Economic Modelling 33 (2013) 940946

Contents lists available at SciVerse ScienceDirect

Economic Modelling
journal homepage: www.elsevier.com/locate/ecmod

Effects of terms of trade on growth performance of India

Syed Tehseen Jawaid 1, Syed Ali Raza
IQRA University, Karachi-75300, Pakistan

a r t i c l e

i n f o

Article history:
Accepted 22 April 2013
JEL classication:
F13
F43
Keywords:
Terms of trade
Economic growth

a b s t r a c t
This study investigates the effect of terms of trade on economic growth of India by using the annual time series
data from the period of 1980 to 2010. The ARDL bound testing cointegration conrms the signicant positive relationship between terms of trade and economic growth in the long run as well as in the short run. Results of
Granger causality, Toda and Yamamoto Modied Wald causality and variance decomposition tests conrm the
bidirectional causal relationship between terms of trade and economic growth in India. Rolling window estimation indicates that the terms of trade is having positive long-run coefcients throughout the sample period. Results of CUSUM and CUSUM of square suggest that there is no structural instability in the residuals of equation of
economic growth in short run. It is suggested that benecial terms of trade is better for economic growth in India.
At this stage we can set the direction of future research that the relationship between commodity groupwise and
countrywise bilateral terms of trade with economic growth should be analyzed. These results would be helpful
for policy makers of India to frame growth enhancing countrywise as well as commodity-wise trade policies.
2013 Elsevier B.V. All rights reserved.

1. Introduction
In India during the last three decades, trend shows that terms of
trade has improved. In the 1980's the average terms of trade was 84,
in the 1990's it increased to 105 and in the decade of 2000 the average terms of trade marginally improved and became 107. Similarly,
in the 1980's the average growth in real GDP was 6%, in 1990's it
again sustained at 6% and in the decade of 2000's in increased to
7.3%. The question is that, is the commodity terms of trade correlated with economic growth? This study examines this question by
using long time series annual data of India covering the period
from 1980 to 2010.
Most of the empirical studies have been conducted under
Prebisch-Singer (PS) hypothesis. 2 Perbish-Singer hypothesis 3 argues
that the terms of trade of primary product specialization countries
will weaken over time as compared to the countries that specialize
in manufactured goods. Declining of terms of trade is one of the
main reason of income gap between developed and developing countries. Increase in terms of trade would lead to increase in investment
and thus economic growth will increase.

Corresponding author. Tel.: +92 333 344 8467.

E-mail addresses: stjawaid@hotmail.com (S.T. Jawaid), syed_aliraza@hotmail.com
(S.A. Raza).
1
Tel.: +92 345 309 4838.
2
Lutz (1999), Hadass and Williamson (2001) and Cashin and McDermott (2002a,b).
3
See Perbisch (1950) and Singer (1950).
0264-9993/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.econmod.2013.04.043

Many studies have been conducted to nd HerzbergerLaursen

Metzler (HLM) effect. 4 HLM effect 5 argued that the declining in
terms of trade will lead to reduce the real income and lower income
will lead to lower savings and investment. Consequently, it affects
the current account. Jawaid and Waheed (2011) explain the channel
between terms of trade and economic growth. According to them, increase in terms of trade results in efcient allocation of resources
which leads to high productivity and economic growth. Higher economic growth makes ground for a country to move their resources
to research and development and this leads to quality improvement
in the country. Consequently, export prices increase resulting to further improvement in terms of trade. In most of the empirical studies
cross country data 6 has been used to analyze the relationship between terms of trade and economic growth, India is mostly not included in these cross country studies. However, some time series
are also done on the same subject. 7 The objective of this study is to
examine the effect of terms of trade on economic growth of India.
The rest of the paper is organized as follows: following introduction Section 2 reviews some selected studies, Section 3 discusses empirical strategy, Section 4 shows estimations and results, Section 5
shows results of rolling window estimation, Section 6 discusses the

4
Arize (1996), Otto (2003), Bouakez and Kano (2008), Hamori (2008) and Misztal
(2010).
5
See Harberger (1950) and Laursen and Metzler (1950).
6
Sea Bleaney and Greenaway (2001) and Cashin and Mecdermott (2002a,b).
7
See Wong (2004, 2010) and Fatima (2010).

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

results of cumulative sum and cumulative sum of square estimations,

Section 7 shows the results of causal relationship between terms of
trade and economic growth and the nal section concludes the
study and provides some policy implications.
2. Review of literature
Many studies suggest the positive effect of terms of trade and negative effect of volatility of terms of trade on economic growth. In this
section some selected studies are discussed.
Arize (1996) uses the cointegration technique to empirically examine the long-run impact of terms of trade on trade balance by
using the data of 16 countries from the period 1973 to 2004. The results suggest the positive relationship between terms of trade and
trade balance in most of the countries.
Mendoza (1997) uses the panel estimation method on the data of
40 industrial and developing countries from the period 1971 to 1991
to empirically examine the endogenous growth model. The ndings
suggest the positive impact of rate of change of terms of trade on economic growth. The negative relationship is found between volatility
of terms of trade and economic growth. Sensitivity analysis conrms
the robustness of the results.
Kaneko (2000) uses endogenous growth model with two factors,
physical and human capital to investigate the relationship between specialization pattern and growth rate of growing economy. Results suggest the positive and signicant relationship between terms of trade
and economic growth in a country that specializes in consumption commodities. Furthermore, if a country specialized in capital commodities,
the economic growth is not affected by the terms of trade.
Bleaney and Greenaway (2001) use stochastic endogenous growth
model to empirically examine the impact of terms of trade, exchange
rate and their volatilities on growth and investment. They use panel
estimations on the data of 14 Sub-Saharan African countries from
the period 1980 to 1995. Volatility of terms of trade and real exchange
rate is estimated by using Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) model. Results show that improvement
in terms of trade and less over value exchange rate have signicant
positive effect on growth and investment while, signicant negative
relationship is found between volatility of terms of trade and economic growth.
Hadass and Williamson (2001) use the data of 19 countries to empirically investigate the relationship between terms of trade and economic growth from the period of 1870 to 1940. The ndings indicate
that the positive movement in terms of trade reduces economic
growth of primary product exporters. They concluded that variation
in terms of trade explain not more than one fth of economic growth
in prewar period.
Cashin and Mcdermott (2002b) use the different quarterly time
series data of ve OECD countries to analyze the relationship between
current account balance and terms of trade shocks. 8 They used Structural vector autoregression (SVAR) model. The results of Canada,
United Kingdom and United States show only a small share of volatility in current account balance by the shocks of terms of trade. On the
other hand in Australia and New Zealand terms of trade shocks are
found to have signicant proportion of variation in current account
balance.
Wong (2004) uses the cointegration and error correction technique
to analyze the long-run and short-run relationship between terms of
trade and economic growth in Malaysia. Annual time series data has
been used from the period 1965 to 2002. The results of cointegration
conrms the signicant positive long-run relationship between terms
of trade and economic growth. The results of error correction model
8
For Canada (1970:21997:4); for Australia from 1970:21997:2; for New Zealand
(1980:21997:2); for the United Kingdom (1970:21997:4); and for the United States
(1973:21997:4).

941

also conrm the positive and signicant relationship between terms

of trade and economic growth of Malaysia in the short run.
Cakir (2009) empirically examines the relationship between
terms of trade and economic growth by using the panel data of 18
emerging economies over the period of 1990 to 2004. Generalized
methods of moments (GMM) has been used. Results indicate the signicant positive relationship between terms of trade and economic
growth.
Wong (2010) uses the annual time series data of Japan and Korea
from 1996 to 2003 and 1971 to 2006 respectively to empirically examine the relationship between terms of trade and economic growth.
To nd the long-run relationship Johansen cointegration technique
has been used. The results suggest that the real GDP per capita and
terms of trade are mutually determined. Results also indicate the signicant negative relationship between volatility of terms of trade and
GDP per capita in both countries.
Jawaid and Waheed (2011) investigate the impact of terms of
trade and its volatility on economic growth by using cross country
data of 94 countries over the period 2004 to 2008. Results indicate
the signicant positive impact of terms of trade and its volatility on
economic growth. Sensitivity analyses have been used to check the
robustness of initial results. The results were found robust despite
the inclusion of additional variables in basic model and use of various
proxies for volatility of terms of trade.
3. Empirical framework
After reviewing the theoretical and empirical work, the model to
examine the impact of terms of trade on economic growth is derived
using the production function framework. The production function in
general form as follows:
Y f A; L; K

3:1

Where Y is the real gross domestic product, L is the labor force, K is

the capital stock and A is the total factor productivity. It has been assumed that effect of terms of trade on economic growth operates
through A. 9
A g TOT

3:2

Substituting Eq. (3.2) in Eq. (3.1)

Y f L; K; TOT

3:3

The empirical models for estimations are developed as follows:

GDP t 0 1 LABt 2 CAP t 3 TOT t t

3:4

Where, t is the error term, GDP is the real gross domestic product,
LAB is the total labor force and TOT represents the terms of trade. Real
gross xed capital formation as percentage of GDP has been used as a
proxy for capital stock because of unavailability of data of capital
stock. 10 The expected signs for labor and capital stock are positive
while, the sign of TOT is to be determined. Annual time series data
have been used from 1980 to 2010. All data are gathered from
World Bank's ofcial database 11 and different issues of economic surveys of India. All variables are used in logarithm form.
3.1. Unit root test
Augmented Dickey Fuller (ADF) and Phillip Perron (PP) unit root
test are used to examine the stationary properties for long-run
9
10
11

See, Kohpaiboon (2003) Jawaid and Raza (2012).

See, Balasubramanyam et al. (1996), Barro (1999) and Kohpaiboon (2003).
The web link of data source is http://data.worldbank.org/indicator.

942

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

relationship of time series variables. Augmented Dickey Fuller (ADF) 12

test is based on equation given below:
k

Y t 0 1 Y t1 dj Y t j t

below the lower critical bound. If long-run relationship between

terms of trade and economic growth is found then we estimate the
long-run coefcients. The following model will be used to estimate
the long-run coefcients:

j1

Where t is pure white noise error term, is rst difference operator,

Yt is a time series, 0 is the constant and k is the optimum numbers of
lags of the dependent variable. Augmented Dickey Fuller (ADF) test determines whether the estimates of coefcients are equal to zero. ADF
test provides cumulative distribution of ADF statistics. The variable is
said to be stationary, if the value of the coefcient is less than the critical values from the Fuller table. Phillip and Perron (PP)13 unit root test
equation is given below:

i1

i1

i1

i1

If we nd evidence of long-run relationship between terms of

trade and economic growth then we estimate the short-run coefcients by employing the following model:
p

i1

i1

i1

GDPt 0 1 GDPt1 2 LABt1 3 CAPt1

p

4 TOTt1 nECt1 t

Y t Y t1 t

i1

The Phillip and Perron unit root test is also based on t-statistics
that is associated with estimated coefcients of *.
3.2. ARDL bound testing approach
The Auto Regressive Distributed Lag (ARDL) method of cointegration developed by Pesaran and Pesaran (1997), Pesaran and Shin
(1999), Pesaran et al. (2000, 2001) has been used with the help of
unrestricted vector error correction model to investigate the long-run
relationship between terms of trade and economic growth. The ARDL
approach has several advantages upon other cointegration methods.
ARDL approach may apply irrespective of whether the underlying
variables are purely I(0), I(1) or mutually co-integrated. 14 ARDL approach has estimated better small sample properties. 15 In ARDL procedure the estimation of results is even possible if the explanatory
variables are endogenous. 16 The ARDL model is developed for estimations as follow:
p

i1

i1

i1

GDPt 0 1 GDPt1 2 LABt1 3 CAPt1

p

GDPt 0 1 GDPt1 2 LABt1 3 CAPt1 4 TOTt1 t

4 TOTt1 1 GDPt1 2 LABt1 3 CAPt1

i1

4 TOTt1 t
Where 0 is the constant and t is the white noise error term, the
error correction dynamics is denoted by summation sign while the
second part of the equation corresponds to long-run relationship.
Schwarz Bayesian Criteria (SBC) has been used to identify the optimum
lag of model and each series. 17 In ARDL model we rst estimate the
F-statistics value by using the appropriate ARDL models. Secondly, the
Wald (F-statistics) test is used to investigate the long-run relationship
among the series. The null hypothesis of the cointegration is (H0 =
1 = 2 = 3 = 4 = 0). The null hypothesis of no cointegration is
rejected if the calculated F-test statistics exceeds the upper critical
bound (UCB) value. The results are said to be inconclusive if the F-test
statistics falls between the upper and lower critical bound. Lastly, the
null hypothesis of no cointegration is accepted if the F-statistics is

The error correction model shows the speed of adjustment needed

to restore the long-run equilibrium following a short-run shock. The n
is the coefcient of error correction term in the model that indicates
the speed of adjustment.
4. Estimations and results
To check the stationary properties we use Augmented Dickey Fuller (ADF) and Phillip Perron (PP) unit root tests. Table 4.1 represents
the results of stationary tests. First, these tests are applied on level
of variables then on their rst difference.
Results of Table 4.1 show that all variables are stationary and integrated at rst difference. This implies that the series of variables may
exhibit a valid long-run relationship.
Autoregressive distributed lag method for cointegration is used to
estimate the long-run relationship between terms of trade and economic growth. The rst step is to determine the optimal lag length
of the variables. The order of optimal lag length is decided by using
the Schwarz Bayesian Criterion. Table 4.2 shows the results of ARDL
cointegration method.
The ARDL results suggest the rejection of null hypothesis of no
cointegration in model because the value of the F-statistics is greater
than upper bound critical value at 1% level of signicance in favor of
alternative hypothesis that the valid long-run relationship is existing
between terms of trade and economic growth in India.
Now we estimate the lag length order of all variables through
unrestricted vector autoregression method. The decision criterion is
based on minimum value of Schwarz Bayesian Criterion.
Table 4.3 represents the results of lag length order of all variables.
Results of Schwarz Bayesian Criterion indicate that the gross domestic
product, labor and terms of trade should be included in the model at
1st lag while capital should be include in the model at 2nd lag. After
having the valid evidence of long-run relationship between terms of
trade and economic growth now we applied the ARDL method to estimate the long-run and short-run coefcients. The model for
long-run coefcients is as follow:
p

i1

GDPt 0 1 GDPt1 2 LABt 3 LABt1 4 CAPt

p

i1

i1
p

i1

i1

i1

i1

i1

5 CAPt1 6 CAPt2 7 TOTt 8 TOTt1 t

12

See, Dicky and Fuller (1979).

13
See, Phillips and Perron (1988).
14
Pesaran and Shin (1999).
15
Haug (2002).
16
Pesaran and Shin (1999) and Pesaran et al. (2001).
17
The SBC is slightly superior to the AIC (Pesaran and Shin, 1999). Besides, SBC is parsimonious as it uses minimum acceptable lag while selecting the lag length and avoid
unnecessary loss of degrees of freedom. Therefore, SBC criterion has been used, as a criterion for the optimal lag selection, in all cointegration estimations.

Table 4.4 shows the results of long-run ARDL estimations. The results of (LAB) and (CAP) are having expected positive sign and are
highly signicant. Results indicate the positive and signicant effect
of terms of trade on economic growth in India. The ndings are consistent with Kaneko (2000), Cakir (2009) and Jawaid and Waheed

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

Table 4.1
Stationary test results.
Variables

GDP
LAB
CAP
TOT

943

Table 4.4
Long-run results using ARDL approach.

Augmented DickeyFuller

PhillipsPerron

I(0)

I(0)

I(1)

I(1)

C&T

C&T

C&T

C&T

2.29
1.83
0.84
1.10

0.96
1.00
2.23
1.24

3.90
5.01
4.29
6.32

5.02
5.35
4.20
6.25

1.97
1.32
0.91
1.78

1.13
0.72
2.07
3.11

3.90
5.01
4.29
7.93

5.17
5.38
4.20
8.25

Note: The critical values for ADF and PP tests with constant (c) and with constant &
trend (C&T) 1%, 5% and 10% levels of signicance are 3.711, 2.981, 2.629 and
4.394, 3.612, 3.243 respectively.
Source: Authors' estimation.

Variables

Coeff.

t-stats

Prob.

C
GDP(-1)
LAB
LAB(-1)
CAP
CAP(-1)
CAP(-2)
TOT
TOT(-1)

0.599
0.863
0.963
0.652
0.211
0.079
0.061
0.245
0.796

0.738
2.119
2.676
1.446
3.510
2.613
0.534
2.975
0.924

0.470
0.048
0.015
0.166
0.003
0.018
0.600
0.008
0.368

Adj. R2
D.W stats
F-stats (Prob.)

0.998
2.209
321.091 (0.000)

Source: Authors' estimation.

Table 4.2
Lag length selection & bound testing for cointegration.

5. Stability of long-run model: A rolling window analysis

Lags order

AIC

HQ

SBC

F-test statistics

0
1
2

5.370
17.136
17.10

5.313
16.850
16.589

5.178
16.176
15.375

81.864

Source: Authors' estimation.

1% level of signicant.

(2011). The coefcient of terms of trade showing that the 1% increase

in terms of trade causes the increases in the economic growth by
0.245%. It is concluded that the terms of trade is a better leading indicator for economic growth in India. The following model is used to
check the short-run relationship among the considered variables
with the different lag length.

6. Stability of short-run model

GDPt 0 1 GDPt1 2 LABt 3 LABt1

p

i1

i1

i1

4 CAPt 5 CAPt1 6 CAPt2 7 TOTt

i1
p

i1

i1

i1

8 TOTt1 nECt t
i1

Table 4.5 represents the short-run relationship between terms of

trade and economic growth. Results indicate that the lagged error
correction term for the estimated economic growth equation is both
negative and statistically signicant. This conrms a valid short-run
relationship between terms of trade and economic growth in India.
The coefcient of error term is 0.35 suggesting that about 35% of
disequilibrium is corrected in the current year. Results indicate the
positive and signicant effect of terms of trade on economic growth
in the short run as well.

Table 4.3
Lags dened through VAR of variables.
Lag

GDP
LAB
CAP
TOT

Selected lags

SBC

SBC

SBC

SBC

1.520
0.560
0.724
1.767

4.651
9.857
2.976
1.556

4.289
9.391
3.063
1.919

1
1
2
1

Source: Authors' estimation.

Indicate minimum SBC values.

The stability of coefcients of the long-run model in the sample

size is evaluated by using the rolling window estimation method.
Fig. 5.1 and Table 5.1 represent the results of rolling window regression method of terms of trade. Two standard deviation bands show
the upper and lower bounds.
Results indicate that the terms of trade is having positive coefcients throughout the sample period. The results of Fig. 5.1 show
that the coefcient of terms of trade increases from 1984 to 1989
then there is a decrease in the coefcient of terms of trade from
1990 to 1996. In the sample range from 1997 to 2005 we have seen
a steady coefcient. In 2006 and 2007 the coefcient of terms of
trade has increased very rapidly but from 2008 to 2010 the coefcient
of terms of trade is continuously decreasing.

The stability of short-run model in the sample size is evaluated by

using the cumulative sum (CUSUM) and CUSUM of square test on the
recursive residuals. CUSUM test detects systematic changes from the
coefcients of regression, while, CUSUM of square test is able to detect the sudden changes from constancy of regression coefcients
(Brown et al., 1975).
Fig. 6.1 and 6.2 represents the results of CUSUM and CUSUM of
square tests respectively. Results indicate that the statistics of both
CUSUM and CUSUM of square test lie within the interval bands at
5% condence interval. Results suggest that there is no structural instability in the residuals of equation of economic growth.

Table 4.5
Short-run results using ARDL approach.
Variables

Coeff.

t-stats

Prob.

C
D(GDP(-1))
D(LAB)
D(LAB(-1))
D(CAP)
D(CAP(-1))
D(CAP(-2))
D(TOT)
D(TOT(-1))
EC(-1)

0.032
0.743
0.986
0.235
0.228
0.066
0.039
0.657
0.803
0.351

0.097
1.524
2.405
1.633
2.335
3.011
0.570
3.283
1.331
3.037

0.924
0.147
0.029
0.122
0.033
0.008
0.577
0.005
0.202
0.008

Adj. R2
D.W stats
F-stats (Prob.)

0.907
1.946
14.167 (0.000)

Source: Authors' estimation.

944

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

15
10
5
0
-5
-10
-15
92

94

96

98

00

CUSUM
Fig. 5.1. Coefcient of TOT and its two S.E. bands based on rolling OLS (dependent variable: GDP).

Table 5.1
Long-run coefcients of terms of trade.
Year

Coeff.

Year

Coeff.

1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995

0.211
0.209
0.200
0.172
0.188
0.330
0.430
0.458
0.754
0.546
0.538
0.564
0.436
0.430
0.417

1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

0.394
0.433
0.497
0.468
0.465
0.509
0.406
0.388
0.382
0.391
0.492
0.722
0.958
0.627
0.561

Note: Coefcient shows long-run elasticity.

Source: Authors' estimation.

7. Causality analysis
In this section three different techniques of causality analysis
namely, Granger causality analysis, 18 Toda and Yamamoto modied
Wald test causality analysis 19 and variance decomposition method 20
have been used to analyze the robustness of causal relationship between terms of trade and economic growth.

04

06

08

Fig. 6.1. Plot of cumulative sum of recursive residuals. The straight lines represent critical bounds at 5% signicance level.

It is assumed that and v are uncorrelated. There are two variables

and dealt with bilateral causality. The above equation states that Y is
related to its lag values and X is related to its lag values.
The results of Granger causality test are reported in Table 7.1.1.
Results show the bidirectional causal relationship between terms of
trade and economic growth.
7.2. Toda and Yamamoto modied Wald test causality analysis
The direction of causality between dependent and independent
variables is analyzed by using the causality test based on Toda and
Yamamoto (1995) procedure. This test uses a modied Wald
(MWALD) test which can be applied irrespective of whether the underlying variables are purely I(0), I(1) or mutually co-integrated.
Toda and Yamamoto (1995) augmented Granger causality test uses
the Seemingly Unrelated Regression (SUR) technique through estimating a two-equation system. The Wald test improves efciency
when SUR models are used in the estimation. So, the model can be
specied as follows:
kd

kd

i1

ti

Y t 1 1i Y ti 2i X ti yt

kd

kd

i1

ti

X t 2 1i Y t i 2i X t i xt

7.1. Granger causality analysis

The direction of causality between dependent and independent
variables is analyzed by Granger (1969) causality test. We determine
the causality analysis of our economic growth model on lag one. Jones
(1989) favors the ad hoc selection method for lag length in Granger
causality test over some of other statistical method to determine optimal lag. The equation of Granger causality model is given below:

02

5% Significance

1.6

1.2

0.8
t

i1

i1

i1

i1

Y i X ti i Y ti

X i X ti i Y ti v

0.4

0.0

-0.4
92

94

96

98

00

02

04

06

08

18

See, Granger (1969).

See, Toda and Yamamoto (1995).
20
The Variance Decomposition method is estimated through VAR framework, it
shows the proportion contribution in one variable caused by the shocks in other variables, Pesaran and Shin (1998).
19

CUSUM of Squares

5% Significance

Fig. 6.2. Plot of cumulative sum of squares of recursive residuals. The straight lines represent critical bounds at 5% signicance level.

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

Table 7.1.1
Results of Granger causality test.

945

Table 7.3.1
Results of variance decomposition approach.

Variables

F-statistic

Probability

Period

LAB

CAP

TOT

GDP does not Granger cause LAB

LAB does not Granger cause GDP
GDP does not Granger cause CAP
CAP does not Granger cause GDP
GDP does not Granger cause TOT
TOT does not Granger cause GDP

4.694
3.784
3.573
3.199
6.676
4.393

0.020
0.027
0.027
0.038
0.018
0.024

Variance decomposition of GDP

1
68.611
2
44.808
3
45.028
4
42.808
5
41.529
6
31.169
7
27.001
8
22.738
9
23.297
10
19.710

0.000
0.568
1.492
3.998
4.845
15.156
19.226
23.151
22.065
25.000

27.671
26.310
25.101
25.266
25.911
26.286
26.181
26.130
26.448
27.209

3.717
28.314
28.379
27.929
27.715
27.389
27.592
27.981
28.190
28.081

Variance decomposition of TOT

1
15.112
2
14.316
3
15.650
4
23.787
5
26.355
6
26.424
7
26.195
8
25.894
9
25.712
10
25.576

0.025
0.141
1.839
2.272
2.160
2.416
2.983
3.739
4.433
4.929

0.339
8.001
23.023
25.539
25.456
25.251
25.271
25.522
25.643
25.597

84.524
77.542
59.488
48.403
46.029
45.909
45.550
44.845
44.212
43.897

Note: The lag length is 1.

Source: Authors' estimations.

Where k is the optimal lag order, d is the maximum order of integration of the series in the system, and yt and xt are error terms that
are assumed to be white noise. Usual Wald tests are then applied to
the rst k coefcient matrices using the standard 2 - statistics.
The results of Toda and Yamamoto (1995) procedure based causality test are reported in Table 7.2.1. Results indicate the bidirectional causal relationship between terms of trade and economic growth.
7.3. Variance decomposition analysis
Generalized forecast error variance decomposition method under
vector autoregressive (VAR) system has been used to analyze the
strength of the causal relationship of terms of trade and economic
growth. The variance decomposition method provides the magnitude
of the predicted error variance for a series accounted for by innovations from each of the independent variable over different time period. Wong (2010), Hye (2012), Shahbaz et al. (2012) and Raza and
Jawaid (2013) have used this approach to nd causal relationship
among considered variables. Table 7.3.1 represents the results of variance decomposition analysis.
Results of Table 7.3.1 show that in the rst round the change in
economic growth explains 68.61% by its own innovations, 27.67% by
capital and 3.72% by terms of trade. In the second period 44.81% explains by own innovation, 0.57% by labor, 26.31% by capital and
28.31% by terms of trade. In period ve the shocks in economic
growth explain 41.53% by own innovation, 4.85% by innovations of
labor, 25.91% by innovations of capital and 27.72% by innovations of
terms of trade. In the tenth period the shocks in economic growth explain 19.71% by own shocks, while, 25.00% is explained by innovations of labor, 27.21% is explained by innovations of capital and
28.08% is explained by innovations of terms of trade.
The shocks in terms of trade explain 84.52%, 77.54%, 46.03% and
43.90% by its own innovations in periods 1, 2, 5 and 10 respectively.
The shocks in terms of trade explain 15.11%, 14.32%, 26.36% and
25.58% by innovation of economic growth in periods 1, 2, 5 and 10 respectively. These ndings suggest the bidirectional causal relationship between terms of trade and economic growth in India.
8. Conclusion and recommendations
This study investigates the effect of terms of trade on economic
growth of India by using the annual time series data from the period
Table 7.2.1
Results of Toda and Yamamoto causality test.
Dependent variable

Modied Wald statistics

GDP

LAB

CAP

TOT

GDP

TOT

6.530
(0.021)

3.208
(0.086)
1.853
(0.189)

5.772
(0.027)
10.790
(0.005)

6.676
(0.018)

Note: The lag length for GDP, LAB and TOT is 1 and CAP is 2 as per Schwartz Bayesian
Criteria (SBC).
Source: Authors' estimations.

GDP

Source: Authors' estimation.

of 1980 to 2010. The ARDL bound testing cointegration conrms the signicant positive relationship between terms of trade and economic
growth in the long run as well as in the short run. Results of Granger
causality, Toda and Yamamoto Modied Wald causality and variance
decomposition tests conrm the bidirectional causal relationship between terms of trade and economic growth in India. Rolling window estimation indicates that the terms of trade is having positive long-run
coefcients throughout the sample period. Results of CUSUM and
CUSUM of square suggest that there is no structural instability in the residuals of equation of economic growth in the short run.
It is suggested that benecial terms of trade is better for economic
growth in India. At this stage we can set the direction of future research
that the relationship between commodity groupwise and countrywise
bilateral terms of trade with economic growth should be analyzed.
These results would be helpful for policy makers of India to frame growth
enhancing countrywise as well as commodity-wise trade policies.

References
Arize, A., 1996. Cointegration test of a long-run relation between the trade balance and
the terms of trade in sixteen countries. North American Journal of Economics and
Finance 7, 203215.
Balasubramanyam, V.N., Salisu, M.A., Sapsford, D., 1996. Foreign direct investment and
growth in EP and IS countries. The Economic Journal 106 (434), 92105.
Barro, R.J., 1999. Determinants of Economic Growth: A Cross Country Empirical Study.
MIT Press, Cambridge, M.A.
Bleaney, M., Greenaway, D., 2001. The impact of terms of trade and real exchange rate
volatility on investment and growth in Sub-Saharan Africa. Journal of Development
Economics 65, 491500.
Bouakez, H., Kano, T., 2008. Terms of trade and current account uctuations: the
HarbergerLaursenMetzler effect revisited. Journal of Macroeconomics 30, 260281.
Brown, R.L., Durbin, J., Ewans, J.M., 1975. Techniques for testing the constancy of regression relations overtime. Journal of the Royal Statistical Society 37, 149172.
Cakir, M., 2009. Terms of Trade and Economic Growth of Emerging Market Economies.
The International Conference on Administration and Business University of Bucharest
(1415 November).
Cashin, P., McDermott, C.J., 2002. The long-run behaviour of commodity prices: small
trends and big variability. International Monetary Fund Staff Papers 49 (2),
175199.
Cashin, P., McDermott, C.J., 2002. Terms of trade shock and the current account: evidence from ve industrial countries. Open Economies Review 13, 219235.
Dicky, D.A., Fuller, W.A., 1979. Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association 74, 427431.
Fatima, N., 2010. Analysis of Terms of Trade Effect for Pakistan. Pakistan Institute of Development Economics: Working Paper, No 59.
Granger, C.W.J., 1969. Investigating causal relations by econometric models and crossspectral methods. Econometrica 37 (3), 424438.

946

S.T. Jawaid, S.A. Raza / Economic Modelling 33 (2013) 940946

Hadass, Y.S., Williamson, J.G., 2001. Terms of Trade Shocks and Economic Performance,
18701940: Prebisch and Singer revisited. National Bureau Economic Research:
Working Paper, No. 8188.
Hamori, S., 2008. Trade balances and the terms of trade in G-7 countries: penal cointegration
approach. Applied Econometrics and International Development 8 (2), 1322.
Harberger, A.C., 1950. Currency depreciation, income and the balance of trade. Journal
of Political Economy 58 (1), 4760.
Haug, A., 2002. Temporal aggregation and the power of co-integration tests: a Monte
Carlo study. Oxford Bulletin of Economics and Statistics 64, 399412.
Hye, Q.A.M., 2012. Long term effect of trade openness on economic growth in case of
Pakistan. Quality and Quantity 46 (4), 11371149.
Jawaid, S.T., Waheed, A., 2011. Effects of terms of trade and its volatility on economic growth:
a cross country empirical investigation. Transition Studies Review 18 (2), 217229.
Jawaid, Syed Tehseen, Raza, Syed Ali, 2012. Workers' remittances and economic
growth in China and Korea: an empirical analysis. Journal of Chinese Economic
and Foreign Trade Studies 5 (3), 185193.
Jones, J.D., 1989. A comparison of lag length selection techniques in tests of Granger
causality between money growth and ination: evidence from the US: 195986.
Applied Economics 21 (6), 809822.
Kaneko, A., 2000. Terms of trade, economic growth, and trade patterns: a small openeconomy case. Journal of International Economics 52, 169181.
Kohpaiboon, A., 2003. Foreign trade regimes and the FDI-growth nexus: a case study of
Thailand. Journal of Development Studies 40 (2), 5569.
Laursen, S., Metzler, L.A., 1950. Flexible exchange rates and the theory of employment.
The Review of Economics and Statistics 32, 281299.
Lutz, M., 1999. A general test of the PrebischSinger hypothesis. Review of Development Economics 3 (1), 4457.
Mendoza, E.G., 1997. Terms-of-trade uncertainty and economic growth. Journal of Development Economics 54, 323356.
Misztal, P., 2010. The HarbergerLaursenMetzler effect. Theory and practice in Poland.
The Romanian Economic Journal 38, 129146 (Year XIII).

Otto, G., 2003. Terms of trade shocks and the balance of trade: there is a Harberger
LaursenMetzler effect. Journal of International Money and Finance 22, 155184.
Pesaran, M.H., Pesaran, B., 1997. Working with Microt 4.0: Interactive Econometric
Analysis. Oxford University Press, Oxford.
Pesaran, M.H., Shin, Y., 1998. Generalized impulse response analysis in linear multivariate models. Economic Letters 58 (1), 1729.
Pesaran, M.H., Shin, Y., 1999. An autoregressive distributed lag modeling approach to
co-integration analysis. In: Strom, S. (Ed.), Econometrics and Economic Theory in
the 20th Century: The Ragnar Frisch Centennial Symposium. Cambridge University
Press, Cambridge.
Pesaran, M.H., Shin, Y., Smith, R.J., 2000. Structural analysis of vector error correction
models with exogenous I(1) variables. Journal of Econometrics 97, 293343.
Pesaran, M.H., Shin, Y., Smith, R.J., 2001. Bounds testing approaches to the analysis of
level relationships. Journal of Applied Economics 16, 289326.
Phillips, P.B., Perron, P., 1988. Testing for a unit root in time series regression. Biometrika
75, 335346.
Prebisch, R., 1950. The Economic Development of Latin American and Its Principal
Problems (Lake Success, NY: United Nations, Department of Economic Affairs).
Reprinted in Economic Bulletin for Latin America 7 (1962), 122.
Raza, S.A., Jawaid, S.T., 2013. Terrorism and tourism: a conjunction and ramication in
Pakistan. Economic Modelling 33, 6570 (July).
Shahbaz, M., Islam, F., Aamir, N., 2012. Is devaluation contractionary? Empirical evidence
for Pakistan. Economic Change and Restructuring 45.
Singer, H.W., 1950. The distribution of gains between investing and borrowing countries.
American Economic Review 40, 473485.
Toda, H.Y., Yamamoto, T., 1995. Statistical inference in vector autoregressions with
possibly integrated processes. Journal of Econometrics 66, 225250.
Wong, H.T., 2004. Terms of trade and economic growth in Malaysia. Labuan Bulletin of
International Business and Finance 2 (2), 105122.
Wong, H.T., 2010. Terms of trade and economic growth in Japan and Korea: an empirical
analysis. Empirical Economics 38, 139158.