. Introduction

The chapter presents research methodology that was used in doing examinations on study

objectives. It is covers research design, empirical model, theoretical framework, data

collection and data analysis techniques and definition and measurement of variables.

3.2 Research Design

This study utilized a non-experimental design. In it, the researcher could not misrepresent,

alter or control the predictor theme or variable but rather relied on observations, interactions

or interpretation to arrive at an inference. Usually, this means non-experimental research

depend on correlations thus tend to have an external validity of high level thus results could

be assumed to represent a bigger population.

3.3 Theoretical Model

3.3.1 Determinants of Government Borrowing in Kenya

The theory underpinning objective one of the study was the model of the members of “new political

economy" depending on positive perspective. It disregards the supposition of the government as a

kind cordial planner, and presumes that those in authority select source of financing on account of

the interests and specific restrictions established by institutional authorities (Heinemann &Illing,

2002). The models therefore include the influence of politics in government borrowing which the

other theories ignore and it is for this reason that this study adopted this model. This model

asserts that fiscal debt and deficit were outcomes of political disagreements between different

groups of citizens (future and present politicians) because of disagreements over capital formation

(Alesina and Tabellini, 1990) or regarding common expenses (Persson and Svensson, 1989).

The models presume that institutions that handle budget decide on the size and deficit and that a

22
huge budget and fiscal shortfall are an outcome of financial indiscipline. Based on the foregoing

arguments, this model of government borrowing could be mathematically stated as follows:

𝐺𝑏 = 𝑓(𝑃𝑓 , 𝑀𝑦 ) ……………………………………………………………………………… 3.1

Where,
Gb = Government Borrowing
Pf = Political influences on debt
𝑀𝑦 = A host of macroeconomic variables such as interest rates, savings influencing debt.

3.3.2 Effects of Government Borrowing on Private Sector Capital in Kenya

The second objective of the study was theoretically anchored in the crowding out theory.

Crowding out in this study is the phenomenon where government borrowing from the

domestic financial market denies private sector credit for investments. Theoretically,

crowding-out generally happens where public authorities borrow from the domestic market

resulting in emergence of a capital crisis because of too much demand which pushes up the

borrowing rate leading to a reduction in private investment. The effects of government

borrowing in domestic market are debatable with regard to crowding out and crowding in.

For example, if borrowed funds are used up to develop services and goods which are

regarded as a replacement for services and goods produced in private, the trust in the private

investors is lowered leading to a decrease in private investment. Alternately, in a situation in

which public institution takes loans to deliver things that complement private sector products,

there is a chance that it could lead to crowding-in impact as opposed to crowding-out effect

even in a stringent money market conditions. Generally, policy makers should have

knowledge on public borrowing in relation to crowding-out effect on investment through

different channels and whether or not the damaging impact of such measures is greater than

their benefits. From the foregoing argument, crowding out effects can formally be stated as

follows:
23
𝑃𝑆𝐶 = 𝑓(𝐺𝑏 , 𝑀𝑦 ) ………………………………………………………………………… 3.2

Where,
PSC = Private Sector Credit (Capital Stock)
𝐺𝑏 = Government Borrowing

𝑀𝑦 = Other macroeconomic variables influencing credit creation by commercial banks

3.4 Model Specification

The empirical models to investigate objectives one and two were based on Equations 3.1 and

3.2 above, and were specified, respectively, as follows:

𝐺𝑏 = 𝛽0 + 𝛽1 𝐵𝐷 + 𝛽2 𝐸𝑇𝐴 + 𝛽3 𝑃𝐼 + 𝜀 …………………………………………….……. 3.3

𝑃𝑆𝐶 = 𝛽0 + 𝛽1 𝐺𝑏 + 𝛽2 𝐿𝑅 + 𝛽3 𝐷𝑆 + 𝛽4 𝐶𝐵………………………………………………. 3.4
Where,
BD = Budget Deficit, ETA = Efficiency of Tax Collection Agency, PI = Political Influence

Gb = Government Borrowing, LR = Annual Central Bank Lending Rate, DS = Domestic

Savings.

CBC = Commercial Banks Capacity while 𝛽0 - constant term and 𝛽1 , 𝛽2 , 𝛽3 𝑎𝑛𝑑 𝛽4 and the
coefficients.
3.5 Definition and Measurement of Variables
Key variables of the study are defined and how each variable is measured are also indicated

in table 3.1. The study also defined other variables affecting government borrowing and

credit availability to the private sector.

24
Table 3.1 provides the definition and measurement of variables.

Variable Definition/Measurement of Variable

Government Borrowing (Gb) Money Value of Domestic Borrowing in a Year
Budget Deficit (BD) The disparity between budgeted expenditure and revenue for
government in a year as per the budget statement.
Efficiency of Tax Collection This is a dummy to be assessed by whether Kenya Revenue Authority
Agency meets its targets or not (1 and o, respectively)
Political Influence (PI) A dummy whose value, 1 or 0, will be determined by perceived
expenditure that will be assessed to be more of a political populist
spending other being economically viable or convincing.
Private Sector Credit (PSC) Total loanable amount by commercial banks in the country in a year.
Lending Rate (LR) The annual Central Bank Lending Rate.
Gross Domestic Savings Money value of level of savings held by commercial banks.
(GDS)
Commercial Banks Capacity This will be measured by the liquidity of the commercial banks
(CBC) utilizing the sum value of money of all assets of commercial banks.
Source: Author Computations

3.6 Data Collection Procedure

This study adopted secondary data in annual economic Surveys, economic reports, annual

statistical abstracts and bank reports developed by the Kenya National Bureau of Statistics

and the Kenya Bankers Association including data from the National Treasury. The data

utilized in this research was for the duration between 1985 and 2020.

3.7 Time series Analysis

The research utilized time series data to find out the correlation amongst the variables of

study within the period of study.

3.7.1 Testing for stationarity

In empirical analysis, Non-stationarity of time series data is a bit complicated. Stationarity

test was conducted to avoid estimation and getting spurious results. This study used ADF test

and check for stationarity to find out the integration sequence. The ADF test for stationarity

25
in the chain involved estimating the equations, (Enders, 2004). Nonetheless, the study used

ARDL formed by Pesaran, Shin, & Smith, (2001) to intergrate I(0) and I(1) variables in the

identical estimation.so if at all the variables were stationary I(0) then OLS would be

acceptable and if at all the variables were non-stationary I(1) then the study would adopt

VECM (Johanson Approach) since it is easier model.

3.7.2 Testing for Co-integration

The study data was evaluated by Johansen Co integration test technique. It is a method used

to find out the availability of long-term links among variables in a non-stationary series. Prior

to assessing co integration, it was critical to determine the integration sequence of the

individual time series. If the series was I (1) the study would use DW-test, where DW value

would be calculated as 𝐷𝑊 = 2(1 − 𝜌′) where𝜌 = 𝜌′ is the estimation first order

autocorrelation.

3.8 Diagnostic Tests

The diagnostic tests carried out by the study were multicollinearity, heteroscedasticity test,

normality test, correlation test and serial correlation tests

3.8.1 Multicollinearity Test

The test was carried out using Variance Inflation Factor (VIF) to test multicollinearity among

the independent variables. The test was necessary to avoid getting spurious results.

3.8.2 Heteroscedasticity Test

The test was carried out using Autoregressive conditional heteroscedasticity (ARCH) Test to

ensure that the error term is normally distributed and independent from the independent

variables in the estimation model

26
3.8.3 Normality Test

The test was carried out using Histogram-Normality test to ensure that the residuals are

normally distributed in the data

3.8.4 Serial Correlation Test

The test was carried out using Breusch-Godfrey Serial correlation (LM) test to ensure that the

error term is serially independent from the independent variables in the estimation model.

3.9 Data Analysis and Presentation

The data was analyzed using Autoregressive Distributed Lags (ARDL) model since the

variables were found to be stationary at different levels using STATA version 16.0 and E-

views version 10. The results were presented in form of tables and figures.

27
 CHAPTER FOUR

EMPIRICAL FINDINGS

4.1 Introduction
The chapter covers the outcomes of the research as follows; descriptive statistics, time series
analysis such as unit root test, correlation and auto-correlation, diagnostic tests as well as
empirical findings.

4.2 Descriptive Statistics

Descriptive statistics like maximum, mean, minimum, standard deviation and median are
given in Table 4.1.
Table 4. 1: Descriptive Statistics Findings

GOVT BUDGET DOMESTIC INFLATIO INTEREST LENDING BANK

BORROWING DCPS DEFICIT SAVINGS N RATE RATE RATE LIQUIDITY

Mean 12.90581 27.12323 -5.035484 14.87097 11.59769 8.282806 19.19032 5.924173

Median 12.66214 25.81000 -5.500000 13.87000 9.234126 7.815000 16.57000 5.911880
Maximum 14.97190 40.20000 2.200000 37.16000 45.97888 21.09600 36.24000 8.557917
Minimum 10.90337 18.50000 -10.10000 7.320000 1.554328 -8.010000 12.10000 2.132627
Std. Dev. 1.137931 5.769438 3.094463 6.736859 9.465027 6.989170 6.830457 1.847258
Skewness 0.230848 0.626391 0.430392 1.850411 1.969099 -0.239788 1.090178 -0.705490
Kurtosis 2.114869 2.706898 2.348215 6.610637 7.008722 2.706996 3.058526 2.921872

Jarque-Bera 1.287300 2.138187 1.505788 34.52984 40.78988 0.407965 6.144941 2.579420

Probability 0.525371 0.343320 0.471002 0.000000 0.000000 0.815477 0.046307 0.275351

Sum 400.0802 840.8200 -156.1000 461.0000 359.5285 256.7670 594.9000 183.6494

Sum Sq.
Dev. 38.84658 998.5925 287.2710 1361.558 2687.602 1465.455 1399.654 102.3708
Observ 31 31 31 31 31 31 31 31
Source: Author Computations

The time series data utilized in the analysis of the research were collected for a period 1990-

2020. The results show that inflation rate in the economy had the maximum value of 45.98

while budget deficit had a minimum value of -10.10. Further, the results show that

28
government borrowing, credit given to private sector locally, domestic savings, inflation rate,

budget deficit and lending rate were skewed to the right (positively skewed). On the other

hand, interest rates and bank liquidity which measured commercial capacity in the economy

were skewed to the left implying that the data was normally distributed but was more skewed

to the left (negatively skewed).

Moreover, the analysis also showed that domestic credit advanced to private sector, interest

rates, lending rates and banks’ liquidity had values close to 3 hence were mesokurtic

implying the data was normally distributed. Variables such as domestic savings and inflation

rates had a kurtosis value more than 3 hence were leptokurtic. Lastly, government borrowing

and budget deficit had kurtosis value less than 3 hence were platykurtic. This imply that only

two variables have high kurtosis and two having low kurtosis, hence, the research concludes

that data gathered was distributed normally hence chances of obtaining spurious results were

minimized.

4.3 Time Series Test Results

The tests done were unit root test, co-integrating tests and correlation test.

4.3.1 Unit Root Test

The test was done by Augmented Dickey Fuller (ADF) at both trend and intercept. The test

was carried out to see to it that all the variables utilized in the analysis were stationary to

avoid chance of getting spurious results. Some study variables became stationary at level

while others only after first difference implying that the study could not use Ordinary Least

Square (OLS) model but Auto-regressive Distributed Lag (ARDL). Based on to the rule of

thumb when the research variables are stationary at both level and first difference then ARDL

is preferred over OLS. The results are presented in Table 4.2

29
Table 4. 2: Unit Root Tests Results
Variable Type of test t-statistics P-value Remarks
Intercept -3.0256 0.0438 Stationary
Budget Deficit (Level)
Trend and Intercept -3.6404 0.0430 Stationary
Intercept -8.0277 0.0000 Stationary
Domestic Savings (1st Difference
Trend and Intercept -6.5473 0.0001 Stationary
Intercept -6.2381 0.0000 Stationary
Inflation (1st Difference)
Trend and Intercept -6.1335 0.0000 Stationary
Intercept -3.9944 0.0045 Stationary
Interest Rate ( Level)
Trend and Intercept -4.0549 0.0174 Stationary
Intercept -5.4076 0.0001 Stationary
Lending rate (1st Difference
Trend and Intercept -5.4074 0.0007 Stationary
Intercept -6.0828 0.0000 Stationary
Bank Capacity (1st Difference)
Trend and Intercept -6.4127 0.0001 Stationary
Intercept -3.3244 0.0236 Stationary
Government Borrowing (1st Difference)
Trend and Intercept -3.6040 0.0470 Stationary
Intercept -5.3815 0.0001 Stationary
Domestic Credit (1st Difference)
Trend and Intercept -5.2949 0.0010 Stationary
Intercept -7.4788 0.0000 Stationary
Political Influence (Level)
Trend and Intercept -7.3557 0.0000 Stationary
Source: Author Computations

The findings reveal that at 5 percent significance level, the entire variables have a P-value

smaller than 0.05 indicating that the entire variables were stationary. According to the rule of

the thumb, a variable with a p-value of less than 0.05 is assumed to be stationary while a p-

value above 0.05 is non-stationary. After testing the stationarity of the variables, the study

found that budget deficit, interest rates were stationary at level while domestic savings,

inflation rate, lending rate, banks’ capacity, government borrowing and domestic credit to

private sector were found to be stationary only after first difference.

4.3.2 Co-integration Test Analysis

It was done to prove if there is an ultimate link connecting the variables utilized to analyze

study findings. The findings are presented in Table 4.3

30
Table 4. 3:Co-integration Test Results
Unrestricted Co integration Rank Test (Trace)

Hypothesized Trace 0.05

No. of CE(s) Eigenvalue Statistic Critical Value Prob.**

None * 0.983267 362.6983 175.1715 0.0000

At most 1 * 0.928274 244.0770 139.2753 0.0000
At most 2 * 0.817838 167.6648 107.3466 0.0000
At most 3 * 0.785919 118.2819 79.34145 0.0000
At most 4 * 0.668936 73.58118 55.24578 0.0006
At most 5 * 0.528678 41.52328 35.01090 0.0088
At most 6 * 0.444315 19.70909 18.39771 0.0326
At most 7 0.087959 2.670026 3.841466 0.1023

Trace test indicates 7 co-integrating equation(s) at the 0.05 level

* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Source: Author Computations
It was conducted using Johansen co-integration test using Akaike Information Creteria (AIC)

at 1 maximum lag length, the results show that seven co-integrating equations have a p-value

of below 0.05 at 5 percent significance level showing no eventual link between the variables

hence no co-integration and the variables are good to be used for the analysis of the study

findings. Similarly, the worth of trace statistics is greater than the crucial value at 5 percent

significance level for the co-integrating equation leading to the assumption that the variables

have no long-run relationship and therefore no spurious results could be obtained. According

to the rule of the thumb, in the presence of n-variables, then co-integrating equations should

be n-1. The study has 8 variables and 7 equations were found to co-integrate at 5 percent

statistical significance level proving that there was no co-integration between the variables.

31
4.3.3 Correlation Test Analysis

It was done to prove if there is an eventual linear link between independent variables. The

findings prove that all the variables have both positive and negative weak correlation except

government borrowing and bank capacity (bank liquidity) which has a strong positive

correlation implying that the two variables cannot be used in the same equation hence only

applicable in different equations. The findings are presented in Table 4.4

Table 4.4: Correlation Test Results

Budget Domestic Interest Lending Bank Government
Deficit Savings Inflation Rate Rate Liquidity Borrowing DCPS

Budget Deficit 1.0000

Domestic Savings 0.1021 1.0000
Inflation 0.0577 0.7326 1.0000
Interest Rate 0.4879 0.0252 -0.3274 1.0000
Lending Rate 0.1859 0.6877 0.3359 0.3329 1.0000
Bank Liquidity -0.4033 -0.7580 -0.6856 -0.0694 -0.5747 1.0000
Government
Borrowing -0.5093 -0.6001 -0.4828 -0.1317 -0.5805 0.9166 1.0000
DCPS -0.4279 -0.6318 -0.5579 0.0330 -0.4933 0.8481 0.8696 1.0000
Source: Author Computations

The models the study employed indicate that government borrowing is a dependent variable

in equation 3.3 used to achieve objective one of the study, therefore, bank liquidity cannot be

used in equation 3.3 but can be used in equation 3.4 in order to attain the second objective of

the research. The challenge of high correlation was addressed by using bank liquidity to

achieve objective two. Based to the law of the thumb, a correlation coefficient below or

equivalent to 0.8 is considered weak or moderately correlated hence no chances of obtaining

spurious results. The study data shows that correlation coefficients were less than or equal to

32
0.8 for most variables hence were used to analyze study findings as there was no indication of

a singular matrix amongst the variables.

4.4 Diagnostic Tests

It was conducted to prove the stability, reliability and validity of the model used in the

research. The tests done were heteroscedasticity test, histogram-normality test, serial

correlation Langrage Multiplier (LM) test, stability test and the multi-collinearity tests.

4.4.1 Heteroscedasticity Test Analysis

Heteroscedasticity test was carried out using Autoregressive Conditional Heteroscedasticity

(ARCH) model to check whether the error term has a constant variance gradually or not. The

findings of the analysis are presented in Table 4.5 and 4.6

Table 4.5: Heteroscedasticity Test Results-Equation 3.3

Heteroscedasticity Test: ARCH
F-statistics 1.31E-06 Prob. F (1,26) 0.9991
Observed R-Squared 1.41E-06 Prob. Chi-Squared 0.9991
(1)
Source: Author Computation

Table 4.6: Heteroscedasticity Test results-Equation 3.4

Heteroskedasticity Test: ARCH
F-statistics 1.31E-06 Prob. F (1,26) 0.9991
Observed R-Squared 1.41E-06 Prob. Chi-Squared 0.9991
(1)
Source: Author Computation

The results show that at 5 percent significance level, the P-value of observed R-Squared and

F-statistics are greater than 0.05 implying that the values are statistically significant hence the

variance value of the error term is constant in the long-run, therefore there was no chances of

obtaining spurious results from the model.

33
4.4.2 Normality Test

It was conducted using Histogram-Normality test to prove if the error term is normally

distributed or not. The test was conducted using Jarque-Bera. The findings are shown in

figure 4.1 and 4.2

8
Series: Residuals
7 Sample 1992 2020
Observations 29
6
Mean 7.64e-15
5 Median -0.140887
Maximum 2.761894
4 Minimum -1.829511
Std. Dev. 1.191476
3 Skewness 0.571860
Kurtosis 2.827427
2
Jarque-Bera 1.616602
1 Probability 0.445615

0
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0

Figure 4.1:Histogram-Normality Test Results-Equation 3.3

Source: Author Computations

12
Series: Residuals
Sample 1991 2020
10 Observations 30

8 Mean -7.73e-16
Median -0.685602
Maximum 4.912404
6 Minimum -3.467226
Std. Dev. 1.977528
Skewness 0.758644
4
Kurtosis 2.901072

2 Jarque-Bera 2.889934
Probability 0.235754
0
-4 -3 -2 -1 0 1 2 3 4 5

Figure 4.2:Histogram-Normality Test Results-Equation 3.4

Source: Author Computations

The results in figure 4.1 and 4.2 show that the p-value of the Jarque-Bera statistics is more

than 0.05 at 5 percent significance level hence are statistically significant. If the p-value is
34
greater than 0.05 then the salvage values are distributed normally and when p-value is lower

than 0.05 then the salvage values are not distributed normally. From the findings the p-values

are greater than 0.05 at 5 percent significance level, this implies there is non-normality hence

the data is normally distributed indicating that there are no outliers or extreme values in the

data that could lead to spurious results.

4.4.3 Serial Correlation (LM) Test

It was conducted to prove if there is continued link in the error term. Breusch-Godfrey Serial

correlation (LM) test was utilized to conduct it. The findings are presented in Table 4.7 and

Table 4.8.

Table 4. 7: Serial Correlation (LM) test results-Equation 3.3

F-statistics 0.515438 Prob. F (2, 14) 0.6081
Observed R-Squared 1.988931 Prob. Chi-Squared 0.3699
(2)
Source: Author Computation

Table 4.8: Serial Correlation (LM) test results-Equation 3.4

F-statistics 0.669350 Prob. F (2, 11) 0.5317
Observed R-Squared 3.146386 Prob. Chi-Squared 0.2074
(2)
Source: Author Computation

The results in the tables show that the F-statistics has a value of 0.515438 and 0.669350 with

a p-values of 0.6081 and 0.5317 respectively at 5 percent significance level. The supposition

is that the error term is serially independent from the independent variables. If the p-value is

above 0.05 at 5 percent significance level, then autocorrelation of the error term is absent and

when the p-value is below 0.05 then the error term is continuously autonomous. The findings

prove that p-values are above 0.05 at 5 percent significance level hence there was no
35
autocorrelation implying the model was good for the analysis of the crowding out of the

private sector by acquisition of loans by government of Kenya.

4.4.4 Stability Test

It was conducted using Ramsey RESET test only as the study adopted ARDL model which

does accommodate CUSUM test. The test was done to check whether the model was well

specified or not. The findings are presented in Table 4.9 and Table 4.10

Table 4. 9: Ramsey RESET Test Results

Value df Probability
t-statistics 0.620146 15 0.5445
F-statistics 0.384581 (1, 15) 0.5445
Source: Author Computation

Table 4. 10:Ramsey RESET Test Results

Value df Probability
t-statistics 0.690438 14 0.5012
F-statistics 0.476705 (1, 14) 0.5012
Source: Author Computation

The results show that the t-statistics value is 0.620146 and 0.690438 with p-values of 0.5445

and 0.5012 respectively. On the other hand, the F-statistics value 0.384581 and 0.476705

with p-values of 0.5445 and 0.5012 respectively. Based on the law of thumb when the p-

value is above 0.05 then the model is well specified and when it is below 0.05 at 5 percent

significance level then the model is not specified. The study results show that p-values were

above 0.05 at 5 percent significance level proving that the model was well specified and

therefore was good and suitable for the analysis of the research results as the changes in the

dependent variable were clarified by the independent variables plus the independent variables

have ability in demonstrating any alteration in the dependent variable.

36
4.5 Regression Results Findings

The research sought to examine crowding out of the private sector by government borrowing

in Kenya. The research had two objectives to attain, the first being to investigate the

determinants of government borrowing in Kenya and the second objective was to determine

the effect of government borrowing on private sector capital in Kenya. The two were

achieved as follows.

4.5.1 The determinants of government borrowing in Kenya.

This is the first objective of the research. To attain the objective, the research conducted a

regression analysis of domestic government borrowing against budget deficit, domestic

savings, inflation rate, interest rates and lending rates while political influence on debt

issuance and efficiency of tax collection agency as control variables using ARDL model. The

estimation method adopted fixed selection using Akaike Information Creteria (AIC) to select

the maximum lag length of two (2, 2) for both the independent and dependent variables. The

findings from the ARDL regression analysis shows that the value of the adjusted R-squared

was 0.814721, proving that 81.47 percent of the alterations in acquisition of loans by

government locally is decided by the level of budget deficit, amount of domestic savings in

the economy, inflation rate, interest rates, lending rates that the government obtains the debt

from domestic source as well as political influence, however, only 18.53 percent changes in

domestic government borrowing are determined by other factors that the study did not

consider and were taken care of by the error term. Further, the study found that the value of

F-statistics was 14.22 with a probability of 0.0000 which is below 0.05 at 5 percent statistical

level means that the model was good and fit to investigate determinants of domestic

government borrowing in Kenya. The findings are shown in Table 4.11

37
Table 4.11: Determinant of government borrowing in Kenya Regression Results
Dependent Variable: Domestic Government Borrowing
Independent Variables Coefficients Std. Error t-Statistics Probability
Domestic government borrowing (-1) 0.213821 0.176318 0.121271 0.0359
Domestic government borrowing (-2) -0.326779 0.166953 -1.957315 0.0079
Budget Deficit -2.995225 1.166822 -2.56699 0.0280
Budget Deficit (-1) 2.527123 0.676708 3.734439 0.0039
Domestic Savings -0.462211 0.129248 -3.576162 0.0050
Domestic Savings (-1) 0.285444 0.103155 2.767124 0.0199
Domestic Savings (-2) -0.243883 0.084349 -2.891354 0.0161
Inflation 0.021715 0.075712 0.286809 0.0080
Interest Rate 0.010571 0.067952 0.15556 0.0879
Interest Rate (-1) 0.081690 0.067427 1.211536 0.2535
Interest Rate (-2) 0.083710 0.059793 1.399997 0.1918
Lending Rate -0.364464 0.150025 -2.429358 0.0355
Lending Rate (-1) 0.790515 0.172209 4.590432 0.0010
Lending Rate (-2) -0.523635 0.173034 -3.026199 0.0128
Political Influence -4.005674 0.794059 -5.044556 0.0005
Political Influence (-1) -0.517746 0.859389 -0.602459 0.5603
Constant 51.23558 12.50675 4.096635 0.0022
R-Squared 0.892400
Adjusted R-Squared 0.814721 F-Statistics 14.22003
Durbin-Watson Statistics 1.999546 Probability 0.000075
Source: Author Computations

The value of Durbin-Watson statistics which measures auto-correlation among the variables

was found to be 2.0 and based to the law of the thumb, Durbin-Watson statistics value above

1.8 signifies absence of autocorrelation, therefore the study concludes that the variables used

had no auto-correlation problems hence no chances of obtaining spurious results.

The constant term was positive (51.24) and statistically significant at 5 percent significance

level showing that when there are no determinants the research deliberated, the level of

38
domestic government borrowing would by 51.24 billion Kenya shillings at period zero, the

coefficient of domestic government borrowing in one-period lag was 0.2138 lower than at

period zero and statistically significance at 5 percent significance level as the p-value

(0.0359) was below 0.05, however, in two-period lag the coefficient of domestic government

borrowing was negative (-0.3268) and significant statistically at 5 percent significance level

showing that in the long-run domestic government borrowing would be negative and

statistically affect other sectors of the economy in terms of investment in essential areas

leading to crowding out of the private sector. The results agree with Cooray (2019) that

increasing loans acquired by the government locally crowd private sector and at the same

time is inflationary in the economy.

The coefficient of budget deficit was negative (-2.9952) and significant statistically at 5

percent significance level showing that a percentage addition in budget deficit government

borrowing would decline by 2.99 points, however, at one-period lag, the coefficient was

positive (2.5271) and statistically significant at 5 percent significance level showing that

eventually, an addition in budget deficit by one percent, domestic government borrowing

would also increase by 2.527 points. This means that budget deficit positive and significantly

determine level of domestic government borrowing in the economy. The findings corroborate

Hasnat & Ashraf (2018) increased budget deficit significantly influence government

borrowing in an economy to finance services such as education, health, agriculture and

infrastructural facilities further widening the budget deficit gap and therefore in order to

bridge the gap, the government has to continue borrowing leading to increased domestic

government borrowing if the government resorts to borrow internally which crowd-out the

private sector. The findings also confirm Zaheer, et Al., (2017) that found a positive and

statistical significance link on the two variables. The research argued that broadening budget
39
shortfall gap improves local government borrowing as the government change focus to

internal sources of funds to finance the deficit and fund the essential sectors such as

education, health and infrastructure with high speed of adjustment of 252.71 percent.

The coefficient of domestic savings was negative (-0.4622) and significant statistically at 5

percent significance level at zero-period lag but positive (0.2854) at one-period lag implying

that the speed of adjustment to correct the negative effect of domestic savings on domestic

government borrowing is low at 28.54 percent meaning domestic savings slightly determines

domestic government borrowing. Additionally, the coefficient of local savings was found to

be negative (-0.2439) and significant at 5 percent significance level at two-period lag

implying that the adjustments declines to 24.39 percent meaning that in the long-run,

domestic savings negatively determine domestic government borrowings. The findings

confirm Lidiema (2018), Ado & Ibrahim (2019) that domestic savings is negative and

significantly influence domestic government borrowing as domestic savings diminishes in the

long-run as the incentives to save would decline leading government to borrow more

domestically to spend on key sectors of the economy.

The coefficient of rate of inflation was positive (0.0217) and significant statistically at 5

percent significance level showing that a percentage addition in rate of inflation results in an

addition in loans acquired by the government locally by 2.17 percent meaning that

government borrowing would increase even more so as to fund the key sectors of the

economy. The finding corroborates Hasnat & Ashraf (2018) high levels of inflation in an

economy is most likely to lead to an increase government borrowing due to sensitivity of

interest rates necessitated by high levels of inflation. However, the study revealed that rate of

lending was negative (-0.3645) and significant statistically at 5 percent significance level
40
showing that a percentage addition in rate of lending would lead to a reduction in government

borrowing domestically by 0.3645 points, similar to a two-period lag where government

borrowing would decline by 0.5236 points, this indicates that the rate of adjustment is

moderately high at 52.36 percent further decline in domestic government borrowing meaning

that government resorts to other source of funds externally meaning that eventually, rates of

lending negatively affect government borrowing from domestic sources, the findings

disagrees with Zaheer et Al., (2017) that ended up with a positive significant link between the

rate of lending and government borrowing domestically temporarily and long-run periods.

Lastly, incorporating political influence as an intervening variable in the model, the variables

the study considered significantly influence domestic government borrowing meaning that

political stability in a country has a role to play in domestic government borrowing.

4.5.2 Effect of government borrowing on private sector capital in Kenya

This is objective two of the research, to attain it, the study carried out ARDL regression

analysis of the local credit to private sector on budget shortfall, rate of interest at which loans

are advanced, rate of local savings, rate of inflation in the economy, lending rates by

commercial banks, level of domestic government borrowing to find out the impact of loans

acquired locally by the government on private sector capital in Kenya. Both efficiency of tax

agency and political influence were used as dummy variables in the model. The findings are

presented in Table 4.12

41
Table 4. 12: Effect of government borrowing on private sector capital results
Dependent Variable: Private Sector Capital
Independent Variables Coefficients Std. Error t-Statistics Probability
Private Sector Credit (-1) 1.148104 0.204033 5.627049 0.0001
Budget Deficit 0.078952 0.213167 0.370375 0.7166
Interest Rate 0.150322 0.69739 2.155502 0.0490
Interest Rate (-1) -0.125228 0.074560 -1.679566 0.0152
Domestic Savings -1.435425 1.300416 -1.103820 0.2883
Domestic Savings (-1) -1.171204 0.808838 -1.448008 0.1696
Inflation 0.093875 0.072487 1.295050 0.2163
Inflation (-1) -0.177559 0.077695 -2.285320 0.0384
Lending Rate -0.661488 0.191187 -3.459904 0.0038
Lending Rate (-1) 0.701279 0.194195 3.611213 0.0028
Government Borrowing -6.866802 1.552419 -4.423291 0.0006
Political Influence 0.632884 0.998637 0.633748 0.5365
Efficiency of Tax Agency (ETA) -2.104911 1.502198 -1.401221 0.1829
Constant -55.21516 13.69560 -4.031599 0.0012
R-Squared 0.921374
Adjusted R-Squared 0.882749 F-Statistics 24.8895
Durbin-Watson Statistics 2.506782 Probability 0.0000
Source: Author Computations
The study adopted ARDL model to estimate equation 3.4 because the variables were co-

integrated of different orders that is at level and at first difference implying that OLS was not

suitable to find out the impact of local government borrowing on private sector credit.

The results show that the value of adjusted R-squared is 0.882749 showing that 88.27 percent

of the variations in private sector credit is determined by the alterations in budget shortfall,

domestic savings, rate of interest, level of inflation in the economy, lending rate by banks,

level of government borrowing from internal sources. However, political influence and

efficiency of tax collection agency was used as control variables to determine whether if

considered there was an effect in the changes. This means that only 11.73 percent of the

42
variations are decided by other variables the study did not capture but was taken care of by

the error term in the model. The value of F-statistics was found to be 24.89 with a p-value of

0.0000 below 0.05 is significant statistically at 5 percent significance level showing that the

model was suitable to decide the impact of loans acquired by government on private sector

capital in Kenya. On the other hand, Durbin-Watson value of 2.5 is higher than 1.8 at 5

percent significance level showing an absence auto-correlation within the variables and hence

the study used the variables to find out the impact of local loans acquired by government on

private sector capital.

The constant figure was found to be negative (-55.21) and significant statistically at 5 percent

significance level, showing that when the factors are absent the research considered, private

sector capital available was zero or in a deficit and therefore not enough for all the borrowers

both public and private, similarly, this also shows that the capital accessible to private sector

is not adequate to be advanced so as to trigger investment in the economy if government also

targets the capital to provide essential services in the economy, this drives out the private

investor in the long run by 114.8 percent as indicated by the coefficient of private sector

credit of 1.148 in the long-run implying the effect of government borrowing from internal

sources are greater and therefore have a huge impact on the economy eventually.

The research also determined that the coefficient of budget shortfall was positive (0.07895)

but significant statistically at 5 percent level of significance, the finding contradicts Hasnat

and Ashraf (2018) that found a negative and statistical significance meaning that at budget

deficit widens, capital available to private sector decreases or diminishes. However, the

current study opines that as budget deficit gap widens, capital available to private sector also

increases meaning in case of budget deficit in the economy, the government would resort to
43
other sources of finance such as borrow externally to finance the budget deficit hence the

amount of capital available to private sector also increases though budget deficit

insignificantly determines the amount of capital available to private sector.

The coefficient of rate of interest in the economy was determined to be positive (0.1503) and

significant statistically at 5 percent significance level, showing that a percentage point

addition in rate of interest, capital available to private sector borrowers also increases by

15.03 percent in short-run this is because a higher interest rates motivates lenders to advance

more loans to as to gain more in future when the borrowers will be making the payment of

principle amount plus interest earned. However, in the long-run, interest was found to be

negative (-0.1252) and statistically significant at 5 percent level of significance implying that

one percent point increase in interest rates, amount of domestic capital available to private

sector decreases by 0.1252 points. The finding confirms that of Ehalaiye et Al., (2017),

Lidiema (2018) as well as Hasnat & Ashraf (2018) that found significant negative

relationship between interest and private sector capital meaning that an increase in interest

rates lowers or reduces amount of capital available to private sector as borrowers would

reduce their credit appetite because of fear of paying more in future when the principal

amount and interest payment matures holding their current borrowing for future when the

interest rates declines. Literatures have also shown that high interest rates negatively affect

investment as a higher interest rates reduces amount of return realized from an investment

opportunity and therefore, potential investors would hold borrowing to invest due to high

interest rates this in turn reduces amount of capital or credit to private sector which the

finding of the study confirms though the analysis in long-run with the rate of adjustment of

12.52 percent.

44
Further, the study revealed that domestic savings negatively and insignificantly affect private

sector capital available both temporarily and eventually. The finding contradicts Lidiema

(2018) that realized a negative and significant impact of local savings on private sector

capital availability, this is because as the amount of domestic savings increases, government

would resort to internal borrowing which significantly affect the amount available for private

borrowers negatively affecting the amount of capital accumulation for private sector.

Additionally, the inflation rate coefficient was determined to be positive (0.0938) and

insignificant statistically temporarily, however, eventually, the coefficient was determined to

be negative (-0.1776) and significant statistically at 5 percent significance level showing that

an addition in inflation rate by one percent point, domestic private sector capital decreases by

0.1776 points. The finding corroborates Hasnat & Ashraf (2018) that found a long-run effect

of inflation rates on domestic capital available to private sector, this is because inflation has

the effect of hiking prices of goods and services and similarly the cost of credit will also go

up diminishing the amount available for private borrowers as credit to private investors

becomes more expensive hence the borrowers resort to hold current borrowing for future

when inflation rate declines so as to invest and earn higher returns.

The study also found that the coefficient of rate of lending to be negative (-0.6615) and

significant statistically at 5 percent significance level temporarily showing that a percentage

point addition in rate of lending, private domestic capital declines by 0.6615 points this

corroborate findings by Zaheeri et Al., (2017),this is because as central bank increase its

lending rates to commercial banks, the amount of capital or funds set aside by the commercial

banks for domestic borrowing decreases immensely affecting the sector negatively. In the

long-run, the coefficient was determined to be positive (0.7013) but significant statistically at

5 percent significance level, this disagrees with Lidiema (2018) and Zaheeri et Al., (2017)
45
that found negative and significant effect both temporarily and eventually. The positive link

eventually based on the present research is because the commercial banks would have

adopted to the rates and set aside enough funds to lend to private borrowers hence domestic

borrowing would increase as lending rates by central banks to commercial banks increases,

this is due to promising investment opportunities available in the economy that motivates

privates investors to borrow more in order to invest therefore, lending rates by the central

bank to commercials banks in long-run have positive impact of local credit given to private

sector.

The coefficient of domestic government borrowing was determined to be negative (-6.8668)

and significant statistically at 5 percent significance level showing that a percentage point

addition in acquisition of loans by government locally, domestic credit available to private

sector decreases by 6.867 points. The finding corroborates Ayturk (2017), Ehalaiye et Al.,

(2017), Tkacevs & Vilerts (2019) and Ado & Ibrahim (2019) determined that a negative and

significant link between loans acquired by government locally and private sector capital, this

is because government borrows at a higher interest rates than private sector and the loan is

guaranteed due to low rate of default as compared to private sector. The continued

government borrowing from internal sources constraint the amount of funds available to

private borrowers hence driving the private sector out of the investment opportunity, the

finding corroborates Lau et Al., (2019) found out that increased domestic government

borrowing crowds out private investment in the economy as financial institutions are more

sensitivity to government borrowing than private borrowing. This is because commercial

banks reduce amount to lend to private sector or otherwise lend at a higher interest rates due

to high rate of default and therefore, the amount of capital to private sector reduces by a

bigger margin as opined by the study findings. However, the finding disagrees with Cooray
46
(2019) that loans acquired by government does not crowd out private sector investment but

rather inflationary as government borrows internally at a higher interest rates to provide

essential goods and services at a higher price in order to generate revenue to repay the

principal amount plus interest rate. The escalation in prices of services and goods leads to

general increase in price levels in the economy which is inflationary as opined by Cooray

(2019) rather than crowding out the private sector.

47
 CHAPTER FIVE

SUMMARY, CONCLUSION AND RECOMMENDATIONS

5.1 Introduction

The chapter covers summary of study, policy implications, conclusion, contributions to

knowledge and suggestions for further research.

5.2 Study Summary

Economic output in Kenya has been low over the years since independence, this is because of

stagnation in the development of key sectors that immensely contributes to economic growth.

For instance, agricultural sector contribution declined to about 3.9 percent while inflation rate

remains high of 100 percent all time in early 90s. In the year 2000s, the government of Kenya

initiated adopted policies that lead to growth of GDP to 7 percent in 2007, the highest since

independence, however, in 2008 the growth declined to negative six percent this forced the

government to revert to acquiring loans from both local and external sources because of

constraint revenue generation. In recent times the government of Kenya has been running on

a budget with shortfall and relies on loans to offer services to her citizens such as education,

health, infrastructure as well as security.

By 2018, the government had incurred debts amounting to 5 trillion Kenyan shillings which

is equivalent to about 60 percent of GDP CBK, report (2019) to continue providing key

service to the citizen. Kenya appetite for loans has increased to about 10 trillion Kenya

shilling as the debt ceiling was increased to Kshs. 10 trillion forcing the government to

increase her loan mixture opting for more domestic borrowing against external borrowing,

this has resulted to decline in private investment as well as business. This has forced private

sector out of investment opportunities what is referred to as crowding out of private sector as

domestic credit available to private sector declined from 78 percent in 2013 to 72 percent in

2017, this is coupled with closure of three commercial banks in 2016 affecting liquidity in the
48
economy. In spite of this, domestic government borrowing has gone up significantly

crowding out private sector. Even though credit to private sector still available, the

continuous increased domestic government borrowing rises alarm and therefore, there is need

to conduct a study to analyze crowding-out of private sector because of the loans acquired by

government of Kenya with specific focus on investigation of determinants of government

borrowing and determine the impact of loans acquired by government on private sector

businesses in Kenya.

To attain the two objectives, the research used non-experimental research design with yearly

time-series data for the duration between 1990 and 2020 from Central Bank of Kenya, World

Bank, Kenya National Bureau of Statistics as well as Kenya Bankers Association. Both pre-

test and post-test analysis were done to ensure that on chances of obtaining spurious results.

The stationary test was done to determine the co-integration order to allow determine the

model to be used in the analysis. The variables were found to be stationary at a point and at

first variation and hence the most appropriate model to be used was Auto-regressive

Distributed lags to approximate the coefficients of the components in the two linear equations

on achieving the two objectives of the research.

To examine the factors influencing local acquisition of loans by government in Kenya, the

study used domestic government borrowing amounts as dependent variable while budget

deficit, domestic savings, inflation rate, interest rate, lending rate as independent variables

with political influence as intervening variable. The research found that budget deficit is

negative and significant temporarily while positive eventually this was similar to local

savings, and borrowing rates. Inflation rate was determined to be positive and significantly

determine government borrowing, however, interest rate was found to positively and
49
significantly determine government borrowing only temporarily while eventually it was

found to insignificantly determine domestic government borrowing. Lastly, the study

established that political influence has no impact on the model investigating the determinant

of domestic government borrowing in Kenya.

To determine the impact of local government borrowing on internal private sector capital, the

research used local credit given to private sector as dependent variable while budget shortfall,

domestic savings, interest rate, inflation, lending rate and domestic government borrowing as

the independent variables while efficiency of tax collection agency and political influence as

control variables. The study found that budget deficit in Kenya positively and insignificantly

affect private sector capital, interest rate positively and significantly affects private sector

capital in short-run while in long-run, is negative and significantly affect private sector

capital. Level of domestic saving insignificantly affect private sector capital both in

temporarily and eventually, further, inflation insignificantly affect domestic credit advanced

to private sector temporarily but eventually inflation rate negative and significantly affect

domestic credit advanced to private sector. Lending rates by central bank to commercial

banks negative and significantly influence private sector capital in short-run while in long-run

it positively influence amount of capital available to private sector. Lastly, the study revealed

that government borrowing from domestic sources is negative and significantly affect amount

of capital to private sector.

5.3 Conclusion

The study sought to carry out an analysis of crowding out of private sector by local

acquisition of loans by government in Kenya with a focus to address two specific objectives.

These objectives were firstly to examine the factors influencing local government borrowing
50
in Kenya and secondly to find out the impact of government borrowing on domestic private

sector capital in Kenya. Depending on the research findings, conclusions were drawn as

follows.

Domestic government borrowing is determined by the level of budget deficit, domestic

savings, inflation rate in economy and lending rates. Budget deficit was found to positive and

significantly determine government borrowing in the long-run, therefore, a huge budget

deficit in an economy stimulates government to resort to borrow to fund the budget deficit

and it becomes more worse when the government resort to borrow internally this is because

the government competes with the private investors on the available funds but since

government has low chances of default therefore commercial banks or lending institutions are

guaranteed of future repayment at a higher interest rates. Similarly, domestic savings was

found to negative and significant in determining government borrowing in the long-run, this

is because domestic savings constitute capital formation in an economy, a high level of

domestic savings encourages government to resort to domestic sources of funds as the debts

have no strings attached as well as favorable borrowing terms as opposed to external

borrowings. A higher capital stock realized through savings than the amount required by the

domestic investors encourages government to borrow internally and invest the funds in the

key sections of the economy like health, education, roads, agriculture and electricity as well

as manufacturing which further yield a higher return in the economy leading to economic

growth as well as economic development. However, in the long-run, as domestic savings

increase, domestic government borrowing decreases, this is due to the fact the government

will have accumulated a higher stock from the investment made from previous borrowing and

therefore any level of domestic savings leads to low domestic government borrowing.

51
On the other hand, inflation rate positive and significantly determine government borrowing,

this is because as inflation rate increases, government would resort to borrow more internally

in order to provide goods and services to avoid over-burdening the citizens of the high cost of

living. High inflation rate in the economy also forces government to borrow internally as

opposed to low inflation rate according to study findings. This government would borrow to

provide goods and services at a lower price which is affordable by the larger population of

which majority are poor. When government provide essential goods and services at lower

prices, then, this ensures that inflation remains within the band encouraging further

borrowing by the government from domestic sources. The study also established that lending

rate negative and significantly determine domestic government borrowing, this is because

low lending rates by the central bank stimulates domestic government borrowing in an

economy as the government is able to incur these debts at affordable cost and investment in

sectors that earns future higher returns therefore repays the debt in time or when fall due

without constraining the economy.

Secondly, the study findings on the impact of loans acquired by government on private sector

capital in Kenya showed that domestic government borrowing negatively and significantly

affect the level of private sector capital. As government increases her borrowing from

domestic sources, the amount of capital advance to private investor decreases implying that

commercial banks would not advance loans to the private investors, this forces the private

sector not to invest and rely on government services and goods and in the end push private

sector investors out of the business commonly referred to as crowding-out of the private

sector investment. The outcomes also show that the crowding out effect is huge as a

percentage increment in loans taken by the government results in six percent of private sector

investors crowded out of investment.

52
Interest rate was also found negative and significantly affect private sector capital level, this

is because a higher interest rates discourages borrowing by the private sector as this

consumes part or the entire returns earned from the investment crowding out the private

sector investors. However, lending rates were found to be negative and significantly affect

private sector capital, this is because high lending rates by central bank reduces the sum of

accessible credit by the commercial banks to private borrowers as commercial banks have

less funds to lend, this makes loans to look less attractive to the borrowers hence fail to

investment thereby crowding out the private sector investment.

5.4 Policy Implications

Depending on the research results, the policy implications made are; Budget deficit should be

minimized as much as possible by reducing unnecessary government expenditure, this is

supported by the fact that a one percent point increase in budget deficit, government

borrowing from domestic sources increases by 2.527 points. Similarly, domestic savings

should be encouraged as this stimulates domestic investment due to availability of stock of

capital for investment, this is backed by the fact that one percent addition in local savings,

government borrowing from domestic sources declines by 0.2439 points. Further, inflation

rate should be maintained as low as possible below 5 percent in order to stimulate

government borrowing, this is affirmed by the findings that a percentage addition in inflation

rate, loans acquired by government goes up by 0.0217 point. The low inflation ensures that

essential services provide by the government are affordable to the majority of the citizens

because inflation rate is known to have the effect of increasing prices pf goods and services in

the economy. Lending rates should be kept low by central banks to increase lending by

commercial banks as this ensures enough money is in circulation to facilitate borrowing and
53
investment by the government in key sectors, this is backed by the findings that a percentage

addition in lending rates, loans acquired by government goes down by 0.5236 points. Lending

rates can be maintained at a level that creates a conducive environment for borrowing to

occur and encourages investment in the economy so as to spur economic growth as well

development.

The research recommends that local government borrowing should be avoided as much as

possible and resort to other sources of funds to finance expenditure such as health,

infrastructural development, education and manufacturing. This is backed up by the fact that

a percentage addition in local government borrowing results in crowding out of private sector

investment by 6.8668 points. Government ability to borrow internally has the effect of

stimulating other factors like interest and inflation rates and eventually widens the budget

deficit gap. Government borrowing would be inflationary when the debts incurred are used

for recurrent expenditures as opposed to capital expenditures, recurrent expenditure has the

effect of increasing the amount of money in circulation, this stimulates demand for goods and

services by the households resulting to excess demand in the economy. The excess demand

leads to price increase according to the low of demand, the spiraling cost of services and

goods is an indication that a lot of money purchasing few goods leading to inflation and this

is why the study recommends that government should borrow to invest in development

projects that earns returns in future and be able to repay the principal amount together with

the interest earned. Similarly, government borrowing also trigger interest rates in the

economy, this is because government borrows at a higher interest rates than privates

borrowers. The high cost of loans to private sector crowd-out the sector and therefore interest

rates should be maintained as low as possible to avoid crowding out of private sector

54
investment. This can be achieved through reviews or legislation of interest rates that

encourages borrowing so as to invest and enhance economic output of the economy.

5.5 Contribution to Knowledge

The study has demonstrated that government acquiring loans from local sources results in

crowding out of the private sector investment and has also recommended ways through which

government acquiring loans from local sources can be reduced to ensure private sector

investment and spur economic growth of the economy. The study has also shown that policy

makers as well as government should design that encourage borrowing externally and at the

same time investment in development projects as opposed to recurrent expenditures that is

only inflationary and also leads to increased cost of borrowing to the private sectors

investors.

Further, the study has also shown that budget deficit forces government to acquire more loans

in order to fund the budget shortfall and that interest rates also reduce borrowing by the

private sector leading to crowding out of the sector from the economic activities in the

economy. Moreover, inflation rates have been found to increase government borrowing

internally as well as crowds out the private sector by a larger margin.

Lastly, this is among the few studies conducted in Kenya that provides a non-experimental

analysis of crowding out of private sector by the acquisitions of loans locally by the

government in Kenya utilizing time-series data with Auto-regressive Distributed Lags

(ARDL) model.

55
5.6 Areas for Further Research

The present research carried out an analysis of crowding out of private sector by acquisition

of loans locally by the government in Kenya and has shown that it crowds out the private

sector by reducing the amount of capital available to private sector. The study suggests that

further research be done to find out the impact of interest rates on domestic government

borrowing in provision of essential services in Kenya. The research also suggests that further

studies be conducted to find out the value of private sector to government revenue generation

and lastly, the study suggests that a similar research should be conducted by utilizing panel

data to show a comparative analysis of the crowding out of private sector across nations so as

to make policies that can be implemented across countries.

56
 REFERENCES
Ado, A., & Ibrahim, A. (2019). Government Borrowing Behaviour: Implications for Private
Sector Growth in Nigeria. Policy, 8(2), 68-82.
Ahmed, H., & Miller, S. M. (2000). Crowding‐out and crowding‐in effects of the components
of government expenditure. Contemporary Economic Policy, 18(1), 124-133.
Alesina, A., &Perotti, R. (1994). The political economy of growth: a critical survey of the
recent literature. The World Bank Economic Review, 8(3), 351-371.
Ayturk, Y. (2017). The effects of government borrowing on corporate financing: Evidence
from Europe. Finance research letters, 20, 96-103.
Barro, R. J. (1979). On the determination of the public debt. Journal of political
Economy, 87(5, Part 1), 940-971.
Barro, R., &Grilli, V. (1994). The public debt. In European Macroeconomics (pp. 343-367).
Palgrave, London.
Beaugrand, P., Loko, B., & Mlachila, M. (2012). The choice between external and domestic
debt financing budget deficits: The case of Central and West African countries. IMF
Working Paper 02/79.8.
Brennan, G., & Buchanan, J. (1983). Predictive power and the choice among regimes. The
Economic Journal, 93(369), 89-105.
Brümmerhoff, W. (1998). Quarterly Bulletin.
Buiter, W. H. (1990). Principles of budgetary and financial policy. Cambridge, The MIT
Press.
Chhibber, A., Van Wijnbergen, S., & Mundial, B. (1988). Public policy and private
investment in Turkey. International Bank for Reconstruction and Development, World
Bank.
Cooray, L. A. N. (2019). Crowding-out effect of public borrowing in Sri Lanka. Journal of
Economics and Business, 2(3), 827-842.
Cruz, B. D. O., & Teixeira, J. R. (1999). The impact of public investment on private
investment in Brazil, 1947-1990. Cepal Review.
Ehalaiye, D., Botica-Redmayne, N., & Laswad, F. (2017). Financial determinants of local
government debt in New Zealand. Pacific Accounting Review.
Enders, W. (2004). Applied time series econometrics. Hoboken: John Wiley and Sons.

57
Fayed, M. E. (2012). Domestic Credit: Between Demand and Supply and Crowding Out
Effect. In Conference On Egypt's Future: Political and Economic Issues. Faculty of
Economics and Political Sciences, Cairo University.
Grilli, V., Masciandaro, D., &Tabellini, G. (1991). Political and monetary institutions and
public financial policies in the industrial countries. Economic policy, 6(13), 341-392.
Hasnat, T., & Ashraf, S. (2018). Is government borrowing crowding out long term debt
resources: An empirical enquiry from India. International Journal of Emerging
Markets.
Heinemann, F., &Illing, G. (2002). Speculative attacks: unique equilibrium and
transparency. Journal of International Economics, 58(2), 429-450.
Kendren, (2009): “Kenya’s Public Debt Status”. Nairobi
Kenya National Bureau of Statistics reports (2013). Facts and figures.
Kenya National Bureau of Statistics reports (2013). Gross Domestic Product first quarter
King’wara, R. (2015). Determinants of dividend payout ratios in Kenya. Research Journal of
Finance and Accounting, 6(1), 48-51.
Krugman, P. (1988). Financing vs. forgiving a debt overhang: Some analytical issues. NBER
Working Paper No. 2486.
Lau, S. Y., Tan, A. L., & Liew, C. Y. (2019). The asymmetric link between public debt and
private investment in Malaysia. Malaysian journal of economic studies, 56(2), 327-
342.
Levy-Livermore, A., Chowdhury, K. (1998), Handbook on Globalization of the World
Economy. Cheltenham, UK: Edward Elgar Publishing Limited
Lidiema, C. (2018). Effects of government borrowing on private investments in Kenya.
Journal of Finance and Economics, 6(2), 49-59.
Lucas Jr, R. E., &Stokey, N. L. (1984). Optimal growth with many consumers. Journal of
Economic Theory, 32(1), 139-171.
Macfarlane, L., Ryan-Collins, J., Bjerg, O., Nielsen, R. H., & McCann, D. (2017). Making
Money from Making Money: Seigniorage in the Modern Economy.
MacLeod, H. D. (1906). The theory and practice of banking (6th ed.). London: Longman,
Greens and Co.
McLeay, M., Radia, A., & Thomas, R. (2014). Money creation in the modern economy. Bank
of England Quarterly Bulletin, Q1.

58
Murungi, S. M., & Okiro, K. (2018). Impact of government debt on economic growth in
Kenya: A critical literature review. European Scientific Journal, 14(1), 240-257.
Ouattara, B. (2004). Modelling the long run determinants of private investment in
Senegal (No. 04/05). Credit Research Paper.
Persson, T., &Svensson, L. E. (1989). Why a stubborn conservative would run a deficit:
Policy with time-inconsistent preferences. The Quarterly Journal of
Economics, 104(2), 325-345.
Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of
level relationships. Journal of applied econometrics, 16(3), 289-326.
Ramírez, M. D. (1994). Public and private investment in Mexico, 1950-90: An empirical
analysis. Southern Economic Journal, 1-17.
Roubini, N., & Sachs, J. D. (1989). Political and economic determinants of budget deficits in
the industrial democracies. European Economic Review, 33(5), 903-933.
Sargent, T. J., & Wallace, N. (1981). Some unpleasant monetarist arithmetic. Federal reserve
bank of minneapolis quarterly review, 5(3), 1-17.
Serven, L. (1999). Does public capital crowd out private capital? Evidence from India. The
World Bank.
Švaljek, S. (2009). Government Revenue Forecasting in Croatia: Why Errors
Occur?. Privrednakretanjaiekonomskapolitika, 19(119), 53-68.
Tkačevs, O., & Vilerts, K. (2019). The impact of government borrowing costs on fiscal
discipline. Kyklos, 72(3), 446-471.
Turner, A. (2012). Credit creation and social optimality. International Review of Financial
Analysis, 25, 142-153. https://doi.org/10.1016/j.irfa.2012.09.004
Were, M. (2011). The impact of external debt on economic growth and private investment in
Kenya: An empirical assessment. WIDER Discussion paper No 2001/116; United
Nation University
Werner, R. A. (2014). Can banks individually create money out of nothing? – The theories
and the empirical evidence. International Review of Financial Analysis, 36, 1-19.
https://doi.org/10.1016/j.irfa.2014.07.015
Werner, R. A. (2016). A lost century in economics: Three theories of banking and the
conclusive evidence. International Review of Financial Analysis, 46, 361-379.
https://doi.org/10.1016/j.irfa.2015.08.014

59
Whittaker, M. (2007). South Africa's National Credit Act: A possible model for the proper
role of interest rate ceilings for microfinance. Nw. J. Int'l L. & Bus., 28, 561.
Zaheer, S., Khaliq, F., & Rafiq, M. (2017). Does government borrowing crowd out private
sector credit in Pakistan. State Bank of Pakistan, Working paper, 83.

60
61