validity of twin deficits hypothesis in pakistan (1980-2007)
DESCRIPTION
This is twin deficit case study of Pakistan for a specific time period during M.Sc Eco(2007-09) in QAU Islamabad,Pakistan. by SYED MUHAMMAD ABDUL REHMAN SHAH. Student MS leading to Ph.D IBF IIUI.TRANSCRIPT
(Mini Thesis)
Validity of Twin Deficits Hypothesis in Pakistan (1980-2007)
SUBMITTED TO;Dr. EATZAZ AHMED
BY SYED MUHAMMAD ABDUL REHMAN SHAH
M.Sc Eco (2007-09)
DEPARTMENT OF ECONOMICS QUAID-E-AZAM UNIVERSITY ISLAMABAD PAKISTAN
With the Name of ALLAH
Contents
Section 1: Introduction
Section 2: Methodology
Section 3: literature Review
Section4: Data, Estimation and our Findings
Section 5: Conclusion
*Figures: Graphical Results of Impulse Response (VAR)
References
Validity of Twin Deficits Hypothesis
in Pakistan (1980-2007)
Introduction: Introduction:
Excessively there is Budget Deficit and as well as Trade Deficit in
Pakistan. In long run these deficits are root cause to each of the ills of our
small economy and it is also found in different studies that unidirectional
relationship from budget deficit to trade deficit exists as in the case of
Thailand (see Baharumshah 2004). Baharumshah also found that
due to budget deficit there exists high interest rate and it leads to the
appreciation of currency rate that is a cause of trade deficit. So in such
channel we can explain the hypothesis of the Twin Deficit in the case of
Pakistan. Not only LDCs,but DCS also faced this severe problem that create
instability in these economies.
The aim of this paper is to investigate this Twin Deficit Hypothesis in the
case of Pakistan by using the data set of (1980-2007).
Section 2 there is model of the theoretical frame work of National
Accounting to explain twin deficits
Section 3 there reviewed literature related to the topic.
Section 4 Annually Data is used because it is believed that effect of one
deficit on the other is more significant in yearly data than in quarterly data.
( Kulkarni, )
Hypothesis of Twin Deficit is tested and results are brought through
different techniques.
Section 5 Concluding remarks of the topic are made briefly.
2:Model:Fiscal and Trade Balance in National Accounts
In the following National Account Identity Twin Deficit hypothesis can be
explained in a good way that is used by different articles ( Aqeel and Nishat
2000 and Baharumshah,Lau & Khalid 2004 ).It can be expressed as
Y= C + I + G + X – M (1)
where Y= gross domestic product (GDP), C = consumption, I
= investment G =government expenditure, X = export and
M = import. Current account (CA) can be defined as the
difference between export (X) and import (M), and
rearranging the variables equation(1) becomes:
CA = Y – (C + I + G) (2)
Where (C + I + G) are the spending of domestic residents
(domestic absorption). In a closed economy savings (S)
equals investment (I) and given that Y – C = S, we have:
S = I + CA (3)
In equation 3 it can be seen that an open economy can
provide sources to invest domestically and internationally
to increase Income. There government can invest more than
domestic savings by borrowing from abroad.
National savings can be divided into private (Sp) and
government savings (Sg)
Sp = Y – T – C (4)
Sg = T – G (5)
there T is the taxes imposed by government ,Using
equations 4 and 5 and substituting into equation (3)yield:
Sp = I + CA + (G-T) (6)
Now by rearranging this equation (6) there the relationship between budget
deficit and current account deficit can be expressed in equation (7)
CA = S(Private)–I – (G – T) (7)
There, CA is for current account balance, S is for private saving ; I for
investment, G is for government purchases; and T for direct taxes collected
from Household firms by the government. The government deficit is expressed
by (G–T).
A rise in the government deficit will increase the current account
deficit if due to budget deficit there occur decrease in national savings.
Now if saving are not increased or increased but offsetted by increasing
Investment and in such a way S-I remain same, then as a result there occurs
positive relation between CA & BD.In this way a government deficit
resulting from increase in purchases reduces the nations current account
surplus. An increase in current account deficit due to budget deficit is the
one aspect of the twin deficit.
In another way we can present that due to budget deficit there
exists high interest rate and it leads to the demand for home currency rise
and results the appreciation of currency rate that is a cause of trade deficit
because cheaper import and more expensive exports are pushing the trade
balance towards deficit. So in such channel we can explain the hypothesis of
the Twin Deficit.
Ricardians and Keynesians have different views about the effects of
budget deficit due to tax cut. Ricardians believe that due to tax cut an
individual does not feel better off because he expects future tax burden imposed
by government to retire the debt which was get to finance the tax cut. So
according to them there no change occurs in major economic variables like
national savings, current account balance, consumption, interest rates and
investment. Keynesian say that a tax cut budget deficit policy affects the
current consumption pattern of an individual who expects high future tax
burden and now increasing consumption. This will reduce national savings,
increase current account deficit and also will affect all macro linkages between
them. Ultimately this leads to twin deficits.
Moreover there is another channel between budget deficit and current
account deficit. As budget deficit increases; government’s borrowing also
increases & the rate of interest increases, that becomes a cause to foreign
capital inflow. This will appreciate the value of the domestic currency that
results in cheaper imports and expensive exports. So there Trade deficit occurs
that begins from government budget deficit thus Twin deficit occurs there.
3: Literature Review:
About Twin Deficit Hypothesis there are different articles and research
papers that are focused mainly by two different point of views (1)
Keynesians (2) Classical (see Baharumshah 2004).
1st one says that due tax increase or tax cut there is clearly change
in consumptions and savings decisions which lead to the change in current
accounts. In the case of fiscal budget deficit financed by foreign borrowing
leads to an increase in interest rate that leads to foreign capital inflow which
comes to be a cause to currency appreciation now there imports are
relatively cheaper and exports are relatively expensive, in this way current
account deficit exists. as (BD_IR_EX_ CAD) (see Volcker, 1984 and
Abell, 1990)
2nd one says that due to Recardian Equellance Hypothesis there is no
change in major all major economic variables mentioned in 1st one so there
is not any kind of relationship between budget deficit and (Aqeel and Nishat
2000 and Baharumshah,Lau & Khalid 2004).
Some of the articles brought the idea of reverse causality in which
trade deficit came to be the cause of budget deficit as it is proved in the case
of Pakistan during time period 1969 to 1997 (Kulkarni and Erickson). In
some countries there is not found even unidirectional causality in the study
of Maxico during time period 1969 to 1997 (Kulkarni and Erickson).
There are often used Granger Causality Tests and VAR technique
to investigate the twin deficit hypothesis and its direction. At stationary data
ADF,Co-Integration, Granger Causality Tests are applied to check the
results(Mukhtar,Zakiria &Ahmad 2007)
The theoretical framework of National Income Identity is used to
have a basic and deep understanding of the topic that can be easily
considered both deficits in such a way.
(4)Data, Estimation and our Findings:
In this article the annual data set of the period (1980 to 2007) is used to
investigate the validity of twin deficit hypothesis in Pakistan. There used four
variables:
(1) BD: Budget Deficit as ratio of GDP
(2)TD: Trade Deficit as the ratio of GDP
(3)ER: Real Effective Exchange Rate
(4)IR: Interest Rate
Unit Root Tests Overall, we found that the variables contain the unit root
and these are Stationary at 1st difference by using ADF.
Impulse Response in VAR Model
VAR model is used to check the impulse responsiveness of
variables among each other especially to check the Twin
Deficit. In this test we shall try to investigate the
responsiveness of BD & TD between each other, there are
different possibilities by which we can test theory. After
estimation of the data different results are found in following
ways:
►(a) In (Figure : 1) there is 1st specification in which only
BD&TD are considered in a control atmosphere IR&ER are
not considered. There is significant result of responsiveness
between BD & TD. These empirical results tells us about that
our data sport the conventional view of that BD implies TD in
long run .Although in short , we can see that as BD increases
then TD till 2 lag remain same, then it decreases slightly up
to 5 lag then increases slightly with BD as our theory
suggests that there is TD due to BD after 5th lag up to end so
there is found the affect of BD on TD in long run.
In (Figure: 2) as there is 1st specification in which only
BD&TD are considered in a control atmosphere where IR&ER
are not considered. There is not more significant result of
responsiveness between BD & TD. In the beginning we found
that there is increasing trend of BD to TD till 4th lag then BD
decreases slightly inverse of TD perhaps some other factors
are affecting it.
► (b) There is used 2nd specification in which all variables are
considered in a open atmosphere where all four variables
are shown affecting each other in (Figure: 3,4,5 &6) in
columns each figure have four variables.
¤In (Figure: 3)there is shown responsiveness of all
four variables to BD increase in separate four graphs, so
responsiveness of CA to BD in open atmosphere is almost
same as that was in controlled atmosphere in figure:1,
having positively movement in long run although now the
movement of TB is more flatter than in 1st figure. There is
also considered the responsiveness of IR to BD,which is
insignificant overall because we can see that at 2nd lag
curve is at its lowest value then IR increases immediately as
our model suggests up to 3rd lag then there is slightly
decrease in IR up to end. This trend is against the theory of
about an increase in IR due to an increase in BD.So we can
say that overall this is a very poor result.
In the case the responsiveness of ER to BD,1st there is some
decrease in ER in 1st two lags, then in next remaining lags ER
moves increasing slightly with BD as model suggests their
positive movement.
◘Note: All Figures are attached at the end of the thesis.
¤ Figure: 4 & 5 show the responsiveness off all four
variables to IR and ER respectively, although all responses
have their own importance as channels of different variables
are made,,,as concern of our specified model is only with affect
of IR on ER in Figure :4, and that of ER on TB in Figure :5 . This
is proposed to construct the bridge between BD and TD, as the
complete channel explained in the model is
BD→IR→ER→TB
as following channels can be seen graphically one by one in
Figures (BD→IR 2nd in Figure:3, IR→ ER 3rd in Figure:4
and ER →TB 4th in Figure:5)
¤In Figure: 6 there is presented graphically in VAR
Model’s Impulse responsiveness of BD, IR, RE,TB to the change
in TB. Our main concern is to observe there TB→BD, if it is
found significant then there may occur bidirectional Twin
Deficit but in our findings there is not any significant trend by
using which we justify the affects of TB on BD.
Conclusion
In the case of Pakistan for the data of time
period of 1980 to 2007, by using Impulse responsiveness of
VAR model Twin Deficit Hypothesis almost satisfies where
Budget Deficit implies Trade Deficit indirectly through the
channel of Interest Rate and Exchange Rate. Bidirectional
between BD and TB or Unidirectional from trade deficit to
budget deficit are also not found. Although estimated results
are not as significant as these should be, so all of the errors in
this concise empirical study are due to my knowledge
constraints and data problems, so considering a rider at the
initial stage in the field of research, kindly guide me.
REFERENCES
Baharumshah (2004) Testing Twin Deficits Hypothesis: Using VARs and Variance Decomposition, Journal of Economic Literature 31: 142-190.
Mukhtar Tahir, Zakria and Mehboob (2007) An Empirical Investigation for the Twin Deficit hypothesis in Pakistan, Journal of Economic cooperation 28, 4 (2007) 63,80
Abell (1990) ‘Twin Deficits during the 1980s: An Empirical Investigation’, Journal ofMacroeconomics 12: 81-96.
AQEEL Anjam and MOHAMMED NISHAT (2000) The Twin Deficits Phenomenon: Evidence from Pakistan The Pakistan Development Review 39: 4 Part II (Winter 2000) pp. 535–550
Kulkarni, Kishore G and Erick Lee Erickson , Twin Deficit Revisited: Evidence From India, Pakistan And Mexico”, The Journal of Applied Business Research Volume 17, Number 2 p:96
Bartolini, Leonardo and Amartya Lahiri(2006), Twin Deficits, Twenty Years Later, FEDERAL RESERVE BANK OF NEW YORK, Volume 12, Number 7 October 2006
Figure:1 Response of BD&TD to BD in VAR model
By using VAR model the impulse response of variables in each other there is investigated affect of the change in TD due to change in BD in controlled atmosphere for the time period 1980-2007 in Pakistan.
-.15
-.10
-.05
.00
.05
.10
.15
.20
1 2 3 4 5 6 7 8 9 10
R e s p o ns e o f L B D to L B D
-.3
-.2
-.1
.0
.1
.2
1 2 3 4 5 6 7 8 9 10
R e s p o n s e o f L T D to L B D
Response to Cholesky One S.D. Innovations ± 2 S.E.
Figure:2
By using VAR model the impulse response of variables in each other there is investigated affect of the change in BD due to
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LBD to LTD
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
.5
1 2 3 4 5 6 7 8 9 10
Response of LTD to LTD
Response to Cholesky One S.D. Innovations ± 2 S.E.
change in TD in controlled atmosphere for the time period 1980-2007 in Pakistan.
Figure:3 Figure:4 Figure:5 Figure:6
-.3
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LBD to LBD
-.3
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LBD to LIR
-.3
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LBD to LER
-.3
-.2
-.1
.0
.1
.2
.3
1 2 3 4 5 6 7 8 9 10
Response of LBD to LTD
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LIR to LBD
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LIR to LIR
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LIR to LER
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LIR to LTD
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LER to LBD
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LER to LIR
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LER to LER
-.08
-.04
.00
.04
.08
.12
1 2 3 4 5 6 7 8 9 10
Response of LER to LTD
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LTD to LBD
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LTD to LIR
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LTD to LER
-.4
-.3
-.2
-.1
.0
.1
.2
.3
.4
1 2 3 4 5 6 7 8 9 10
Response of LTD to LTD
Response to Cholesky One S.D. Innovations ± 2 S.E.
Real effective exchange
Overall Deficit As % of
IR
rate index (2000 = 100) GDPER BD IR TD
1980 188.6825 6.3 8.971981 213.7525 5.3 8.61 8.71982 195.7975 5.3 9.86 10.31983 189.1025 7 8.69 9.31984 193.1117 6 8.1 9.41985 180.3475 7.8 9.13 111986 148.8275 8.1 7.26 81987 131.6008 8.2 6.26 5.11988 129.5558 8.4 6.27 51989 121.2667 7.4 6.34 5.91990 117.3733 6.5 6.77 4.91991 116.0483 8.7 7.12 3.31992 114.3933 7.4 7.36 4.81993 113.9958 8 9.81 6.11994 111.4392 5.9 9.18 3.41995 110.7217 5.6 10.33 3.71996 107.2792 6.5 11.16 4.91997 108.7875 6.4 12.97 5.71998 106.7425 7.7 12.23 2.41999 99.4875 6.1 7.84 2.82000 100.0008 5.4 8.52 2.42001 91.4775 4.3 8.96 2.12002 94.77833 4.3 6.74 1.72003 91.78333 3.7 4.23 1.32004 91.11167 2.4 1.86 3.32005 94.04917 3.3 4.34 5.5
2006 4.3 6.83 9.52007 4.3 8.89 9.4