Pak. J. Agri. Sci., Vol.

47(1), 82-87; 2010

ISSN (Print) 0552-9034, ISSN (Online) 2076-0906
http://www.pakjas.com.pk

TECHNOLOGY TRANSFER AND AGRICULTURAL GROWTH IN PAKISTAN

Zahoor Hussain Javed1, Muhammad Farooq2* and Hamid Ali1
1
Department of Economics, GC University, Faisalabad
2
Department of Sociology, GC University, Faisalabad
*Corresponding author’s e-mail: [email protected]

This study was designed to measure the impact of the technologies on agricultural growth. Agricultural
technologies and knowledge have been restructured and distributed by public institutions at large scale. But over
the past two decades, use of pesticides, fertilizers, tractors and tube wells for agricultural production has
increased rapidly, and the world economy has become more efficient and well-organized. These factors have
positive impact on investment in agricultural research, technology and have influenced the growth of economy of
Pakistan. This study shows empirical relationship between tube wells, pesticides, tractors, fertilizers and
agriculture growth over the period 1971-2007. For empirical analysis simple OLS, (Ordinary Least Square),
Johnson’s Co-integration and Error correction model techniques have been used to find relationship between
agriculture growth and explanatory variables. The public sector still has a role to play, particularly in managing the
new knowledge, supporting research to fill up the left over gaps, encouraging and regulating farmers associations,
and ensuring their effects on the agriculture growth are adequately reviewed.
Keywords: Extension services, agriculture growth, farm machinery, agrochemicals

INTRODUCTION made in reducing the use of chemicals in agriculture,

represents a serious environmental hazard in the
Pakistan is a developing country and agriculture is the United States and Pakistan. Clevo (2000) focus
backbone of country’s economy. It is currently the attention on the environmental and human costs of
seventh most populous country in the world. Its commercial agricultural production, especially the
agriculture sector occupies an important position in its Green Revolution technology in South Asia during the
economy and contributes about 21 percent of the gross last few decades. Modern commercial agricultural
domestic product (Economic Survey, 2007-08). Since practices involving chemical inputs such as fertilizers
the end of World War II, the public institutes of and pesticides have been associated with huge
developed countries have helped transfer agricultural increases in food production never witnessed before
technologies to developing countries. During this and, in the case of cereal production under Green
period, most of the developing countries in Latin Revolution technology, recorded spectacular growth.
America and Africa, as well as some countries in Asia As statistics show, production and productivity have
(like India, Thailand and Pakistan), have depended increased. MacDonld (2005) discusses the importance
heavily on agricultural production to sustain their of biotechnology to addressing the problem of
economies (Piñeiro, 2007). Robert (2004) and universal starvation. Technologies developed at the
Thurston (1999) state that the fertilizer supplier has research institutes remain unproductive if not
been busy, oilseed rape and winter barley is both transferred and adopted by the farmers. Technology
complete, leaving oats and wheat to be given the transfer projects ensure better and efficient transfer of
balances. improved technology from the research institutes to the
Technologies have become very important in physical stakeholders. To increase the production, government
products, like farm machinery or agrochemicals. has started many projects like farm field schools (FFS)
Speedy growth in such sectors has led to a rapid which are also promoted by the World Bank as it is
development of private firms that make, produce and more effective approach to extend the science based
sell technology and products based on agrochemicals. knowledge and practices. For an effective transfer of
Innovation and technology development has always technology, Government of Pakistan has also launched
been main source of agriculture because agricultural different extension approaches under the umbrella of
progress and enlargement depends upon interference agriculture department. In government sector many
of modern technology tools by agricultural scientists efforts have been made to improve the situation of
and experts. Robbins and Julie (2003) suggest that extension in the country. Inspite of serious deficiencies
the increasing use of fertilizers and pesticides to and financial constraints agricultural extension has
housing lawns, which has started to offset the gains made significant contribution in increasing agricultural

82
 Javed, Farooq and Ali

production. Until 1978, a general extension approach comprehensive evaluation be taken at the tendencies
was in operation in the country. This approach was in aggregate use of inputs in agricultural production in
basically one of “Technology Transfer” from Pakistan. The main group of crops in Pakistan is
government to agriculture sector. Well organized wheat, cotton, pulses, rice and maize. The emphasis
extension services can bridge the gap between the will be given on food crops and other major crops
potential productivity and the current productivity because of the urgent need to re-establish food
Nevertheless the farmers of Pakistan have started to inefficiency in Pakistan and the consequent poverty
use tractors, pesticides, tube wells and fertilizers to situation. The importance and relevance of these
enhance agriculture growth. This will certainly raise the factors may differ from country to country and may also
spirit of research for the development and change overtime. Time series data are used for the
advancement of agriculture in Pakistan. Therefore, we estimation of regression results of dependent and
see advancement of Agriculture by using pesticides, explanatory variables. Therefore, using neoclassical
fertilizers, tube wells and tractors on nominal production function, in log-linear form the agricultural
agriculture growth of Pakistan. Thomas et al. (2007) growth equation is:
concludes that optimizing behavior of individual Ln (AG) = α + β1 ln (Pes) + β2 ln( Fer) + β3 ln(Tract) +
organism’s short-run ecosystem equilibrium depends β4 ln( Tub) +€
on the farmer's use of fertilizers and pesticides and on β1, β2, β3, β4, > 0
the rat’s population which is affected by pesticides. The The expected sign of all coefficients are positive.
planning and organization of agriculture has been Where
adjusted when prices and markets are introduced into Ln= natural logarithm
agricultural production and the direct planning is AG = Agriculture growth
restricted. Pes = pesticides
The study is an attempt to discover farmer’s behaviors Fer = fertilizers
with regard to use of agrochemical. They found that Tract = tractors
judgments on fertilizer, pesticides, and use of agro-film Tub = tub wells
have different impacts on crop pledge contribution, and € = white noise error term
are influenced by the latter in different ways. It is also The data for agriculture growth, tube wells, fertilizers,
implied that comforting farmer’s participation in crop pesticides, and tractors had been taken from various
insurance under current low-premium and low- “Economic Surveys of Pakistan” (1987-88, 1995-96
indemnity terms does not have a significantly negative and 2006-07). The agriculture growth has been taken
impact on the atmosphere Zhong et al. (2007), Parker as dependent variable, while pesticides, fertilizers,
(2005), Nanjingad and Fuzhou (2006) however, Childs tractor and tube wells have been taken as explanatory
(2004), Thurston (1999) and Hollis (2004) declare that variables. The units of agriculture growth, pesticides,
the market performance of the agricultural products fertilizers, tractors and tube wells are thousands tons,
depend on use of chemical fertilizers. Growth of the thousands liters, thousands tons, thousands numbers
pesticides market in the U.S. may contribute to the and thousands numbers respectively. Since quarterly
decline in U.S. markets for chemical insecticides. Hollis data were not available, therefore annual data had
(2004) and Doyle (2007) have analyzed the financial been used. The impact of technology transfer on
risks associated with dependence on purchased farm agriculture growth in Pakistan had been covering the
inputs regardless of the benefits of increased efficiency period of 1971 to 2006. All involved series were
and greater productivity. transformed into log form. Log form was used to
reduce the problem of heteroskedasticity. Simple least
MATERIALS AND METHODS square regression method, Jhonsen co- integration
and error correction model were used to analyze the
Inputs are a subset of the independent variables for trends and patterns dependent and explanatory
agricultural out put growth. The concert of the major variables in Pakistan.
input factors is an essential indicator to determine the
Unit root, OLS, Johnson’s Co integration and Error
most efficient development interference. These inputs
correction Model
contain pesticides, fertilizers, tractors and tube wells.
The productivity of the major crops grown and the The order of co-integration can be detected by using
following production levels are clearly affected by the unit root. If it is found that all series of variables are
combination of resources used. In order to effectively based on non-stationary at I(0) and are stationary at
determine the effect of technology transfer on I(1). Nevertheless, both tests give the same result after
agriculture growth in Pakistan, it is essential that a first variation of series, that is all series are stationary

83
 Technology transfer and agricultural growth

at I(1). So it is clear that Johnson’s co- integration The OLS results in Table 2 show that from the four
technique can be used to detect the relationship basic variables two variables do not have the expected
between the variables at I(0). The choice of lag length sign, this is so because in case of tractors, some
is based on the lower the values of Akaike and tractors are being used for commercial purpose
Schwarz statistics. The lowest the value of Akaike and similarly in case of pesticides, the majority of farmers
Schwarz shows the goodness of the model. do not use pesticides as per directions of agriculturist .
Before testing the co-integration, first researcher Nevertheless, two variables have accepted positive
estimated whether the time series was stationary or sign and they are highly significant. However, there are
nonstationary? Several tests had developed to check problems in the above regression results from the point
the stationary or nonstationary status in the time series of view of standard econometric assumptions. The
econometrics literature. In most of these tests the null equality of R2 and DW involved that the regression
hypothesis was a unit root, and it was rejected only might be spurious regression that arises in the
when there was strong evidence against it. For this presence of non-stationary variables. Furthermore, the
purpose, we test each series by well known above regression results do not take into consideration
Augmented Dickey-Fuller unit root test. dynamic aspects and problem of serially co-related
∆Y = α0 + µ Yt -1 + α∑∆ Yt -1 +€t errors making parameters estimates unreasonable.
Where ∆Y =Yt - Yt -1, €t is the error term and µ is chosen Breusch-Godfrey serial correlation LM test shows
to ensure serially uncorrelated residuals. The variables normality of model and also shows that there is no
containing in Table 1 are examined for stationary. correlation among variables. Though the sample size is

Table1: Results of unit root test for examination of stationary

Variables With trend level With trend 1st diff Without trend Level Without trend 1st diff
AG -1.089161 -4.745327 -4.554422 -4.852108
Pes -1.586693 -10.11765 -4.914867 0.793
Fer -2.747133 -4.119332 -0.696755 -4.855187
Tract -2.181647 -6.941059 -2.473929 -7.163104
Tub -0.220578 -5.107973 -2.223542 -5.033139
AG = Agricultural Growth, Pes = Pesticides, Fer = Fertilizers, Tract = Tractors, Tub = Tube wells

The unit roots results of the Augmented Dickey Fuller not large (37 observations), researcher subjected the
test are presented in Table.1 The table shows the residuals of regression (1) to Q-statistics, LM test for
results of all variables in level form and at first serial co-relation and ARCH test. F-version of these
difference with trend and without trend. To determine tests shows significance. Therefore, results were not
the order of integration, we also applied ADF unit root appropriate for analysis in this version. The results
test to examine the variables in their first differences. were displayed in Table.3 and Table.4 respectively.
The null of stationarity is accepted for all the variables For that reason, the data is reexamined for time series
for their first differences. Therefore, all the variables properties.
are first difference stationary I (0) thus integrated of
order 1. Table 3. Breusch-Godfrey Serial Correlation LM Test
The ordinary least squares (OLS) estimates of
equation (1) are as under: Breusch-Godfrey Serial Correlation LM Test
F-statistic 11.94401 Probability 0.000164
Table 2. Ordinary least squares regression results
for selected variables Obs*R-squared 16.26018 Probability 0.000295

Predictor Coef SE Coef T P Table 4. ARCH Test for Significance of the Model
Constant -11.076 0.784 -14.124 0.000
ARCH Test
Pes -0.015 0.058 -0.264 0.793
Fer 0.766 0.112 6.825 0.000 F-statistic 8.355065 Probability 0.006752
Tract -0.060 0.052 -1.162 0.254 Obs*R-squared 7.071136 Probability 0.007834
Tub 1.561 0.126 12.354 0.000
R2 0.99 D.W 0.78 If the hypothesis of nonsatationarity is recognized for
the basic variables, it is striking and significant that the
Source: Author’s own calculations
time series data are examined for co-integration.

84
 Javed, Farooq and Ali

Tabe 5. Johansen Maximum Eigenvalue Test (AG, Pes, Trac, Tub and Fer)
LOG ( AG) LOG (Pes) LOG (Trac) LOG (Tub) LOG (Fer)

Lags interval: 1 to 1
Hypothesized Eigenvalue Trace statistics 5 Percent Hypothesized
No. of CE(s) Critical Value Critical Value No. of CE(s)
None ** 0.605539 81.70979 68.52 None **
At most 1 * 0.547418 50.08176 47.21 At most 1*
At most 2 0.368149 23.12705 29.68 At most 2
At most 3 0.176351 7.517578 15.41 At most 3
At most 4 0.026731 0.921224 3.76 At most 4
Max-eigenvalue test indicates 2 co-integrating eqn(s) at the 0.05 level
*denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

Therefore, the maximum likelihood method of and they are highly significant. Tractor and pesticides
Johansen can be used to find co-integration have insignificant impact on agriculture growth; this is
relationship. The eigenvalue statistic tests Ho that the so because, many tractors are being used for
number of co-integrating vectors is r against the commercial purpose similarly in case of pesticides, the
alternative of r +1 co-integrating vectors. In view of the majority of farmers do not use pesticides as per
fact that Johansen approach has become standard in directions of agriculturist. The results of this study
the econometric literature. The Johansen co- coincide with the empirical consequences of Prakash
integration results are shown in Table 5. Trace test and Gregory (2004), Thomas et al (2007). Zhong et al.
indicates two co integrating equations while Maximum- (2007) respectively.
Eigenvalue test also indicates two co-integrating
relationship. Therefore, Agriculture growth and tubwell CONCLUSION
and fertilizer are co-integrated thus having long-run
relationship with agriculture growth. This paper has examined the impact of technology
transfer on the agricultural growth of Pakistan for the
Error-Correction Model
period 1971-2007. The findings in regression model
∆E Gt = α + ∂Zt+1 + ∑ βi∆ Ln EGt-i + ∑ µi Ln ∆ (PES)t-I + indicate that the equality of R2 involved that the
∑ €i∆ Ln (TRACT) t-I + regression might be spurious regression that arises in
∑ φi∆ Ln (TUBLE) t-I + ∑ ηi ∆Ln (FER) t-I +ψt the presence of non-stationary variables. Furthermore,
the regression results do not take into consideration
Zt+1 is the error correction term generate from the dynamic aspects. These have problem of serially
Johansen multivariate procedure and the parameter ∂ correlated errors. Johansen’s co integration method
is the error correction coefficient that measures the shows fertilizer and tub well were co-integrated and
response of the regressed in each period to departures have long-run equilibrium relationship with agricultural
from equilibrium. The error correction model’s results growth, while pesticide and tractor were statistically
are shown in Table 6. insignificant and show negative relationship with
To choose a suitable lag length, we used AIC, SC and agricultural growth, notwithstanding, error-correction
the optimal lag length was 1. Error correction results model results also support the co integration results. It
show that the error correction term Zt+1 has the means raise of quantity of tube well and fertilization
correct positive sign and is significant for agriculture may increase agricultural growth of the Pakistan and
growth, tub well and fertilization and indicates the long- use of tractors and pesticides may be rehabilitated if
run equilibrium relationship between these variables. authorities and farmers give proper attention.
An estimate of 0.19 for agriculture growth indicates that Pakistan’s economy will definitely get advantage from
19% of the preceding year disequilibrium included in technology transfer provide that the country chases
the current year. Moreover, use of fertilization and tub concrete policies and infrastructures program.
well raise agriculture growth 34% and 19% respectively

85
 Technology transfer and agricultural growth

Table 6. Error Correction Results (AG, Tract, Pes, Fer and Tub)
Dependent Variable> D (LOG(AGRIGROWTH)) D (LOG(TRACTOR)) D(LOG(PES)) D(LOG(FER)) D(LOG(TUBLE))
α 0.196704* 0.511549 -0.382971 0.095092 -0.060492
(0.04251) (0.36648) (0.46245) (0.11080) (0.05888)
(4.62757) (1.39584) (-0.82814) (0.85820) (-1.02743)
D(LOG(AGRI GROWTH(-1) -0.108680 -0.585153 1.172268 -0.114248 -0.180360
(0.16511) (1.42352) (1.79628) (0.43039) (0.22869)
(-0.65823) (-0.41106) (0.65261) (-0.26545) (-0.78866)
D(LOG(AGRI GROWTH(-2) -0.228043 -0.526683 0.865931 0.032529 0.513436
(0.13274) (1.14443) (1.44412) (0.34601) (0.18386)
(-1.71798) (-0.46021) (0.59963) (0.09401) (2.79257)
D(LOG(TRA(-1) 0.095070* 0.236801 -0.010037 -0.005606 -0.048011
(0.02234) (0.19258) (0.24301) (0.05823) (0.03094)
(4.25620) (1.22962) (-0.04130) (-0.09627) (-1.55180)
D(LOG(TRA(-2) 0.002943 0.001375 0.051668 0.122089 -0.026449
(0.02584) (0.22279) (0.28113) (0.06736) (0.03579)
(0.11389) (0.00617) (0.18379) (1.81251) (-0.73896)
D(LOG(PES(-1) 0.016448 0.060838 -0.145726 -0.021537 -0.087004
(0.01929) (0.16628) (0.20983) (0.05028) (0.02671)
(0.85283) (0.36587) (-0.69450) (-0.42839) (-3.25686)
D(LOG(PES(-2) -0.027267 -0.044621 -0.137704 0.048121 -0.005237
(0.01830) (0.15778) (0.19909) (0.04770) (0.02535)
(-1.48998) (-0.28281) (-0.69165) (1.00875) (-0.20660)
D(LOG(FER(-1) 0.199302* -1.119562 -0.606774 0.023610 -0.092911
(0.08552) (0.73735) (0.93044) (0.22293) (0.11846)
(2.33039) (-1.51836) (-0.65214) (0.10590) (-0.78433)
D(LOG(FER(-2) 0.131683 -0.853359 -0.413528 -0.321002 0.291011*
(0.09153) (0.78914) (0.99578) (0.23859) (0.12678)
(1.43869) (-1.08138) (-0.41528) (-1.34540) (2.29544)
D(LOG(TUBLE(-1) 0.347912* -1.406109 1.120997 0.220010 0.583383*
(0.14558) (1.25513) (1.58381) (0.37948) (0.20164)
(2.38985) (-1.12029) (0.70779) (0.57976) (2.89316)
D(LOG(TUBLE(-2) 0.142635 -1.211826 2.190824 -0.326362 0.469506*
(0.14818) (1.27758) (1.61213) (0.38627) (0.20525)
(0.96256) (-0.94853) (1.35896) (-0.84491) (2.28751)
Note: Standard errors & t-statistics in parentheses. Figures within parentheses are t – statistic and * indicates significance at 5%

REFERENCES Hollis, P.L. 2004. Purchased inputs raise risk levels.

Southeast Farm Press 31(9): 12-13.
Childs, R.D. 2004. Pesticide and Other Agricultural MacDonald, V. 2005. Agchem and Fertilizers Grow.
Chemical Manufacturing. American Nurseryman Chemical Week 167(1): 24-26.
200(4): 36-46. Nanjingand, W. and J. Fuzhou. 2006. What’s growing
Clevo, W. 2000. Environmental and human costs of on in market research? American Nurseryman
commercial agricultural production in South Asia. 204(10): 11-17.
International Journal of Social Economics 27(7-10): Parker, R. 2005. Non-crop uses–the way ahead for
816-831. pesticides. Specialty Chemicals 25(1): 20-21.
Doyle, R. 2007. Food Boom. Scientific American Piñeiro, M. 2007. Agricultural technology transfer to
296(5): 34-34. developing countries and the public sector.
Government of Pakistan. 2007. Economic Surveys of www.yahoo.com. Blog. Iphandbook.org. Access date
Pakistan: 1987-88, 1995-96, 2006-07 and 2007- 07-08-09.
08. Ministry of Finance, Islamabad. Prakash. S. and C. Gregory. 2004. Technology for life:
Gregory, T.R. 2004. Genome size is not correlated How biotech will save billions from starvation.
positively with Longevity in fishes. Exp. Gerontol. American Enterprise 15(2): 16-20.
39(3): 859-860.

86
 Javed, Farooq and Ali

Robbins, P.S. and T. Julie. 2003. Producing and Thomas, C., T. Eichner and P. Rüdiger. 2007. Optimal
consuming chemicals: The moral economy of the pest control in agriculture. Journal of Economic
American lawn. Economic Geography 79(4): 425- Dynamics and Control 31(12): 3965-3985.
451. Zhong, F., M. Ning and L. Xing. 2007. Does crop
Robert, F. 2004. Crops now responding to the warmer insurance influence agrochemical uses under
weather. Crops 79(4): 23-28. current Chinese situations? A case study in the
Thurston, C. 1999. Brazilian pesticide and fertilizer Manasi watershed, Xinjing. Agricultural Economics
imports increasing in double digits. Chemical 36(1): 103-112.
Market Reporter 255(3): 25-30.

87