DO BUDGET DEFICITS CROWD OUT PRIVATE INVESTMENT? AN ANALYSIS OF THE SOUTH AFRICAN ECONOMY

ecuadorianaceΔιαχείριση

28 Οκτ 2013 (πριν από 4 χρόνια και 12 μέρες)

92 εμφανίσεις


DO BUDGET DEFICITS CROWD OUT PRIVATE INVESTMENT? AN
ANALYSIS OF THE SOUTH AFRICAN ECONOMY

By


Biza R. A
1
;

Kapingura F.M
2

and Tsegaye A
3





ABSTRACT

South Africa has been experiencing unprecedented budget deficits since the 1960s, in light of
this, the

study

investigates whether budget deficits crowd out or crowd in private investment
in South Africa, using quarterly data covering the period 1994 to 2009. An empirical model
linking private investment to its theoretical variables is specified and used to

assess the
quantitative effects of budget deficits on private investment. Th
e

study augments the co
-
integration and vector auto
-
regression (VAR) analysis with impulse response and variance
decomposition analyses to provide robust long run and short run dy
namic effects on private
investment. The variables have been found to have a long run relationship with private
investment. Results suggest that budget deficits significantly crowds out private investment.
These results corroborate the theoretical predicti
ons and are also supported by previous
studies.


Keywords:

Economic growth, Budget Deficits, Investment, South Africa, Cointegration, VECM


JEL Classification: H30, H62, H63












Paper prepared for presentation at the
Financial Globalisation and
Sustainable Finance:
Implications for Policy and Practice

(
29
to

3
1 May, 2013, Cape Town, South Africa)




1

British High Commission, 255 Hill Street, Arcadia Pretoria, South Africa

2

University of Fort Hare, Department of Economics, East London Campus
,

East London, South Africa
,

Emai l: fkapi ngura@ufh.ac.za

3

Uni versi ty of Fort Hare, Department of Economi cs, Ea
st London Campus, East London, South Afri ca


1.

INTRODUCTION

Although increased government involvement in the economy may crowd
-
out private
investment as supported by the neoclassical school, public

spending may help in developing
infrastructure for encouraging private investment. Accordingly, the Keynesian model argues
that an increase in the government spending stimulates the domestic economic activity and

crowds
-
in”

private investment (Darrat, 19
89). The model asserts that the government might
help lay the ground for the development of private sector through the provision of legal
infrastructure that ensures physical and intellectual property rights. The government also
helps by undertaking invest
ments that intensifies the physical and human capital
infrastructure in the country.

One of the bones of controversy between Keynesians and Monetarists is on the effectiveness
of fiscal action in stimulating economic activity.
According to Easterly and Re
belo (1993)
fiscal deficits have been blamed for the variety of ills that beset developing and industrial
countries. These ills include overindebtness, high inflation and poor investment and growth
performance.
The driving force behind economic policymakin
g lies in the macroeconomic
objectives which are price stability, employment, economic growth, balance of payments
surplus and equity (Chowdhury and Hossain, 1998: 130). Economists have a common belief
that budget deficits are harmful for the total functio
ning of the economy.

The South African economy is a prominent emerging economy and its government
expenditures show direction to infrastructure and include investment in public works,
education, and health care facilities. Consequently to this expenditure

in productive
investments and human capital infrastructure by the state, the expenditure might crowd in,
rather than crowd out, private sector investments. In addition, the South African government
may use fiscal policy that involves increased spending in

infrastructure projects as an
aggregate demand management tool. If such policies turn out to be successful, decreased
macroeconomic volatility and a more stable level of aggregate demand may provide an
incentive for private businesses.

The decline in gov
ernment budget deficits in South Africa has had implications for the public
sector borrowing requirement, which declined substantially to around 0.9 percent of GDP in
2008. At the same time, although the government remains cautious about financing the defi
cit
through foreign borrowing, increased recourse has been made to this source of finance. The
combined effect of increased foreign borrowing and lower deficits has resulted in less
pressure on long
-
term bond rates and reduced the cost of government borrow
ing. Marcus
(2003) proposes that, in fact, the declining deficit and the moderately increased foreign
borrowing have led to a shortage of paper in the market, which has reinforced the downward
pressure on long
-
term interest rates.

It is also important to n
ote that a
lthough there is rich literature on the structure of interest
rates, the empirical impact of
this on budget
deficits
has received less academic attention.
Persistent budget deficits will result in tax cuts leading to expansion of aggregate
demand,
thus, private savings rise by less than the tax cut so that desired national savings declines.
Expected real interest rate will rise to restore equality between national savings and
investment demand; as a result
,

crowds out private investment whic
h will affect economic
growth. Budget deficits automatically increase government debt, requiring income to finance
the debt. Financing of the debt by domestic borrowing may raise interest rates which crowd
out private investment. Reliance on high taxes pla
ce a burden on future generations as it
reduces their income and spending accordingly. The higher taxes also distort economic
incentives and thus weaken future economic performance.

It is interesting to note that s
ince the transition to democracy, South A
frica’s macroeconomic
performance has been solid but not spectacular. Between 1994 and 2003 GDP growth has
fluctuated over this period, it has neither declined nor exceeded 4.5 percent in any calendar
year since 1993. In this respect, South Africa appears
locked into a path of sustained but
moderate growth.


Despite an improving situation, it must be emphasized that investment spending remains low.
Private investment spending averaged 12.1% of GDP between 1994 and 2003. Optimists
point to a number of facto
rs that could sustain the current acceleration and allow South Africa
to approach this target


higher levels of government spending “crowding in” further
investment on the part of the private sector. However, these higher levels of spending bring
about fi
scal deficits which will need to be financed through borrowing thus raising external
debt.


In light of the above, the study will seek to answer the following questions. Firstly, does

a
higher level of public spending and budget deficit crowd out or crowd

in private investment
in South Africa?
Secondly, g
iven the structure of government spending in South Africa,
increased government intervention also raises the level of employment and anchors growth of
the economy
,

thus does this crowding out effect really

matter?

Thirdly, t
o what extent will
interest rates rise in response to the greater demand for money and the supply of bonds
p
rompted by the government deficits?
And lastly, h
ow will investments fall in response to the
rise in interest rates?


The rest
of the paper is organised as follows. Section 2 presents a

background and

overview
of
budget deficits in South Africa
. Section 3 presents a selective review of relevant
theoretical and empirical literature on the
impact of budget deficits on investment
. Se
ction 4
provides the methodology employed in the stu
dy. Section 5 draws conclusions.


2.

BACKGROUND AND OVERVIEW OF BUDGET DEFICIT TRENDS IN
SOUTH AFRICA


The relationship between the government’s fiscal position and market interest rates breeds
much
controversy as claims have been made by some researchers and economists that there is
a strong relationship between the two, while others argue that there is no relationship at all.

Generally interest rates in South Africa have been higher than the deficit to GDP ratio. In the
early 1960s both the deficits and the rates of interest were low, a trend that continued until the
mid
-
1970s.The treasury bill rate was about 3% in the 1960s wh
en the yield on government
bonds was close to 5% and the lending rate of commercial banks was almost 6%. The rates
were stable at treasury bills being a little over 4%, government bonds at 6% and the lending
rate of commercial banks a little over 8% until
1970 (SARB, 1994). During the early and
mid
-
1970s, the rates increased such that by 1981, the Treasury Bill Rate (TBR) was about
10%, government bonds yielded 13% and lending rates 14%. This period coincided with
rising budget deficits. Interest rates were

still generally high in the 1980s, and so were budget
deficits. By the 1990s, interest rates began to fall such that by 1996, the TBR was 16%,
almost on a par with the yield on government bonds, although the commercial banks’ lending
rate was close to 20%

(SARB, 1996). From the 1960s to the early 1980s, it appears that the
interest rates and the deficit to GDP ratio moved in the same direction. Changes in the deficit
to GDP ratio have been followed by similar changes in interest rates, or vice versa. The
e
xtent of the variation in interest rates was more distinct in the early 1980s and beyond.
Whereas the deficit to GDP ratio changed slowly, there were sharp reductions in interest rates
around 1986

7 and 1992

3 (SARB, 1996).


Figure 1

shows the relationship

between budget deficits and interest rates on government
bonds of 0
-
3 years and 10 years and above for the period 1993 to 2009. Series 1 shows the
trends in budget deficits as a percentage of GDP. Series 2 shows trends in interest rates on
government bond
s of 0
-
3 years and series 3 shows the interest rate tendencies on government
bonds of 10 years and above. The variations in the different interest rates are analysed in line
with the trends in budget deficits to establish the relationship between these mac
roeconomic
variables.


Figure1: Budget deficit
-
interest rate relationship 1993
-
2009


Source: Department of trade and industry: Economics statistics, 2010


In 1993 both the deficits and the rates of interest were high, a trend that continued until mid
-
1994
. As budget deficits decreased from 7.3% of GDP in 1993 to 4.9% in 1996, interest rates
on government bonds of 0
-
3 years rose from 10.97% to 15.95% and rates on government
bonds of 10 years and above increased from 12.34% to 16.19%. At a level of 2.1% of b
udget
deficits in 1999, interest rates on bonds of 0
-
3 years decreased to 12.92% from a high level of
16.91% in the previous year. Interest rates on government bond of 10 years and above were
13.96% in 1999 from a high level of 16.36% in 1998. This trend s
upports the conventional
view that high budget deficits cause an increase in the long term interest rates.


Budget deficits continued to decline as there was an improvement in revenues as well as
economic performance. Budget deficits were at their lowest i
n 2005 at 0.5% of GDP. Interest
rates on bonds of 0
-
3 years declined to 7.27% and that of bonds of 10 years and above were
7.57%. The general trend at this period was that, as budget deficits declined interest rates also
declined but at a slower pace. In 2
006 South Africa experienced its unanticipated surplus of
0.3% of GDP; however interest rates (on bonds of 0
-
3 years) increased to 8.87% and rates on
10 year bonds and above slightly rose to 7.81%.


-10
-5
0
5
10
15
20
25
30
35
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
percent (%)

Interest rates on government bonds of 10
years and above
Interest Rates on government bonds of 0-
3 years
Budget Deficits (% of GDP)
In 2007 t
here was an increase in the
budget
surplus to 0
.7% of GDP
but on the other hand
interest rates on 0
-
3 year bonds increased significantly to 10.56% and those on government
bonds of 10 years and above increased to 8.29%. The rise in interest rates contradicted with
the economic activities as real output
was bolstered due to the improved economic
performance. However the global financial crisis impacted negatively on other sectors of the
economy as a result the surplus of 2007 turned into a fiscal deficit of 0.4% of GDP in 2008.
Interest rates on governmen
t bonds of 0
-
3 years declined to 9.87% in 2008 and rates of
government bonds of 10 years and above reduced to 7.82%. The budget deficit increased
substantially in 2009 to 4.9% of GDP due to the forceful performance of the economy as well
as increased gover
nment expenditure to finance infrastructural developments. Interest rates in
2009 on 0
-
3 year government bonds declined significantly to 7.56% while those on
government bonds of 10 years and above increased to 9.03%.


The external events associated with th
e East Asian crisis in 1998 negatively impacted on the
GEAR objectives (Freund and Padayachee, 1998). The rand depreciated extensively and
forced a monetary policy response that resulted in short term interest rates being increased by
400 basis points whil
e long term bond rates rose slightly. As a result the real rate of interest
moved from 3% in 1994 to almost 14 % in 1998 (Ibid: 82). Yan Kuo et al. (2003) assert that
under a repressed financial sector, taxes on financial assets are a major source of reven
ue for
the government.

On the other hand, in a liberalized financial system, where the government
finances its deficits via domestic borrowing, public sector will compete with the private
sector for loans. This puts upward pressure on interest rates.


According to Akinboade (2004) from the 1960s to the early 1980s, interest rates and the
deficit to GDP ratio moved in the same direction. Changes in the deficit to GDP ratio have
been followed by similar changes in interest rates, or vice versa. The magnit
ude of the change
in interest rates was more pronounced in the early 1980s and beyond. The change in deficits
from negative to positive percentages resulted to lower interest rates. There were sharp
reductions in interest rates around 1992

3 with interest
rates declining from 16.66% to 14.9%

The Reserve Bank has practiced inflation targeting by keeping interest rates high in order to
control domestic demand. In 1993
-
1998 the nominal interest rates rose while inflation was
declining, leading to a sharp
increase in real interest rate. Data indicate that the nominal
interest rate on long
-
term bond financing rose considerably in the 1970s and remained high
until around 2000, before declining quite considerably. The main reason for the increase in
nominal bo
nd yields in the 1970s was the acceleration in inflation to double
-
digit rates from
around 1974 (SARB, 2010). The SARB raised interest rates by 200 basis points in the last six
months of 2006 while the government controlled growth in expenditure. As econom
ic
conditions deteriorated with the unfolding global recession, the Monetary Policy Committee
(MPC) of the SARB progressively reduced the repo rate. The repo rate was lowered by 500
basis points, from 12% in November 2008 to 7% by August 2009. South Africa
’s commercial
banks tagged along, lowering the prime overdraft rate to 10.5% over the same period (SARB,
2010). This monetary reduction was supported by an improved inflation position, but the
MPC had also recognized the need to provide some recovery stimu
lus to a poor performing
economy. The MPC left the repo rate unchanged at its meeting held on 17 November 2009.



The relationship between budget deficits and interest rates remains
sceptical

as the trends
show different movements as a result of different
ways of financing. South Africa uses
domestic borrowing and foreign borrowing as a way to finance its debt and deficits thus, the
effects of the various forms of financing on interest rates and hence private investment are
sceptical
. The deficit
-
interest r
ate relationship
in figure 1

generally shows a negative
relationship between the two variables such that an increase in budget deficits results
in

lower
short and long term interest rates. It shows correlation between the two macroeconomic
variables.


Beyo
nd the relationship between fiscal deficits and interest rates, analysis o
f the impact of

fiscal deficit
s

on private investment has been
an area of interest
. Writing on the
macroeconomic determinants of domestic private investment in Africa, Mlambo and
Osh
ikoya (1999), using a sample of 18 African countries for the period running from 1970 to
1996, found that fiscal, financial and monetary policy, macroeconomic uncertainty, and trade
variables were significant determinants of private investment in Africa. F
igure
2

below shows
the budget deficit
-
interest relationship in South Africa from 1991 to 2009.


Figure2: Budget deficit
-
private investment relationship


Source: Department of trade and industry: Economics statistics, 2010

Figure 2 indicates that c
hange
in private investment was
-
5.53 and
-
2.38 percent in 1991 and
1992

respectively
. After a period of declining, change in private investment made a sharp rise
in 1993 and 1994 to 4.25 and 12.71 percent and began single
-
handedly to determine the trend
of Sout
h African investment. On the flip
-
side, budget deficits were 1.9 percent in 1991, 3.7
percent in 1992 and 7.3 percent in 1993. Structural adjustment programmes were put in place
in 1994 which put emphasis on fiscal adjustment and this resulted in the budge
t deficits
falling to 5.6 percent of GDP.


From 1995, private investment began declining before reaching a low change of
-
3.17 percent
in 1999. Change in private investment increased to 8.14 percent and continued to fluctuate
reaching its peak of 14.74 per
cent in 2004. Private investment continued to be positive and
increasing until in 2009 when it reached its low of
-
9.58 percent. This was due to the global
economic crisis that saw many economies plunging and as a result economic performance
was weak.


The

main shift in fiscal stance occurred in 1999 as the government sought to achieve fiscal
stability through the reduction of the deficit. Budget deficits continued to decline to 2.1
-15
-10
-5
0
5
10
15
20
percent (%)

Budget Def icits (% of GDP)
GFCF-Private Enterprises (%
change)
percent in 2000, 1.1 percent in 2003 and 0.3 percent in 2006. South Africa
experienced a
budget surplus in 2007 and 2008 of 0.7 percent and 0.9 percent before recording a 0.7 percent
budget deficit in 2009 as a result of deteriorating economic conditions as a result of the global
financial crisis which started in 2007/8. As budge
t deficits steadily declined due to better
revenue performance, change in private investment increased showing fiscal policy as a
significant determinant of private investment in South Africa.


3.

LITERATURE REVIEW AND THEORETICAL FRAMEWORK

Literature is abou
nding, with studies on the impact of budget deficits on private investment
and the crowding out effect. A large body of evidence comes from developed countries and
African literature is scarce. In particular, there are a few studies that have been done to
find
whether budget deficits have an impact on private consumption and investment and whether
the crowding out paradigm holds in the South African context.


Cebula et al. (1981) used three models in their study of the 'crowding out' effect of Federal
gove
rnment outlay decisions. Their study examined the crowding out effect of aggregate
federal government spending decisions upon purchases of new physical capital by private
firms. In particular, the study examined crowding out by determining to what degree t
he
proportion of actual GNP that was devoted to private investment in new physical capital was
affected by the proportion of actual GNP devoted to federal government spending. Their new
approach to the empirical dimension of the crowding out sought to prov
ide further insight
into whether or not the crowding out issue is substantive. The three alternative models were
estimated, all of which found evidence of (a) a definite pattern in which private investment is
crowded out by government spending and (b) only

partial crowding out.
These findings are
inconsonance with
Arestis (1979), Abrams and Schmitz (1978), and Zahn (1978)
.


Dewald (1983) tested for the effects of deficits on both short
-

and long
-
term interest rates,
using two different estimating technique
s. In one approach, annual data were used; in the
second approach, data were averaged over the business cycle. In either case, deficits had a
statistically significant effect on long
-

but not short
-
term rates.
The author concluded
that
deficits did not hav
e a large and consistent effect on interest rates. However, an equally
important conclusion would seem to be that deficits have an effect on long
-
term interest rates,
although not on short
-
term rates. This meant that deficits had an impact on long term
con
sumption and investment and did not affect short term investments, hence fiscal deficits
crowded out long term investments.


On the other hand, Hoelscher (1983) constructed an empirical model of the short
-
term credit
market, with the interest rate on three
-
month Treasury bills a function of the money base,
inflation expectations, a cyclical variable
-
proxied by the unemployment rate and net
borrowing by the federal government. The results indicated that the coefficient on
government borrowing was quite small

and not statistically significant.


Makin (1983) used a simple univariate

regression equation for the changes in the three
-
month
Treasury bill rate as a function of either the change in the actual deficit (relative to GNP) or
the change in the high employment deficit. The coefficients for the deficit variable were not
statistic
ally significant. Similar results of non
-
significance for the fiscal variable were also
reported for long
-
term interest rates. However, the analyses made were said to be faulted for
not including other important determinants such as the monetary base and i
nflationary
expectations.


According to Motley (1983), the budget deficit financed by borrowing from the private sector
leads to an increase in the supply of government bonds, and to attract the private sector to buy
these bonds, the government has to offe
r them at a low price, which essentially implies an
increase on interest rates, which causes crowding out of investment in the private sector. It
was noted that the loanable funds model predicts that, in the nonexistence of debt
monetisation, the effects o
f large fiscal deficits could lead to large effects on interest rates.
However, Brunner (1984) notes that portfolio analysis suggests that interest rates are
determined by stock demand and supply, and not by flow demands and supply, as suggested
by the loa
nable funds model as this could lead to either significant or negligible effects of
deficits on interest rates, depending on the sizes of accumulated stock of debts and deficits.

In Tanzi’s (1985) study of American statistics for the effects of federal fis
cal deficits on the
interest rate and macroeconomic performance, the results found a mixed association between
higher fiscal deficits and interest rates. The statistical significance of several alternative fiscal
measures in a model of interest rates on on
e
-
year Treasury bills was tested. In the
formulation, the interest rate was made a function of inflationary expectations using the
Livingston Index, the gap between real GNP and potential real GNP, government debt, total
private investments as a percentage

of real GDP and measures of the fiscal deficit, including
the deLeeuw and Holloway (1985) measure of the structural deficit. The study did not
include a specific variable to capture the effects of monetary policy, although the inflationary
expectations va
riable may indirectly capture some aspects of monetary policy. The results
found were mixed. The sign on the structural deficit was negative, suggesting that higher
deficits lowered interest rates and positively impacted on investment, a result that is dif
ficult
to reconcile with any theory. On the other hand, the actual debt variable had the hypothesized
positive and statistically significant effect.


Consistent with
Tanzi
’s results
Evans
(1985)
used a conventional Keynesian model to explain
why deficits m
ight be expected to affect interest rates which have an impact on private
investment. In his formulation, the nominal interest rate was a function of real government
spending, the real deficit, the real money stock, and expected inflation. While much of hi
s
analysis pertained to wartime, the period analysed of most interest was from October 1979 to
December 1983 (Evans, 1985). The study used monthly data and two
-
stage least squares
estimation to deal with the problem of the endogeneity of the deficit. In an
y case, the
coefficient on the deficit was usually negative and statistically significant. This result is
difficult to explain with any theory, and stopped short arguments that deficits lower interest
rates. The author did, however, argue that no empirical

support could be found for the notion
that deficits raise interest rates. The explanation was that the Ricardian equivalence theorem
must have held. But other interpretations were also possible, such as that international capital
mobility dominated intere
st rate movements for the period, or that price controls and
rationing accounted for the results for some of the wartime periods.


Easterly et al. (1994) in their study on public sector deficits and macroeconomic performance
pointed out that there was a po
sitive relationship between fiscal balances and interest rates.
This assertion was in contrast with the common prediction that deficits cause high interest
rates and surpluses low interest rates. There were a large number of negative real interest
rates in

the sample study and their finding was explained by the relationship between
financial repression and fiscal deficits. Economies which go through budget deficits risk the
possibility of financing their debt through seigniorage resulting to high inflation.

The study
showed little correlation between fiscal balances and inflation rates. In their study they
acknowledged the work done by Haan and Zelhorst (1990) in their study on the impact of
government deficits on money growth in developing countries that th
e correlation between
interest rates and inflation holds for countries with high inflation.


Ball and Mankiw

(1995) published one of the studies to test empirically for a relationship
between deficits and interest rates. Their model was in the Keynesian tradition, emphasizing
income determination. In their model, the ten
-
year government bond rate was made a
func
tion of the log of real per capita government debt, the log of real per capita
nongovernment GNP, expected inflation, the monetary base, and the lagged change in the
dependent variable. The results reported showed that the government debt variable had a
st
atistically significant though small effect, thus, revealing that budget deficits had a small
effect on the level of interest rate.


The

literature discussed above shows that the relationship between budget deficits and
investment is ambiguous. However, th
e
literature
has indicated

a number of significant
variables that are essential to investigate, in order to explain to what extent the literary
theories on the crowding out effect can explain the effect of the unprecedented budget deficits
on private inves
tment in the South African economy. It is obvious that economic indicators
such as budget deficit, interest rates, change in real gross domestic product and inflation are
relevant to investigate, in order to explain the impact on the South African economy.

Thus
the literature has provided a guideline in the selection of the relevant variables to be used in
the study.


4.

METHODOLOGY

Th
e

study
seeks to

investigate the crowding out effect of budget deficits on private
investment. The reason is because South Afri
ca has experienced budget deficits during most
parts of the study period and the study seeks to analyse whether these have detrimental
effects on private investment.


The study will adopt a model developed by

Blanchard and Perotti (2002).
T
he model is
spe
cified as follows:



………………………………………………………………
……… (4.1)

Where: PI is private investment, BD is budget deficits, IR is interest rates, PCY is change in
GDP and P is inflation. For analytical convenience the variables are all expressed either in
ratios or percentages. Because changes in the natural logarithm are (almost) equal to
percentage changes in the original series, it follows that the slope of a trend line fitted to
logged data is equal to the average percentage growth in the original serie
s. Percentage
changes in the variables and rates are not logged because of their negative nature. When
expressed, equation (1) becomes:


……………………………………………… (4.2)

Budget deficits resulting from public investment can act as a substitute (negativ
e effect on
private investment) to or a complement (positive effect on) for private investment. The sign
of the effect depends on the area in which the government executes the investment projects.
Public investment may encourage private investment when suc
h expenditure contributes to
increasing private owned firms’ productivity. On the other hand, it may crowd out private
investment when: (i) the government invests in inefficient state
-
owned firms; (ii) private
investors expect higher taxes to finance such
expenditures; and/or (iii) the public sector
competes with the private sector for domestic loanable funds. Thus the relationship between
the two variables is ambiguous. The anticipated sign on budget deficits is not clear


it can
either be positive or neg
ative because it is dependent on the area in which the government
executes the investment projects.


Interest rates are expected to have a negative impact on private investment as proved by
empirical theory (Easterly and Rebelo, 1993). This stems from the
assortment that if interest
rates are high then it makes it expensive to borrow money. This will deter investment because
investment is often financed through borrowing. Also when interest rates are high it makes it
more attractive to save money. Investmen
t is often financed out of retained profit. High
interest rates mean that investment is relatively less attractive than saving money in a bank.


Theoretically, the relationship between private investment and growth can be derived from an
accelerator model
with the underlying assumption that the production function has a fixed
relationship between desired capital stock and a level of real output. The a priori expectation
is that a percentage change in real gross domestic product will have a positive impact o
n
private investment. The expected sign for the percentage change in real GDP coefficient (

in
equation 4.2) is a positive sign. This follows from the path that a positive change in real GDP
will result in increased private investment and hence growth. Stu
dies by Hernandez
-
Cata
(2000), Kormendi and Mcguire (1985:145
-
163) prove beyond doubt that there is a positive
linkage between private investment and the rate of growth.


As an indicator of macroeconomic instability, inflation can hamper private investment

as it
spells out the level of government efficiency. Greene and Villanueva (1991:33
-
58) and
Oshikoya (1994:573
-
596) point out that, high and unpredictable inflation distorts the
information content of relative prices and increases the riskiness of longer
time investment.
Therefore, a negative association between inflation and investment is proposed thus a
negative sign of the coefficient is expected.


4.1
Data Sources and Measurement of Variables

The study will employ time series data for the period
between the first quarter of 1994 to
fourth quarter of 2009 by tracking trends, seasonal fluctuations and variations in data or
statistics. This period will be studied to establish the crowding out effect of budget deficits on
private investments, when the

government sought a leading era and enabling role in guiding
the mixed economy through reconstruction and development. The variables, in level form,
are of quarterly frequency and are obtained from various publications of the World Bank, The
South African

Reserve Bank (SARB) electronic data delivery systems and Department of
Trade and Industry (DTI).


4.2
Analytical framework and Estimation Techniques

The first step in our analysis is to test for stationarity of our variables. Gujarati (2003) suggest
that
a stationary stochastic process implies that the mean and variance are constant overtime,
and the covariance between two periods depends only on the lag between the two time
periods and not the actual time at which the covariance is computed. This implies
therefore
that a non
-
stationary time series will have a varying mean or varying variance or both.


The statistical and time series properties of the data set were first carried out using the
Augmented Dickey
-
Fuller (ADF) and Phillips
-
Perron (PP) to test f
or unit root. Mallik and
Choudhry (2001) and Ahmed and Mortaza (2005) point out that the PP test can properly
distinguish between stationary and non
-
stationary time series with high degree of
autocorrelation and presence of structural break.


Having determ
ined the order of integration of out variables, the next procedure would be to
test for the possibility of co
-
integration among the variables used. Cointegration is an
econometric technique for testing the correlation between non
-
stationary time series
var
iables. If two or more series are themselves non
-
stationary, but a linear combination of
them is stationary, then the series are said to be cointegrated.


Gujarati (2003:830) points out that the cointegration of two or more series suggests that there
is a
long run or equilibrium relationship between them. This means that even though the
series themselves may be non
-
stationary, they will move closely together over time and their
difference will be stationary. The long run relationship between the series is t
he equilibrium
to which the system joins over time and the disturbance term can be interpreted as the
disequilibrium error or the distance that the system is away from equilibrium at time t.


The Johansen technique is preferred to Engle Granger in this
study since it captures the
underlying properties of time series data and provides all co
-
integration relationships between
variables. Since our model is multivariate, there is a likelihood of having more than one co
-
integrating vector. According to Seddig
hi et al. (2000: 297) in the presence of more than one
co
-
integrating relationships, the Engle
-
Granger approach would produce inconsistent
estimates. Thus, in light of these problems, the Johansen methodology, developed in
Johansen (1991, 1995) is preferre
d as it allows hypothesis tests to be done using the vector
autoregressive co
-
integration based (VAR) tests.


In addition, a VAR is a useful method for analysing the impact of a given variable on itself
and on all other variables in the system by using for
ecast error variance decompositions
(FEVD) and impulse response functions. By breaking down the variance of the forecast error
for each variable into its components, FEVD are useful tool to analyse the impact of budget
deficits on private investment. IRFs
are also useful in tracing out the effects of one
-
time
shock to deficits and private investment.


5.

ECONOMETRIC PROCEDURE
,

RESULTS

AND CONCLUSION

Time series properties of the data were carefully evaluated through the Augmented Dickey
Fuller (ADF) and
Phillip
-
Peron (PP) te
sts. All variables
were regarded as non
-
stationary at
their levels. The variables were tested for stationarity at first differences. The results
indicated that all variables are stationary. The results confirmed therefore that differen
cing
once was all that was required to bring these variables to stationarity at all levels of
significance. This suggests that our variables are integrated of order one
I(1)
. Having
established the existence of unit roots, cointegration tests were conducte
d.


The optimal lag order was determined empirically. Based on several criteria (AIC, SIC, FPE,
LR and HQ), a lag order of 1, which produced a stable VECM, was selected.

The Johansen
test proved evidence of three cointegrating vectors. Based on the results

of cointegration, the
VECM was specified which provided the parameter estimates for the long
-
run relationship.
The main concern is to determine which variables have the greatest impact on private
investment demand and hence our discussion is mostly concer
ned with the influence of those
other variables on private investment rather than how all the variables influence each other.
We proceed to estimate a VECM normalized on private investment where we have the
private investment in the VEC as a function of th
e remaining variables. The long run
regression test results are given as follows:


PI = 50.17484 + 1.585597BD + 0.281064IR
-

6.290677PCY + 4.747208P

(0.93802)

(0.72552)

(1.43529)

(1.02707)

[1.69037]

[0.38739]

[
-
4.38287]

[4.62209]

Notes:

( )
standard error


[ ] t
-
statistic


The empirical findings, as reported by the positive coefficient of the variable BD in the
equation above, indicate that budget deficits have a negative long run relationship with
private investment. Real lending rate and
inflation also have a negative long run impact on
private investment demand whilst percentage change in real gross domestic product has
positive long run implications on private investment demand.


The empirical results indicate that there is a negative re
lationship between budget deficits and
private investment. This supports the theory that postulate that an increase in deficits reduces
private investment since higher deficits is seen as reflecting an increase in government
spending
signalling

higher gove
rnment debt. In this case the deficits are seen as deflationary
in the sense that they depress private investment hence economic growth. The debt financed
deficits crowd out interest sensitive private sector spending, in particular investment in homes
and
in new productive capacity.


Increase in government spending, given that government revenue is fixed, which brings about
a budget deficit, results to higher interest rates, thereby crowding out private investments.
The result of increase in interest rate c
omes from the neoclassical loanable funds theory. In
the theory, bond
-
financed government spending creates an insufficient fund for private
investment. Since it is assumed that the supply of fund is fixed in the theory, the competition
between government a
nd private sector for the available funds gives rise to higher interest
rates. As a result, (at least some of) private investors leave the loan market and hence
depresses private investment. This result is consonance with Gale and Orszag (2004) and
Anoruo
and Braha (2005) who discovered a negative relationship between budget deficits and
private investment.


Interest rates even though insignificant have a negative long run relationship with private
investment. This shows that an increase in interest rates r
esults in a decrease in private
investment as investment is interest sensitive. This confirms to theory and a
-
priori
expectation again as in the neoclassical theory of investment, increasing the interest rate
reduces investment by raising the cost of capit
al. This means that a fall in interest rates
should decrease the cost of investment relative to the potential yield and as a result, planned
capital investment projects on the margin may become worthwhile. A firm will only invest if
the discounted yield ex
ceeds the cost of the project. When interest rates are high, investment
becomes more expensive. The relationship between the two variables is represented by the
marginal efficiency of capital investment (MEC) curve. The IS/LM model also shows that
investme
nt is sensitive to interest rate changes. This is consistent with Bljer and Khan (1984)
and Greene and Villanueve (1991:33
-
58) who in their studies in developing countries
deduced a negative relationship between investment and interest rates. Fielding (199
7: 349
-
369), after concluding a study on the aggregate investment in South Africa, concluded that
investment is a negative function of the real interest rate.

The empirical results show that the inflation rate can have an adverse long run impact on
private

investment and the coefficient is highly significant. This is consistent with theory as
an increase in the inflation r
ate results in the central bank
increasing the repo rate to reduce
inflationary pressures thereby discouraging private investment. As a v
ariable also influenced
by government spending, it can be considered as an indicator of government efficiency. High
and unpredictable inflation also distorts the information content of relative prices and
increases the riskiness of longer time investment w
hich discourages potential investors. This
result corroborates the outcomes that Greene and Villanueve (1991:33
-
58) found that a higher
inflation rate had a negative impact on private investment for 23 developing countries in their
pooled time series/ cros
s sectional study. Similar studies by Abbas (2004) also confirmed the
negative relationship between the two variables.


Change in GDP has postulated a positive long run relationship with private investment and
the coefficient is highly significant. This sh
ows that an increase in GDP growth results in an
increase in private investment. This conforms to a
-
priori expectation. Khaled (1993) confirms
the positive relationship between private investment and GDP growth rate.


Impulse response and variance decompos
ition functions were also constructed to trace the
temporal and directional response of private investment to structural innovations in the
macroeconomic variables and as well as tracing the proportion of the movements in a
sequence of private investment t
o its own shocks versus shocks to other variables. It was
therefore established that innovations in BD, IR and P are negatively related to private
investment whilst innovations in PCY are positively related to private investment in both the
short and long
run. The most interesting result which emerged from this analysis and which is
supported by previous research is that among other variables, GDP growth explains the
largest proportion of the variation in private investment. On balance, the evidence therefo
re
suggests that private investment fluctuations in quarterly data are predominantly equilibrium
responses to real and monetary shocks rather than fiscal policy shocks.


CONCLUSION


The study investigated whether budget deficits crowd out private
investments in the South
African economy in the period 1994Q1
-
2009Q4.

Based on an extensive review of the
literature on the determinants of private investments, a background of the unprecedented
budget deficits in South Africa, an empirical model that link
s the budget deficits to private
investment was specified. All the variables have a long run relationship with private
investment. An increase in the budget deficit, real interest rate and inflation negatively impact
private investments in the long run, wh
ile change
s

in real GDP positively impacts on private
investment. The budget deficit coefficient was significant at 5% significant level revealing
that budget deficits had a significant effect on the South African economy. This means that if
an economy con
tinuously experiences fiscal deficits, private investment deteriorates, thus
being crowded out. It follows from these findings that private investment is largely a function
of both real and nominal variables in the long run. Both the theory and previous re
search are
contradictory and ambiguous as other researchers found that fiscal deficits caused by public
investment crowds in private investments and deficits caused by public consumption crowds
out private investment. These results therefore, for the most
part, agree with part of both the
theoretical predictions and findings from previous research which assert that fiscal deficits
caused by public consumption crowds out private investment.


The results from this study imply that that the monetary authoriti
es have the ability to
influence the changes in private investment. The authorities may however reduce the impact
of this shock, in the long run, by utilising stringent monetary policies to promote private
investment and acting on other fundamentals. More
generally, coordination of monetary and
fiscal policies will be important for an effective policy response to the crowding out effect as
rising interest rates may also add to spending pressures. Also low inflation and sustainable
debts contribute to strong

economic performance.




REFERENCES
:


Abbas, V. 2004. “What determines private investments in Iran?”
International Journal of
Social Economics
, Vol.
31.


Abrams, B.A. & Schmitz, M.D. 1978.

“The 'crowding out' effect of government transfers on
private charitable contributions”.
Public Choice

33

(1): 29
-
39.


Ahmed, S., and Mortaza, G. 2005. Inflation and Economic Growth in Bangladesh:1981
-

2005.
Policy Analysis Unit (PAU) Working Paper
0604


Akinboade, O.A. 2004. “The relationship between budget deficit and interest rates in South
Africa: some econometric results”.
Development Southern Africa,

21
(2): June
.


Anoruo, E. & Braha, H. 2005. “Corruption and Economic Growth: The African Experience”.
Journal of Sustainable Development in Africa
, Vol.
7
(1): 43
-
55.


Arestis, P. 1979. “The 'crowding out' of private expenditure by fiscal actions: An empirical
investigation”.
Public Finance/Finances Publiques
, No.
1.


Ball, L., Elmendorf, D.W. & Mankiw, N.G
. 1995. “The Deficit Gamble”,
NBER Working
Paper 5015
.


Blanchard, O & Perotti, R (2002) “ An Empirical characterization of the Dynamic effects of
Changes in Government Spending and Taxes on Output” Quarterly Journal of Economics,
November 2002, Vol.117, N
o.4, Pages 1329


1368.


Blejer, M.I. & Khan, M.S. 1984. “Government policy and private investment in developing
countries”.
IMF Staff Papers
,
31
(2).


Brunner, K. 1984. “Deficits, interest rates and monetary policy”. Shadow open market
committee policy statement and position papers (March). Cited in McMillin, W.D. 1986.
“Federal deficits and short
-
term interest rates”.
Journal of Macroeconomics
,
8
(4): 40
3

22.


Cebula, R.J., Carlos, C. & Koch, J.V. 1981. “The 'Crowding out' Effect of Federal
Government Outlay Decisions: An Empirical Note”.
Public Choice
, 36(2): 329
-
336,
Springer.


Chowdhury, A. & Hossain, A. 1998.

Open Economy Macroeconomics for Developing
Countries. Edwarel Elgar Press. Northampton, UK

Darrat, A.F. 1989. “Fiscal Deficits and the Long
-
Term Interest Rates: Further Evidence from
Annual Data”. Southern Economic Journal, Vol. 56(2): 363
-
374.

DeLeeuw, F

& Holloway, T. 1985 "The Measurement and Significance of Cyclically
Adjusted Federal Budget and Debt".
Journal of Money, Credit and Banking
, Vol.
17

(2): 323
-

42.


Dewald, W.G. 1983. "Federal Deficits and Real Interest Rates: Theory and evidence",
Federal
Reserve Bank of Atlanta Quarterly Review,

Vol.
68
: 20
-
29.


Easterly, W. & Rebelo, S.

1993. “Fiscal policy and economic growth: An empirical
investigation.”
Journal of Monetary Economics
,
32
: 417
-
458.


Easterly, W. & Rebelo, S. 1993. “Fiscal policy and econ
omic growth: An empirical
investigation.” Journal of Monetary Economics, 32: 417
-
458.

Easterly, W.R, Rodríguez, C. A. & Schmidt
-
Hebbel, K. 1994. Public Sector Deficits and
Macroeconomic Performance
,
Washington, D.C: Oxford University Press.


Evans, P. 1985
.

Do large Deficits Produce High Interest Rates
?”

American Economic
Review
,
75(1)
:68
-
87


Fielding, D. 1997. "
Aggregate Investment in South Africa: A Model with Implications for
Political Reform
,"
Oxford Bulletin of Economics and Statistics
, Department of Economics,
University of Oxford, vol.
59
(3): 349
-
69, August


Freund, B. & Padayachee, V. 1998. “Post
-
Apartheid South Afri
ca: The Key Patterns
Emerge”.
Economic and Political Weekly
, Vol.
33
(20): 1173
-
1180.


Gale, W.G. & Orszag, P.R. 2004. “Economic Effects of Sustained Fiscal Deficits.”
National
Tax Journal
56
: 463

485.


Greene, J. & Villanueva, D. 1991. “Private
investment in developing countries”.
IMF Staff
Papers,
38
(1): 3358.


Gujarati, D. N. 2003. Basic Econometrics (4e). New York: McGraw
-
Hill Inc.


Haan, J. D. & Zelhorst, D. 1990. “The Impact of Government Deficits on Money Growth in
Developing Countries”.
Jo
urnal of International Money and Finance
9
(4).


Hernandez
-
Cata, E. 2000. “Raising Growth and Investment in Sub
-

Saharan Africa: What
Can be done?”
Policy Discussion Paper:

PDP/
00/4
,
International Monetary Fund,
Washington, D .C.


Hoelscher, G. 1983.
"Federal Borrowing and Short
-
Term Interest Rates".
Southern Economic
Journal
, Vol.
50
(2): 319
-
33.

Ibid.

Khaled, S. 1993. “Determinants of private investment in Pakistan”.
IMF working Paper

30

Kormendi, R.C. & Mcguire P.G. 1985.

“Macroeconomic Determinants of Growth: Cross
-
Country Evidence”.
Journal of Monetary Economics
, Vol.

16
: 141
-
163.

Makin, J. H. 1983. “Real interest rate, money surprises, anticipated inflation and fiscal
deficits”.
Review of Economics and Statistics
,
65:

374

84, August

Mallik, G. and Chowdhury
,
A. (2001). Inflation and Economic Growth: Evidence from South
Asian Countries,”
Asian Pacific Development Journal,
Vol.
8,

No.1. pp. 123
-
135.


Marcus, G. 2003. “The Implications of Fiscal Issues for Central Banks: T
he South African
Experience”. BIS Papers No. 20: 230
-
233.

Mlambo, K. & Oshikoya, T.W. 1999. “Macroeconomic Factors and Investment in Africa”,
Journal of African Economies
, Vol.
10
, Supplement 2, pp 12
-
47.

Motley, B. 1983. "Real Interest Rates, Money and
Government Deficits".
Federal Reserve
Bank of San Francisco Quarterly Review
, Vol.
5
(2): 31
-
45.

Oshikoya, T. W. 1994 “Macroeconomic Determinants of Domestic Private Investment in
Africa”
Journal of Economic Development and Cultural Change
,
42
(3): 573
-
596.

South African Reserve Bank (SARB). 1994.

Quarterly Bulletin,

Pretoria: SARB.

South African Reserve Bank (SARB). 1996.
Quarterly Bulletin,

Pretoria: SARB.

South African Reserve Bank (SARB).2010. “Annual report”
,

Pretoria: SARB

Tanzi, V. 1985.

“Fiscal deficits and interest rates in the United States: An empirical analysis
1960

84”.
IMF Staff Papers
,
32
(4): 551

76.

Yan Kuo, C, Jenkins, G.P & Mbenjamin, M. 2003. “The Economic Opportunity Cost of
Capital for South Africa”
The South African Journal

of Economics Die Suid
-
Afrikaanse
Tydskrif vir Ekonomie

Vol.
71
(3), September.

Zahn, F. 1978. “A flow of funds analysis of crowding out”.
Southern Economic Journal
, July.