ArticlePDF Available

Are Agricultural Markets in the Punjab Technically Efficient?

Authors:

Abstract and Figures

We testthe technical efficiency, measured by the degreeof integration,of agriculture marketsfor five crops in the Punjabprovince of Pakistanusing daily wholesale market prices from the Agriculture Management Information System (AMIS). We findthat potato, onion and mango markets are well integrated both horizontally and vertically, with the speed of price adjustment in most cases (mango isthe exception) being very rapid. We also find that kinnow and basmati rice markets are bothvertically fairly well integrated. Furthermore, we findthat trends in cropping patternsover the period 2000 to 2014 are in line with the changing market demand and government priceinterventions. The reformsintroducedby thePunjab Agriculture Marketing Regulatory Authority (PAMRA) Act 2020,aimed atincreasing competition in agriculture markets,have the potential tosignificantly improve economic efficiency.
Content may be subject to copyright.
The Lahore Journal of Economics
25 :1 (Spring 2020): pp. 89138
Are Agricultural Markets in the Punjab Technically
Efficient?
Mahniya Zafar*, Naved Hamid** and Fatima Arshad***
Abstract
We test the technical efficiency, measured by the degree of integration, of
agriculture markets for five crops in the Punjab province of Pakistan using daily
wholesale market prices from the Agriculture Management Information System
(AMIS). We find that potato, onion and mango markets are well integrated both
horizontally and vertically, with the speed of price adjustment in most cases
(mango is the exception) being very rapid. We also find that kinnow and basmati
rice markets are both vertically fairly well integrated. Furthermore, we find that
trends in cropping patterns over the period 2000 to 2014 are in line with the
changing market demand and government price interventions. The reforms
introduced by the Punjab Agriculture Marketing Regulatory Authority
(PAMRA) Act 2020, aimed at increasing competition in agriculture markets,
have the potential to significantly improve economic efficiency.
Keywords: Agricultural prices, market integration, price transmission,
market efficiency, agriculture marketing.
JEL Classification: Q110, Q111, Q113 and C110.
1. Introduction
Agriculture plays a significant role in economic development not
only for ensuring food and nutritional security but is the major source of
* Teaching and Research Fellow, Centre for Research in Economics and Business (CREB), Lahore
School of Economics, Pakistan. Email: mahniyazafar@gmail.com
** Professor and Director Centre for Research in Economics and Business (CREB), Lahore School
of Economics, Pakistan. Email: navedhamid@gmail.com
*** Research Associate, Centre for Economic Research in Pakistan (CERP). Email:
afatima.0013@hotmail.com
We would like to thank Dr. Syeda Rabab for her tremendous support and help in understanding,
running and interpreting the results of VAR models (or framework). However, it goes without
saying, that any errors that remain are entirely the responsibility of the authors.
Technical Efficiency in Punjab's Agricultural Markets
90
rural employment and contributes substantially to earnings from exports.
While the relative importance of agriculture has been declining in South
Asia, it is still considerable. The agriculture sector contributes about 40
percent of total employment in Bangladesh, India and Pakistan, and over
50 percent in Bhutan and Nepal (International Labor Organization, 2018)
1
.
Agriculture also accounts for over 20 percent of GDP in Pakistan and
Nepal, and about 15 percent in Bangladesh, Bhutan and India (World
Development Indicators, 2018)
2
.
In Pakistan, agricultural sector growth has slowed significantly
since 2000, with the slowdown being greater in the crops sub-sector (see
Figure 1)
3
. There is no consensus on why the decline in the growth rate
has occurred, but factors such as “inequality in farm sizes, limited
investment in irrigation systems, the slowing of adoption of new
technology and techniques and a weak extension service” have been cited
as likely causes (Valdes, 2013).
We suggest that the lack of efficiency of agricultural markets is an
additional factor responsible for this slowdown. We know that there are
multiple players at each stage of the agricultural marketing chain, that the
legal and regulatory framework of agricultural markets in the Punjab (and
the rest of the country) is archaic (Ahsan 2018) and that marketing margins
are high
4
. It is our view that inefficient agricultural markets could be
eroding the incentives for the producers to invest in productivity enhancing
inputs and technologies. We posit that market efficiency can be divided into
two components, i.e. economic efficiency and technical efficiency.
An economically efficient agricultural marketing system, defined as
a system where competition throughout the marketing chain, results in
total marketing costs of agricultural products being minimized and profits
earned by each of the players in the marketing chain being no more than
normal; and a technically efficient agricultural market being defined as one
where the various agricultural markets in the region are well integrated.
1
https://data.worldbank.org/indicator/SL.AGR.EMPL.ZS
2
http://wdi.worldbank.org/table/4.2
3
Agricultural growth has been declining since the 1990s, but it was still fairly healthy until 2000.
4
According to the World Bank, prior to the recent [in 2020] reforms “Farmers’ produce used to
pass through seven or eight different hands before reaching the consumer. Consequently, market
margins were high, but producers were left with little” (https://blogs.worldbank.org/
endpovertyinsouthasia/modernizing-punjabs-farming-benefit-farmers-and-consumers)
91 Mahniya Zafar, Naved Hamid and Fatima Arshad
In this article, we will focus on testing for technical efficiency of the
agricultural marketing system because, unfortunately, due to the lack of
data on farm gate prices and margins at different stages of the marketing
chain it is difficult for us to say much about its economic efficiency.
5
Figure 1: Historical Growth Rates for Pakistan's Agriculture and Crops
Source: Pakistan Economic Survey (1980-2016).
Market integration has been defined as the tradability or
contestability between markets (Barret and Li, 2002). It can be interpreted
as the extent to which price shocks are transmitted between spatially
separate markets (Goodwin, 2006) and can be measured in terms of
strength and speed of price transmission between markets across various
regions of a country (Beag and Singla, 2014). Market integration is
undoubtedly important because until agricultural markets are integrated,
producers and consumers will not realize their potential gains (Reddy,
2012) and the degree to which consumers and producers can benefit
depends on how domestic markets are integrated with world markets and
how the regional markets are integrated with each (Varela et al., 2012).
5
The economic efficiency of agricultural markets in the Punjab is expected to improve following
the approval in March by the Punjab Assembly of the PAMRA Actshort for Punjab Agriculture
Marketing Regulatory Authority Act 2020. The new law establishes a more transparent legal
regime to market agricultural produce to help safeguard the free flow of crops and stimulate food
supply (https://blogs.worldbank.org/endpovertyinsouthasia/modernizing-punjabs-farming-benefit-
farmers-and-consumers).
1980's 1990's 2000's 2010's
Agriculture 5.4 4.4 3.2 2.4
Crops 3.8 3.3 2.1 0.9
0.0
1.0
2.0
3.0
4.0
5.0
6.0
Percent
Technical Efficiency in Punjab's Agricultural Markets
92
The concept of market integration is often used as a measure of
market efficiency; however, in our view it is more appropriate to think of
it as a measure of the ‘technical efficiency’ of a market; this is how we use it
in this paper, with technical efficiency of the market for various crops
being evaluated in terms of transmission of price information among the
producer markets and between producer and consumer markets. Our
research not only adds to our understanding of the working of the
agriculture markets in Pakistan, but it also contributes to the overall
literature on agricultural markets because our analysis of market
integration is based on a unique data set that has daily price information,
in contrast to most of the existing research on integration in agricultural
product markets, which is based on analysis of monthly price data
6
.
We selected five crops for analysis, namely, potato, onion, mango,
kinnow and basmati rice. To test for strength and speed of price
transmission between agricultural markets, we use vector auto regressive
(VAR) models. It is seen that potato, onion and mango markets are well
integrated both horizontally and vertically, with the speed of price
adjustment in most cases (mango is the exception) being very rapid.
Therefore, we can say that these three markets are technically efficient. As
far as kinnow and basmati rice markets are concerned, both are vertically
fairly well integrated but we are unable to satisfactorily measure the
extent of horizontal integration due to lack of data.
In Section 2, we review the literature on market integration from the
perspective of methodologies used and the extent of market integration
estimated for different crops in other countries. Section 3 provides a
description of the data and research methodology and in Section 4 we
discuss the results. Section 5 provides a reality check on the impact of
efficient agricultural markets and Section 6 concludes.
2. Review of Literature
There is considerable literature on market integration and price
transmission. Markets are said to be integrated when a price increase or
decrease (shock) is transmitted to vertically or between spatially connected
6
Kinnucan and Forker, 1987; Goletti, Ahmed &Farid, 1995; Parsley and Wei, 1996; Dawson and
Dey, 2002; Kaabia et al., 2002; Rapsomanikis et al., 2003; Goodwin and Holt, 2006; Weber and
Lee, 2006; Trung et al., 2007, Baulch 2008; Bakucs et al., 2013.
93 Mahniya Zafar, Naved Hamid and Fatima Arshad
distinct markets (Jena, 2016), whereas price transmission is the extent to
which market shocks are transmitted up and down in the marketing chain
(Goodwin, 2006). The degree to which a price shock in one market affects a
price in another market can indicate whether efficient arbitrage exists
between the two markets (Rapsomanikis et al., 2004).
Different authors have explained price transmission through two
ways: 1) on the basis of the concept of the Law of One Price (Baffes, 1991;
Yang et al., 2000) and 2) in terms of market integration, an approach that
has been far more commonly used
7
. The Law of One Price (LOP) states,
“In markets linked by trade and arbitrage, homogeneous goods will have
a unique price, when expressed in the same currency, net of transaction
costs” (Ibid, p. 83). Under market integration a further division that can
be made is the extent of spatial and vertical market integration. “Spatial
market integration refers to co-movement of prices, and more generally,
to the smooth transmission of price signals and information across
spatially separated markets” (Goletti et al., 1995). It implies that the
difference between prices in different marketplaces will never exceed
transaction costs (Listorti & Esposti, 2012). Vertical price transmission
means movement of price along the supply chain from the consumer to
the producer level (Rapsomanikis et al., 2004).
Studies on spatial market integration show how regional markets
are linked using data on agricultural products. In the case of markets for
cereals in developing countries, generally the evidence is of strong or
perfect spatial integration (Dawson and Dey, 2002 for Bangladesh; Ghosh,
2003, Makama et al., 2016 for India; Zahid et al., 2007 for Pakistan; Baulch,
2008 for Vietnam); but in a few cases only weak evidence was found
(Trung et al., 2007 for Vietnam). There were only a few studies on
vegetables or fruit markets, but in those as well there is evidence of
strong spatial integration (Ramadas et al, 2014, KC and Rajalaxmi, 2019,
bothfor India).
Speaking of vertical price transmission, studies on vegetable markets
found stable long run relationship between prices either between producer
7
Ravallion, 1986; Palaskas and Harriss 1993; Gardner & Brooks, 1994; Baulch 1997; Dawson &Dey,
2002; Kaabia et al., 2002; Rapsomanikis et al., 2003; Ghosh, 2003; Weber & Lee, 2006; Trung et al.,
2007; Zahid et al., 2007; Baulch, 2008; Bakus, 2013; Ramadas et al., 2014; Paul et al., 2015; Tadesse,
2016; Kharin et al., 2017; Usman & Haile, 2017; KC &Rajalaxmi, 2019; Ozturk, 2020.
Technical Efficiency in Punjab's Agricultural Markets
94
and consumer markets (Tadesse, 2016 for Ethiopia) or between export and
domestic markets (Paul et al., 2015, for India). A few studies on cereals and
grains market found weak evidence for vertical price transmission in the
long run among domestic markets and domestic and international markets
(Usman and Haile, 2017 for Ethiopia; Ozturk, 2020, for Turkey). A meta-
analysis of European agriculture found that vertical price transmission is
asymmetric in both the long and shortrun (Bakus, 2013). Studies on meat
and dairy markets found vertical integration between the farm, wholesale
and retail markets in the long run and full transmission of all supply and
demand shocks to prices prevalent in the system (Kaabia et al., 2002, for
Spain; Kharin et al., 2017, for Slovakia). The studies explaining price
transmission base their results on monthly price data.
Studies on a variety of goods, both agriculture and non-agriculture,
using quarterly data of prices, found that vertical price convergence takes
place faster for tradable goods than for non-tradable goods (Yazgan and
Yilmazkuday, 2011; Parsley and Wei, 1996, all for USA).
Studies on price transmission and market integration use
numerous time series techniques. Techniques such as vector auto
regressive and error correction models have become the standard
instruments for investigating market relationships (Jena, 2016). While
vector auto regressive (VAR) models check for size and speed of price
adjustment among markets (Rapsomanikis et al., 2004), vector error
correction (VECM) models check for long-run relationships mainly
through the estimation of cointegration
8
among price series (Maitra,
2019). Both methods are used commonly in literature: Dawson and Dey,
2002, VAR; Ramadas et al., 2004, VAR; Baulch, 2008; Zahid et al., 2007;
Trung et al., 2007; Tadesse, 2016; Usman and Haile, 2017; KC and
Rajalaxmi, 2019; Ozturk, 2020, all use VECM.
According to Rapsomanikis et al. (2004), a commonly used
method to estimate causality between prices is the Granger causality test.
It provides information on which direction, if any, price transmission is
8
Cointegration implies the theoretical notion of a long run equilibrium relationship. If two price series
are cointegrated, there is a trend of co-movement in the long run given their linear relationship. In the
short run, the prices may vary, as shocks in one market may not be immediately transmitted to other
markets or due to transportation delays, however, arbitration prospects confirm that these deviations
from the long run equilibrium relationship are temporary” (Rapsomanikis et al., 2004, p. 58).
95 Mahniya Zafar, Naved Hamid and Fatima Arshad
occurring between two series. If two markets were integrated, the price
in one market would generally Granger-cause the price in the other
market and vice versa. Two price series may deviate from one another
because of factors such as transaction costs and yet Granger causality
may exist since some price signals may be transmitted from one market to
the other. However, lack of Granger causality may not indicate an
absence of transmission since price signals may be transmitted
immediately under special conditions. Causality tests commonly use
post-market integration estimation, as in the following studies: Blank and
Schmiesing, 1988; Baulch, 2008; Nazlioglu, 2011; Beag and Singla, 2014.
3. Data and Methodology
Data
Our research contributes to the literature by using a unique data
set, the Agriculture Management Information System (AMIS)
9
, that has
daily price information of crops for the years 2010-17. This research will
bridge the gap in literature by carrying out the following analysis using
time-series economic modeling on the crop subsector including cereals
(rice), fruits (mango and citrus) and vegetables (onion and potato): 1)
checking for market integration through horizontal and vertical price
transmission; that is, firstly whether price signals are being transmitted
between production centers, and secondly whether price signals are being
transferred from the consumer center to the producers and vice versa,
respectively, using daily price data; and 2) understanding market
efficiency, mainly by analyzing the speed at which horizontal and vertical
price transmission takes places among the markets for the above
mentioned crops.
The Agriculture Management Information System (AMIS) data set
provides district-wise daily wholesale market price information.
However, since AMIS reports price data only for districts in Punjab, the
analysis unfortunately has to be restricted to this one province.
Restricting the analysis to Punjab does not invalidate our results since
agricultural marketing is a provincial subject and Punjab accounts for
over 75 percent of Pakistan’s production of 4 out of the 5 selected crops
9
http://amis.pk/
Technical Efficiency in Punjab's Agricultural Markets
96
(see table 1) and 53 percent of Pakistan’s population, i.e., 110 million out
of 208 million (Pakistan Bureau of Statistics
10
, 2018). But it needs to be
noted that the conclusions with regards to integration of agriculture
markets of this analysis may not be fully applicable to rest of Pakistan,
particularly as both agriculture markets and transport infrastructure in
the Punjab are more developed than in the other the three provinces.
Table 1: Provincial Shares in Total Pakistan Production (000 ' tonnes)
for 2016-2017
Punjab
Sindh
Baluchistan
Khyber
Pakhtunkhwa
Pakistan
Punjab's %
share in Total
Production
3660
6
22
143
3831
96
370
748
532
184
1833
20
1375
405
1
3
1784
77
2117
26
7
30
2180
97
2524
78
95
42
2739
92
Source: Agricultural Statistics of Pakistan 2017-18 (2019), Ministry of National Food
Security & Research Islamabad.
In the analysis, we look at crops in three categories of agricultural
produce: cereals (rice), fruits (kinnow
11
and mangos) and vegetables
(onions and potatoes). These items are selected because these are
important crops in each category, both with regards to the country’s
agricultural production and exports (Ministry of National Food Security
and Research, 2019a) and the regularity of data reported in the AMIS
system. Even though wheat is the most important crop in Pakistan, it has
not been included in our analysis because the government intervenes in
the wheat market through a minimum support price (MSP) policy, which
would bias any analysis of the market price data.
12
10
http://www.pbs.gov.pk/content/provisional-summary-results-6th-population-and-housing-census-
2017-0
11
Kinnow which is similar to a mandarin orange is the dominant form of citrus gown in Pakistan
12
Under the support price program, the government usually announces a MSP in November,
procures a substantial share of the output during the harvest period (April to June) and releases it to
the flour mills during the lean season (December to March).
97 Mahniya Zafar, Naved Hamid and Fatima Arshad
The daily prices of products are available for an 8-year period from
2010 to 2017
13
. The data provides market price values for weekdays only and
the reported prices of rice, mangos, onions and potatoes are per 100 kg while
those for kinnow are per 100 pieces. For the purpose of analysis the main
consumer district for all crops is taken as Lahore, which in 2017 had a
population of over 11 million, i.e., 27.5 percent of Punjab’s urban population,
while five districts with the highest production of the selected crops in the
Punjab (for the year 2016-2017) are chosen as the producer districts for that
crop
14
. Some limitations of the data are: i) price data is not available for all
the selected districts and, ii) price data is available primarily for the months
in the harvest period but there are still missing values for some days within
the harvest period. The harvest period for purposes of analysis is taken as:
basmati rice, September-October; kinnow, January-March; mango, July-
August; potatoes, January-February, April-May, August and October;
onions, May-June, August and November-December (Pakistan Bureau of
Statistics, 2016). Missing values, up to a maximum of two days, have been
replaced with an average of the previous two days.
Methodology
The two dimensions of price transmission that will be discussed
for each crop are vertical and horizontal (or spatial) price transmission.
While discussing vertical transmission, we will try to understand the
linkage between the prices in the main consumer market and the largest
producer markets (up to a maximum of five) for each crop. For
horizontal price transmission, the extent of integration that exists within
the producer markets will be discussed. Price changes in the producer
markets selected for each crop are analyzed to see whether there is any
visible direction of transmission of price signals among the producers. In
the discussion, the selected producer markets are considered as clusters if
they are spatially close to each other.
Each crop is analyzed as follows. We conduct Granger causality
tests to see the causality of the relationship that exists between market
13
Data for mango, rice, potatoes and onions is from 2010-2017, while in the case of kinnow price
data for 2014 is unavailable.
14
If price data is not available for a particular district, the district with the next largest production
of that crop in the Punjab is selected. However, for selection of producer districts a minimum 5
percent of Punjab’s production of the crop rule is applied, and as a result for some crops there may
be fewer than 5 districts included in the analysis.
Technical Efficiency in Punjab's Agricultural Markets
98
prices in both horizontal and vertical frameworks. In order to carry out the
Granger causality tests, the following steps have to be executed. Each
market is checked for its order of integration using the Augmented Dickey-
Fuller (ADF) test for each series of prices. ADF tests the null hypothesis
that a unit root exists and if this is rejected, the series is said to be stationary
(Elliott et al., 1996). In the case that the pair of series (both while examining
vertical and horizontal price transmission) are found to be Integrated of
order 0, I(0), we conclude that the series are not cointegrated and use a
VAR framework to check for size and speed of the price adjustment among
markets (Rapsomanikis et al., 2004). We then test for Granger causality
within a VAR framework to assess vertical and horizontal price
transmission. If the market pairs are integrated at order 1, I(1), they may be
cointegrated and several tests are conducted to check for that
15
. Once the
cointegration of markets has been determined, the series are tested for
Granger causality. If the series are cointegrated, a VECM is usually
estimated, and if they are not cointegrated, a VAR model is estimated and
then checked for Granger causality. Cointegration itself cannot be used to
make conclusions about the direction of causation between prices therefore
causality tests are necessary (Ibid, 2004). Since there are missing values in
our data, a VECM model could not be estimated. Therefore, only a VAR
model is run to estimate the integration among markets.
4. Results
In this section, the discussion of each of the five selected crops is
organized as follows: first, we discuss the nature of the crop, i.e.,
production, shelf life/storage, importance of exports or imports, etc.;
second, we analyze the horizontal (spatial) price transmission among
producer districts; and finally we look at the vertical price transmission
between consumer and the producer markets.
15
“The Johansen test is used to assess the pair-wise co-integration rank of producer-consumer
markets. The cointegrating rank (r) is determined based on acceptance/rejection of null and
alternative hypotheses. Next, cointegration is tested using the Two-Way Engle Granger Approach.
This involves testing the cointegration of two markets based on the fact that deviations from
equilibrium condition of two non-stationary variables should be stationary. This implies that, while
price series may wander extensively, pairs should not diverge from one another in the long-run”
(Rapsomanikis, 2004, p. 59).
99 Mahniya Zafar, Naved Hamid and Fatima Arshad
Potato
Potato is an important and expanding vegetable crop in Pakistan
with an area and production of 178 thousand hectares and 3,831 thousand
tons, respectively, in 2016 (Ministry of National Food Security and
Research, 2019). Punjab province is the leading potato producer with a
total production of3,660thousand tons (i.e., 96 percent of Pakistans total
production) followed by Khyber Pakhtunkhwa at 143 thousand tons,
Baluchistan at 22 thousand tons and Sindh at 6 thousand tons (Table 2).
Potato has three crops namely autumn (September-February), summer
(March-October) and spring (January-May), with the three contributing
70-75 percent, 15-20 percent and 7-10 percent of the total production
respectively (Trade Development Authority of Pakistan, 2010). The main
potato producing districts in Punjab are Okara, Sahiwal, Kasur,
Pakpattan and Chiniot
16
(see Table 2).
Table 2: Potatoes: Area, Production and Share by Major Producer
Districts (2016-2017)
District
Area
(in 000 hectares)
Production
(in 000 tons)
% Share of
Punjab’s Production
Okara
54.1
1269.7
34.7
Sahiwal
25.0
522.7
14.3
Kasur
19.3
429.2
11.7
Pakpattan
18.6
417.6
11.4
Chiniot
9.4
196.5
5.4
Punjab
166.4
3660.5
100
Sources: Directorate of Agriculture, Crop Reporting Services, 2018 and Ministry of
National Food Security & Research, 2019.
In Punjab, potatoes are primarily produced for sale in urban
markets and it can be safely stored up to 6 months (Arain, n.d.). The
autumn crop, in addition to feeding the instant market, is placed in cold
storage. The stored potatoes are gradually released during the lean crop
periods generally from June onward. Pakistan is an exporter of potatoes
and about 12 percent of the production is exported, with Afghanistan,
UAE and Sri Lanka being the main markets (See Table 3).
16
Chiniot is not included in the analysis because of non-availability of price data.
Technical Efficiency in Punjab's Agricultural Markets
100
Table 3: Exports of Potatoes from Pakistan
Annual
Average
Partner
Country
Quantity
(000' tons)
Trade Value
(million US $)
2015-2017
Afghanistan
166
50
2015-2017
Sri Lanka
71
12
2015-2017
United Arab Emirates
94
14
2015-2017
World
426
94
Source: UN Comtrade database, 2015-2017.
Horizontal Price Transmission in the Potato Market
To measure horizontal price transmission in the potato market, we
look at the relationship between the prices in the producer markets. It is
likely that horizontal price transmission occurs through transfer of
information rather than actual movement of the product between the
producer districts. Three of the four largest producers (Okara, Sahiwal
and Pakpattan) are relatively close to each other
17
, and this cluster
contributes over 60 percent of the total production of potatoes in Punjab.
The Granger causality results show that the largest producer, Okara,
causes a change in price in the other two producer districts in the cluster,
implying horizontal price transmission occurs in this direction (Table 4).
The other two districts in the cluster, Sahiwal and Pakpattan, have a
bidirectional relationship with each other. As far as Kasur is concerned,
Sahiwal and Pakpattan have a unidirectional relationship with it and the
two districts cause a change in price in Kasur. However, there doesn’t
seem to be any horizontal price transmission taking place between Okara
and Kasur. But as Kasur is more or less a suburb of Lahore, vertical
transmission between the largest producer (Okara) and the main
consumer market (Lahore) may be muddling the horizontal relationship
between Okara and Kasur.
17
Okara is at a distance of about 60 km and 40 km from Pakpattan and Sahiwal, respectively, while
Pakpattan is about 45 km from Sahiwal.
101 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 4: VAR Descriptive Statistics Potato Granger Causality Tests-
Horizontal Transmission
Dependent Variable in Regression
(1)
Regressor
(2)
Okara
(3)
Sahiwal
(4)
Kasur
(5)
Pakpattan
Okara
-
0.067
0.349
0.077
Sahiwal
0.171
-
0.000
0.000
Kasur
0.147
0.500
-
0.101
Pakpattan
0.829
0.000
0.000
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether lagged
values of one variable help to predict another variable. Column 1 shows the regressor while
columns 2-5 show the dependent variables. The results were computed from a VAR model
with an average of three lags and a constant term over the 2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
As far as the speed of adjustment in prices under horizontal
transmission is concerned, prices in Pakpattan adjust to prices in Okara
within two days while prices in Sahiwal adjust to prices in Okara within four
days. The two smaller producers in the cluster (Pakpattan and Sahiwal) are
well integrated among themselves as the adjustments take place within a
day in both directions. Pakpattan and Sahiwal are also well integrated with
Kasur as price adjustments are taking place within a day (see Table 5).
Technical Efficiency in Punjab's Agricultural Markets
102
Table 5: Vector Auto Regression Model Results for Price Adjustment in
the Potato Market- Horizontal Price Transmission
(1)
Regressor
(2)
Day of
Adjustment
(3)
Dependent Variable
Okara
Sahiwal
Kasur
Pakpattan
Okara
1
-
-0.097
0.098a
0.208
2
-
0.147
-
0.344**b
3
-
-0.015
-
-
4
-
0.436***
-
-
Sahiwal
1
0.009
-
0.196***
0.243***
Pakpattan
1
-0.026
0.178***
-
0.133***
Kasur
1
-0.142
0.063
0.180***
-
Source: Author's calculations using AMIS data set 2010-2017.
a: For presentation purposes, one lag was selected (according to AIC criteria) for
regression of (Okara- Kasur) therefore only day 1 adjustment coefficient is reported.
b:For presentation purposes, two lags were selected (according to AIC criteria) for
regression of (Okara-Pakpattan) therefore only day 1 and day 2 adjustment coefficients
are reported.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. That is * representing significance of
p value at 10%, ** representing significance of p value at 5% and *** representing
significance of p value at 1%.
Vertical Price Transmission in the Potato Market
To examine vertical price transmission in the potato market, we
look at the relationship between the prices in the consumer market
(Lahore) and the producer markets. Vertical transmission is a result of both
a transfer of information and the commodity (potatoes). The price of
potatoes in Lahore determines the price in the largest producer market
Okara as well as the other producers. While the consumer market drives
prices in the largest producer market, there exists a bidirectional
relationship between Lahore and the other three producers Sahiwal,
Pakpattan and Kasur, which implies that these producer markets also drive
the prices in the consumer market (See Table 6).
103 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 6: VAR Descriptive Statistics-Potato Granger Causality Tests-
Vertical Transmission
Dependent Variable in Regression
Regressor
(1)
(2)
Lahore
(3)
Okara
(4)
Sahiwal
(5)
Kasur
(6)
Pakpattan
Lahore
-
0.008
0.000
0.000
0.000
Okara
0.967
Sahiwal
0.003
Kasur
0.028
Pakpattan
0.000
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the columns 2-6 show the dependent variables. The results were
computed from a VAR model with an average of two lags and a constant term over the
2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
Looking at vertical adjustment of prices, prices in Okara, Sahiwal,
Pakpattan and Kasur adjust to prices in Lahore within a day. Even
though Sahiwal, Pakpattan and Kasur prices adjust in both directions
with Lahore, the adjustment is relatively small, i.e. less than 10 percent.
The size of the coefficients implies that 20-30 percent of the adjustment
takes place within a day from Lahore to the producer markets (except for
Kasur) thereby implying that this channel is dominant in transmitting the
price signals (Table 7). In a study conducted in Ethiopia using monthly
price data, potato producer markets were adjusting to consumer market
prices within 3.5 months and bidirectional causality was also observed
(Tadesse, 2016).
Technical Efficiency in Punjab's Agricultural Markets
104
Table 7: Vector Auto Regression Model Results for Price Adjustment in
the Potato Market- Vertical Price Transmission
(1)
Regressor
(2)
Day of
Adjustment
(3)
Dependent Variable
Lahore
Okara
Sahiwal
Kasur
Pakpattan
1
0.213***
0.245***
0.143***
0.303**
Lahore
Okara
1
0.059
Sahiwal
1
0.054**
Kasur
1
0.084***
Pakpattan
1
0.074***
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. That is * represents significance of p
value at 10%, ** represents significance of p value at 5% and *** represents significance of
p value at 1%.
To sum up, it is seen that the potato market is connected both
vertically and horizontally. Generally, most adjustment in the prices
between markets takes place within a day of the change taking place in
the other markets. Therefore, we can say that the potato market in the
Punjab is well integrated and adjustments are rapid.
Onion
The total area and production of onions was 340 thousand
hectares and 1833 thousand tons, respectively, in 2016 (Ministry of
National Food Security and Research, 2019). Sindh province is the leading
onion producer with a total production of 748 thousand tons followed by
Baluchistan at 532 thousand tons, Punjab at 370 thousand tons and
Khyber Pakhtunkhwa at 184 thousand tons (Table 1). In other words,
only 20 percent of onions are produced in the Punjab, and since it has
over 50 percent of Pakistan’s population, probably a major part of the
onions sold in its consumer markets, such as Lahore, comes from Sindh
105 Mahniya Zafar, Naved Hamid and Fatima Arshad
and Baluchistan, which produce 40 and 30 percent, respectively, of the
country’s output. The main onion producing districts in Punjab are
Khanewal, Rajanpur, Rahim Yar Khan and Bahawalpur
18
(Table 8).
Table 8: Onions: Area, Production and Share by Major Producer
Districts (2016-2017)
District
Area (in 000
hectares)
Production (in
000 tons)
% Share of
Punjab’s
Production
Khanewal
6.4
50.1
13.5
Rajanpur
1.7
35.5
9.6
Rahim Yar Khan
3.1
34.6
9.3
Bahawalpur
3.0
24.2
6.5
Punjab
42.8
370.4
100.0
Sources: Directorate of Agriculture, Crop Reporting Services, 2018 and Ministry of
National Food Security & Research, 2019.
The agro-ecological diversity in the country enables onions to be
produced almost year-round. Due to limited shelf life and absence of cold
storage facilities in the country, onions cannot be kept for an extended
period of time and have to be sold in the domestic or international markets
soon after the time of harvest (Agriculture Market Information Service,
2004). Therefore, Pakistan both exports and imports onions each year, with
the two quantities being about the same, i.e., equivalent to about 6 percent
of its production. Exports are primarily to UAE, Malaysia and Afghanistan
while imports are almost entirely from Afghanistan and China (See Table 9).
Imports from Afghanistan probably largely supply Khyber Pakhtunkhwa
and the northern /central districts of Punjab, including Lahore.
18
No other district in the Punjab produces as much as 5 percent of Punjab’s output therefore only 4
producer districts are included in the horizontal (spatial) price transmission analysis.
Technical Efficiency in Punjab's Agricultural Markets
106
Table 9: Export of Onions from Pakistan
Annual
Partner
Quantity
Trade Value
Average
Country
(000' tons)
(million US $)
2015-2017
United Arab Emirates
34
6.3
2015-2017
Afghanistan
13
5
2015-2017
Malaysia
33
7.2
2015-2017
World
121
27.5
Import of Onions from Pakistan
2015-2017
Afghanistan
77
13.7
2015-2017
China
38
70.8
2015-2017
World
118
87.3
Source: UN Comtrade database, 2015-2017.
Horizontal Price Transmission in the Onion Market
To measure horizontal price transmission in the onion market, we
look at the relationship between the prices in the producer markets. The
selected producer districts. i.e., Khanewal, Rajanpur, Rahim Yar Khan,
Bahawalpur and Lodhran are all in southern Punjab and lie along the road
links from Sindh/Baluchistan to Lahore. The Granger causality results show
a bidirectional relation between all producers. This shows that horizontal
price transmission occurs well across all the producer districts, implying
that each producer causes a change in price in the other producer within the
southern Punjab cluster (see Table 10).
107 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 10: VAR Descriptive Statistics- Onion Granger Causality Tests-
Horizontal Price Transmission
Dependent Variable in Regression
Regressor
(1)
(2)
Khanewal
(3)
Rajanpur
(4)
Rahim Yar Khan
(5)
Bahawalpur
Khanewal
-
0
0.002
0
Rajanpur
0
-
0
0.627
Rahim Yar Khan
0
0.006
-
0
Bahawalpur
0
0
0.011
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-6) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
As far as the adjustment period in prices under horizontal
transmission is concerned, one-day adjustments are found among all
markets. This implies that markets are well integrated, as rapid
transmission of price information among all producer markets is found.
Khanewal being the largest producer also efficiently adjusts to prices of
other smaller producers and vice versa (see Table 11).
Technical Efficiency in Punjab's Agricultural Markets
108
Table 11: Vector Auto Regression Model Results for Price Adjustment
in the Onion Market- Horizontal Price Transmission
(1)
Regressor
(2)
Day of
Adjustment
(3)
Dependent Variable
Khanewal
RajanPur
Rahim Yar Khan
Bahawalpur
Khanewal
1
-
0.185***
0.185***
0.166***
RajanPur
1
0.279***
-
0.229***
0.261***
Rahim Yar Khan
1
0.406***
0.129**
-
0.190***
Bahawalpur
1
0.279***
0.044
0.119*
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. That is * represents significance of p
value at 10%, ** represents significance of p value at 5% and *** represents significance of
p value at 1%.
Vertical Price Transmission in the Onion Market
To examine vertical price transmission in the onion market, we look
at the relationship between the prices in the consumer market (Lahore) and
the producer markets. Vertical transmission is a result of both a transfer of
information and the commodity (onion). The results show that the price of
onions in Lahore determines the price in all producer markets and, at the
same time, all producers (except Bahawalpur) determine the price in the
Lahore. This means all producers are well connected with the consumer
market and vice versa (see Table 12).
109 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 12: VAR Descriptive Statistics- Onion Granger Causality Tests-
Vertical Price Transmission
Dependent Variable in Regression
Regressor
(1)
(2)
Lahore
(3)
Khanewal
(4)
RajanPur
(5)
Rahim Yar Khan
(6)
Bahawalpur
Lahore
-
0.011
0.000
0.000
0.000
Khanewal
0.000
RajanPur
0.000
Rahim Yar Khan
0.000
Bahawalpur
0.166
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-7) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
The onion market is well connected as adjustment in the prices
among markets takes place within a day of the change that takes place in
price. The onion market is therefore is well integrated and adjustments
are rapid. Within a day, the price in Lahore (the consumer market)
adjusts to prices in a producer market and vice versa. However, the size
of the coefficient in most districts (with exception of Rahim Yar Khan and
Bahawalpur) shows a larger effect (of 50-70 percent of the adjustment) in
the direction of producer markets impacting the consumer market. This
implies a supply driven effect showing that the price is set in the
producer market and that in turn determines the price in the consumer
market the next day (See Table 13).
Technical Efficiency in Punjab's Agricultural Markets
110
Table 13: Vector Auto Regression Model Results for Price Adjustment
in the Onion Market- Vertical Price Transmission
(1)
Regressor
(2)
Day of
Adjustment
(3)
Dependent Variable
Khanewal
Rajanpur
Rahim Yar Khan
Bahawalpur
Lahore
1
0.064**
0.044**
0.303***
0.296***
Lahore
Khanewal
1
0.549***
RajanPur
1
0.705***
Rahim Yar Khan
1
0.166***
Bahawalpur
1
-0.03
Bahawalpur
2
0.142**
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. That is * represents significance of p
value at 10%, ** represents significance of p value at 5% and *** represents significance of
p value at 1%.
To sum up, it is seen that the onion market is also well connected
both vertically and horizontally. Generally, most adjustment in the prices
between markets takes place within a day of the change taking place in
the other markets. Therefore, we can say that the onion market in the
Punjab is well integrated and adjustments are rapid.
Mango
The total area and production of mangos was 419 thousand
hectares and 1784 thousand tons, respectively, in 2016 (Ministry of
National Food Security and Research, 2019). Mango is the second largest
fruit produced in Pakistan after citrus. Punjab produces 77 percent of
Pakistan’s total mango output, while the rest is largely produced in Sindh
(Table 2). More than 200 varieties of mangoes are cultivated in Pakistan.
Sindhri (primarily in Sindh) and Chaunsa (primarily in the Punjab) are
the most famous varieties in the country (Javed et.al, 2012). Because of its
111 Mahniya Zafar, Naved Hamid and Fatima Arshad
dominance among the varieties in the Punjab and the non-availability of
data for other varieties our analysis is based on the Chaunsa variety. The
main mango producing districts in Punjab are Multan, Rahim Yar Khan,
Khanewal and Muzaffargarh
19
. Area, production and percentage of
Punjabs output produced in each district can be seen in Table 14.
Table 14: Mango: Area, Production and Share by Major Producer
Districts (2016-2017)
District
Area (in 000
hectares)
Production (in
000 tons)
% share of
Punjab’s
Production
Multan
31
420
31
Rahim Yar Khan
24
310
23
Khanewal
14
180
13
Muzaffargarh
19
269
7
Punjab
106
1,375
100
Sources: Directorate of Agriculture, Crop Reporting Services, 2018 and Ministry of
National Food Security & Research, 2019.
Mangoes have a extremely short shelf life, which is measured in
days rather than in weeks and this has implications for the direction of
vertical price transmission. Also, as mangoes are highly perishable, they
are exported by air (Baloch et.al, 2011). Pakistan has a very weak system for
managing the cool chain for effective transportation of fresh mangoes from
producers to the airports as well for meeting the international
phytosanitary standards for export of fresh fruits and therefore less than 5
percent of the mango crop is exported. Pakistan mainly exports mangoes to
United Arab Emirates, United Kingdom and Saudi Arabia (See Table 15).
19
As price data is not available for Muzaffargarh, it is not included in the analysis. Also as no other
district in the Punjab produces as much as 5 percent of Punjab’s output, only 3 producer districts
are included in the horizontal (spatial) price transmission analysis.
Technical Efficiency in Punjab's Agricultural Markets
112
Table 15: Exports of Mango Crop from Pakistan
Annual
Average
Partner
Country
Quantity
(000' tonnes)
Trade Value
(US $)
2015-2017
United Arab Emirates
26
19.3
2015-2017
United Kingdom
7
9.9
2015-2017
Saudi Arabia
4
5.1
2015-2017
World
54
51.0
Source: UN Comtrade database, 2015-2017.
Horizontal Price Transmission in the Mango Market
To measure horizontal price transmission in the mango market,
we look at the relationship between the prices in the producer markets.
Horizontal transmission occurs by means of a transfer of information. In
the mango market, the large producers are all in southern Punjab with
two of them being fairly close to each other (Multan and Khanewal),
together contributing 44 percent of total production in Punjab. The largest
producer Multan causes a change in the price of Rahim Yar Khan
whereas Khanewal causes a change in price in Multan. Khanewal and
Rahim Yar Khan Granger-cause changes in price within their markets,
implying a bi-directional relationship (see Table 16).
Table 16: VAR Descriptive Statistics MangoGranger Causality Tests-
Horizontal Transmission
Dependent Variable in Regression
(1)
Regressor
(2)
Multan
(3)
Rahim Yar Khan
(4)
Khanewal
Multan
-
0.030
0.505
Rahim Yar Khan
0.227
-
0.002
Khanewal
0.084
0.000
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-4) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017-sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
113 Mahniya Zafar, Naved Hamid and Fatima Arshad
As far as the time period of adjustment in prices under horizontal
transmission is concerned, prices in Rahim Yar Khan adjust to prices in
Multan within 3 days while Khanewal causes a change in price in Multan
within 2 days. Khanewal and Rahim Yar Khan both adjust to each other’s
prices. However, the mechanism is such that Multan causes a change in
price in Rahim Yar Khan, and Khanewal causes a change in price in both
Multan and Rahim Yar Khan (See Table 17). Despite being a smaller
producer, Khanewal plays a more central role in horizontal transmission
of prices. A possible explanation maybe that Khanewal, as one of new
settlements at time when the British developed the canal colonies in the
Punjab, has always been agriculturally the most progressive district in
Southern Punjab and probably seen as the trend-setter by other districts
in the region.
Table 17: Vector Auto Regression Model Results for Price Adjustment
in the Mango Market- Horizontal Price Transmission
(1)
Regressor
(2)
Day of
Adjustment
(3)
Dependent Variable
Multan
Rahim Yar Khan
Khanewal
Multan
1
-
-0.027
0.087
2
-
0.112
0.075
3
-
0.210**
-
Rahim Yar Khan
1
0.034
-
-0.038
2
-0.008
-
0.0699
3
0.006
-
0.107*
Khanewal
1
-0.043
0.104*
-
2
0.063**
-0.053
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017. Column 1 shows the regressor, Column 2 shows the day of adjustment and
column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. That is * represents significance of p
value at 10%, ** represents significance of p value at 5% and *** represents significance of
p value at 1%.
Technical Efficiency in Punjab's Agricultural Markets
114
Vertical Price Transmission in the Mango Market
To examine vertical price transmission in the mango market, we
look at the relationship between the prices in the consumer market
(Lahore) and the producer markets. Vertical transmission is a result of
both a transfer of information and the commodity (mango). Granger
causality tests help shows a unidirectional relationship between Lahore
and Multan where the price of mangoes in Lahore determines the price in
the largest producer market, Multan. The fixed supply of mangoes at any
time and their short-shelf life means that the largest producer (Multan)
has to be the price taker. Khanewal being virtually a suburb of Multan,
which is also a big city, has the option to ship the mangoes either to
Lahore or Multan. Thus, the decision by Khanewal producers whether to
sell in Lahore or Multan has an impact on the prices in the former. As a
result, in the case of Khanewal, the direction of the price signal seems to
be from the producer to the consumer market
20
. The relationship between
Lahore and Rahim Yar Khan (which is the 2nd largest producer in the
Punjab) is bidirectional. The reason for this maybe that the distance
between Lahore and Rahim Yar Khan is almost 600 km, i.e., about the
same as its distance from Karachi (population over 15 million) in Sindh
and therefore producers in Rahim Yar Khan could choose whether to
send mangoes to Lahore or Karachi based on the prices in the two cities
and their decision in turn would impact on the prices of mangoes in the
two cities (See Table 18).
20
It is also likely that, as discussed for horizontal transmission, Khanewal’s position as a price
setter for Lahore is because of its central role in Southern Punjab.
115 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 18: VAR Descriptive Statistics- Mango Granger Causality Tests-
Vertical Price Transmission
Dependent Variable in Regression
(1)
Regressor
(2)
Lahore
(3)
Multan
(4)
Rahim Yar Khan
(5)
Khanewal
Lahore
0.036
0.016
0.488
Multan
0.104
Rahim Yar Khan
0.001
Khanewal
0.012
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-5) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
When looking at adjustment of prices between consumer and
producer markets, prices in Lahore determine prices in Multan and vice
versa, after three days. However, the size of the coefficient shows that the
dominant transmission channel is also an adjustment in prices from the
producer to the consumer market. The relationship between Lahore and
Rahim Yar Khan is significant in both directions within 5 days; however,
the size of coefficient shows that the dominant transmission channel is an
adjustment in prices from the consumer to the producer market.
Khanewal, on the other hand, has the most rapid adjustment mechanism
and prices in Lahore adjust to prices in Khanewal within a day (see Table
19). In India, mango markets are found to be well integrated where
adjustments take place within a month of the change that takes place in
prices (Pardhi, 2016).
Technical Efficiency in Punjab's Agricultural Markets
116
Table 19: Vector Auto Regression Model Results for Price Adjustment
in the Mango Market- Vertical Price Transmission
(1)
Regressor
(2)
Day of Adjustment
(3)
Dependent Variable
Lahore
Multan
Rahim Yar Khan
Khanewal
1
-0.044
0.054
0.097
2
-0.067
0.054
0.012
3
0.087*
-0.014
-0.061
4
0.04
-0.124
0.006
5
0.117*
-0.0001
Lahore
Multan
1
0.075
2
0.053
3
0.173**
Rahim Yar Khan
1
0.767
2
0.005
3
0.007
4
-0.052
5
0.093**
Khanewal
1
0.010**
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. The * represents significance of p
value at 10%, ** represents significance of p value at 5% and *** represents significance of
p value at 1%.
Kinnow
Kinnow (mandarin) is one of the most important fruit crops in
Pakistan with a total area and production of 206 thousand hectares and
2180 thousand tons, respectively, in 2016 (Ministry of National Food
Security and Research, 2019). Citrus is almost entirely grown in the
Punjab with 97 percent of total production occurring in this province
(Table 2). The peak kinnow harvesting months are January to March. The
main kinnow producing districts in Punjab are Sargodha, Toba Tek
117 Mahniya Zafar, Naved Hamid and Fatima Arshad
Singh
21
, Mandi Bahauddin, and Khanewal. Area, production and
percentage produced in each district can be seen in Table 20.
Table 20: Kinnow: Area, Production and Share by Major Producer
Districts (2016-2017)
District
Area
(in 000
hectares)
Production
(in 000 tons)
% Share of
Punjab’s
Production
Sargodha
83
1,077
56
Toba Tek Singh
12
215
11
Mandi Bahauddin
9
116
6
Punjab
150
1917
100
Sources: Directorate of Agriculture, Crop Reporting Services, 2018 and Ministry of
National Food Security & Research, 2019.
Being a non-climacteric fruit, kinnow without treatment has a low
shelf life even in cold storage and may lose its quality because of some
physiochemical changes (Haider et. al, 2017). In the 1990s, the adoption of
a new technology, imported from Italy, for waxing of kinnow upon
harvesting greatly extended its shelf-like and that initiated the era of
kinnow exports for Pakistan. Currently, over 20 percent of the production
is exported with main markets Afghanistan (and possibly onward to
other Central Asian Countries), the Russian Federation and United Arab
Emirates (see Table 21).
Table 21: Exports of Kinnow Crop from Pakistan
Annual Average
Partner
Quantity (000'
tons)
Trade Value (US
$)
2015-2017
Afghanistan
162
62
2015-2017
Russian Federation
85
45
2015-2017
United Arab Emirates
46
17
2015-2017
World
389
165
Source: UN Comtrade database, 2015-2017.
21
As price data is not available for Toba Tek Singh and Mandi Bahauddin, they could not be
included in the analysis. Thus, as no other district in the Punjab produces as much as 5 percent of
Punjab’s output, we only one producer district, i.e. Sargodha, and therefore no horizontal (spatial)
price transmission analysis is carried out.
Technical Efficiency in Punjab's Agricultural Markets
118
Vertical Price Transmission in the Kinnow Market
To examine vertical price transmission in the kinnow market, we
look at the relationship between the prices in the consumer market
(Lahore) and the producer market (Sargodha). Vertical transmission is a
result of both a transfer of information and the commodity (kinnow). The
Granger causality test shows a unidirectional relationship between
producer and consumer market. The price of kinnow in Lahore
determines the price of kinnow in Sargodha. The opposite channel is also
significant at around 10 percent significance level implying that there is
bi-directional relationship (see Table 22).
Table 22: VAR Descriptive Statistics- Kinnow Granger Causality Tests-
Vertical Transmission
Dependent Variable in Regression
(1)
Regressor
(2)
Lahore
(3)
Sargodha
Lahore
-
0.005
Sargodha
0.101
-
Source: Author's calculations using AMIS data set 2010-2017 (excluding 2014).
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-5) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017 (excluding 2014) sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
We measured the speed of adjustment of prices between markets as
three days. But the size of the coefficient shows that the adjustment of
prices from the producer, Sargodha, to Lahore dominates, i.e. Sargodha is
the price setter (see Table 23). The reason for this may be that the price in
Sargodha is determined by demand and prices in the export markets. Also
as kinnow’s shelf-life is significantly extended by processing and storage,
the sellers are not in a hurry to sell in the local market because the sellers
know that any kinnows in storage that they are unable to export can
always be sold in the local market in the off-season at a premium.
119 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 23: Vector Auto Regression Model Results for Price Adjustment
in the Kinnow Market- Vertical Price Transmission
(1)
Regressor
(2)
Day of Adjustment
(3)
Dependent Variable
Sargodha
Lahore
1
-0.002
2
-0.018
3
0.03*
Lahore
Sargodha
1
0.113
2
0.369
3
0.532**
Source: Author's calculations using AMIS data set 2010-2017 (excluding 2014).
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 (excluding 2014) sample period. Column 1 shows the regressor, Column 2
shows the day of adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. A * represents significance of p value at
10%, ** represent significance of p value at 5% and *** represent significance of p value at 1%.
Rice
The total area and production of rice was 4291 thousand hectares
and 6849 thousand tons, respectively, in 2016 (Ministry of National Food
Security and Research, 2019). Rice is grown primarily in Sindh and
Punjab, with 64 percent being produced in the latter. Basmati rice is the
most famous of the rice varieties grown in Pakistan and is known for its
flavor and quality (Gain Report- USDA Foreign Agriculture Service,
2018). Basmati rice is harvested from September to October and Punjab
produces over 90 percent of the basmati rice grown in Pakistan (Table 2).
The main rice producing districts in Punjab are Sheikhupura, Hafizabad,
Sialkot, Nankana Sahab, Gujranwala and Okara
22
. Area, production and
percentage produced in each district can be seen in Table 24.
22
Due to non-availability of price data for Sheikhupura, Hafizabad, and, Nankana Sahab, the only
markets that could be considered for this analysis are Sialkot, Gujranwala, and Okara.
Technical Efficiency in Punjab's Agricultural Markets
120
Table 24: Basmati Rice: Area, Production and Share by Major Producer
Districts (2016-2017)
District
Area
(in 000 hectares)
Production
(in 000 tons)
% Share of
Total Production
Sheikhupura
158
290
11.5
Hafizabad
104
213
8.4
Sialkot
115
206
8.2
Nankana Sahab
99
194
7.7
Gujranwala
102
184
7.3
Okara
72
143
5.6
Punjab
1353
2524
100
Sources: Directorate of Agriculture, Crop Reporting Services, 2018 and Ministry of
National Food Security & Research, 2019.
Pakistan is among top ten rice producers in the world and it
exports just under 60 percent of its rice production. Basmati rice as a
percentage of total rice exports from Pakistan is about 26 percent by value
and about 13 percent by quantity (Rice Exporters Association of Pakistan,
2015-2017, see table 25). Farmers harvest paddy, which can only be kept
for short period unless it is dried in a mill. Rice millers acquire most of
the crop, dry and polish it, and then store it. Once milled, rice can be
stored for more than a year. Rice millers are also the primary exporters of
rice, with some of the large exporters owning many rice mills spread over
the main rice growing areas. These large rice exporters also directly
market basmati rice domestically under their own brand names. Most of
the IRRI rice and some of the basmati rice sold in the domestic market is
unbranded.
121 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 25: Export of Basmati Rice as a percentage of Total Rice in Pakistan
Basmati Rice
Total Rice
Basmati Rice
percentage share
of Total Rice
Quantity
(000'
tons)
Trade
Value
(million
US$)
Quantity
(000'
tons)
Trade
Value
(million
US $)
Quantity
Trade
Value
2015
503
455
4,262
1,860
11.8
24.5
2016
480
427
3,585
1,607
13.4
26.5
2017
501
525
4,024
2,000
12.5
26.2
Average
495
469
3957
1,823
12.6
25.7
Source: Calculations based on data from the Rice Exporters Association of Pakistan,
Retrieved 10 January, 2020 from http://reap.com.pk/download/index.asp
The most important export markets for Pakistan’s rice are Kenya,
Afghanistan, China and United Arab Emirates (see Table 26).
Table 26: Exports of Rice Crop from Pakistan
Annual
Average
Partner
Quantity (000'
tons)
Trade Value
(million US $)
2015-2017
Kenya
533
209
2015-2017
United Arab Emirates
204
150
2015-2017
China
456
161
2016-2017
Afghanistan
1,298
122
2015-2017
World
3,890
1,791
Source: UN Comtrade database, 2015-2017.
Horizontal Price Transmission in the Rice Market
To measure horizontal price transmission in the rice market, we
look at the relationship between the prices in the producer markets.
Surprisingly, we find no causality in any market, in any direction. This
implies that the signals are not being transferred and it can be said that
the price transmission mechanism is weak (see Table 27). In the basmati
rice market, there are five producers together in a cluster in central
Punjab (i.e., Sheikhupura, Hafizabad, Sialkot, Nankana Sahab and
Gujranwala) and together they contribute 43 percent of total production
Technical Efficiency in Punjab's Agricultural Markets
122
in Punjab. Unfortunately, we have price data from only for two producers
in the cluster, which does not include the top two producers, and that
limits the usefulness of the analysis. But finding no connectedness
between Sialkot and Gujranwala certainly indicates that horizontal price
transmission is weak at best. The probable explanation is in the nature of
the rice crop and market. A few rice millers in each district control the
market in their area and there is a lack of price competition in the market.
Also, the rice traded in the producer wholesale markets is a small
proportion of total production and is largely for local consumption. Thus,
we can conclude that the rice market in Punjab is fragmented and not
well integrated.
Table 27: VAR Descriptive Statistics- Rice Granger Causality Tests-
Horizontal Transmission
Dependent Variable in Regression
Regressor
(1)
(2)
Sialkot
(3)
Gujranwala
(4)
Okara
Sialkot
-
0.257
0.201
Gujranwala
0.51
-
0.888
Okara
0.516
0.663
-
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger-causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-4) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017-sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
As far as the period of adjustment in prices under horizontal
transmission is concerned, no adjustment of prices takes place either
within the market cluster or with Okara, reinforcing the finding that the
producer districts are not integrated with each other (See Table 28).
123 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 28: Vector Auto Regression Model Results for Price Adjustment
in the Rice Market- Horizontal Price Transmission
Regressor
(1)
(2)
Day of Adjustment
(3)
Dependent Variable
Sialkot
Gujranwala
Okara
Sialkot
1
-
0.073
-0.022
Gujranwala
1
0.062
-
0.021
Okara
1
0.082
0.07
-
Source: Author's calculations using AMIS data set 2010-2017 (excluding 2014).
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017 sample period. Column 1 shows the regressor, Column 2 shows the day of
adjustment and column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The number of * against each
coefficient shows the significance using the p values. A * represents significance of p value at
10%, ** represent significance of p value at 5% and *** represent significance of p value at 1%.
Vertical Price Transmission in the Rice Market
To examine vertical price transmission in the rice market, we look
at the relationship between the prices in the consumer market (Lahore)
and the producer markets. Prices in the two producer districts, Sialkot
and Gujranwala, Granger-cause changes in price in the consumer market,
thereby implying a supply driven effect from producers to consumers
(see Table 29).
Technical Efficiency in Punjab's Agricultural Markets
124
Table 29: VAR Descriptive Statistics-Rice Granger Causality Tests-
Vertical Price Transmission
Dependent Variable in Regression
(1)
Regressor
(2)
Lahore
(3)
Sialkot
(4)
Gujranwala
(5)
Okara
Lahore
-
0.477
0.771
0.498
Sialkot
0.000
Gujranwala
0.006
Okara
0.107
Source: Author's calculations using AMIS data set 2010-2017.
Note 1: The table shows results from Granger causality statistics that examine whether
lagged values of one variable help to predict another variable. Column 1 shows the
regressor while the remaining (2-5) show the dependent variables. The results were
computed from a VAR model with an average of three lags and a constant term over the
2010-2017 sample period.
Note 2: The entries in the columns show the p-values for F-tests. P value is a measure of
significance and it is significant at 10% if ρ<0.1.
The rice market is vertically well connected as the adjustment of
prices from Gujranwala and Sialkot to Lahore takes place in one day (see
Table 30). This, unlike the findings with regards to horizontal
transmission, corresponds to the findings for Bangladesh by Dawson and
Dey (2002) that the law of one price holds in the rice market in that
country since the rice prices in Dhaka and each regional market were so
perfectly integrated with each other that a change in price in one market
was mirrored somewhere else.
125 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 30: Vector Auto Regression Model Results for Price Adjustment
in the Rice Market- Vertical Price Transmission
(1)
Regressor
(2)
Day of Adjustment
(3)
Dependent Variable
Sialkot
Gujranwala
Okara
Lahore
1
0.083
0.033
-0.038
Lahore
Sialkot
1
0.171***
Gujranwala
1
0.144**
Okara
1
0.108
Source: Author's calculations using AMIS data set 2010-2017 (excluding 2014).
Note 1: The table shows the results from the Vector autoregressive model. It shows the
speed at which the vertical and horizontal price adjustments take place. The results were
computed from a VAR model with a minimum of one lag and a constant term over the
2010-2017. Column 1 shows the regressor, Column 2 shows the day of adjustment and
column 3 shows the dependent variables.
Note 2: The entries in the columns show the coefficients. The * against each coefficient shows
the significance using the p values. * represents significance of p value at 10%, ** represents
significance of p value at 5% and *** represents significance of p value at 1%.
Summary of Findings
We have found that potato, onion and mango markets are well
integrated both horizontally and vertically, with the speed of
adjustment generally (with the exception of mango) being very rapid
(see Table 31). Therefore, we can say that these three markets are
technically efficient. However, based on the results it is difficult to say
whether kinnow and basmati rice markets are technically efficient. The
reasons for this are two-fold.
One, in the case of both the crops a large part of the output is
exported and, therefore, it is likely that international prices of these crops
play a major role in determining the local price. But since we do not have
data on the daily international prices of these crops we are unable to
examine the extent of their integration with the world market. However,
as there are no government restrictions on their export, we expect that the
local producer markets are probably well integrated with the world and
regional markets.
Technical Efficiency in Punjab's Agricultural Markets
126
Two, for both the crops there are limitations with regards to
availability of price data for the producer markets. In the case of kinnow,
we have data for only one producer market and, therefore, it is not
possible to examine extent of horizontal integration. In the case of
basmati rice we do not have price data for the two largest producer
markets. Therefore, the result that the basmati rice market is not
integrated horizontally and price signals are not transmitted among
producer markets is subject to the caveat that for other crops, the larger
producer markets generally play a central role in horizontal transmission,
and data on these is missing in the case of rice.
As far as vertical integration is concerned, both markets seem to be
fairly well integrated. Therefore, it is likely that the markets for these two
crops are also technically efficient, but we cannot categorically say so based
on our data. With regards to efficiency of the markets, an interesting
finding is that in 40 out of the 44 relationships that have significant
causality
23
, the adjustment in prices takes place within one day. This rapid
adjustment is probably because of widespread use of mobile phones and
the resulting real-time communication of price information from one
market to another. In other words, the introduction of new communication
technologies in the last two decades has probably played an important role
in improving the technical efficiency of agricultural markets in the Punjab,
and possibly other developing countries.
23
There are 33 market-pairs for horizontal and vertical transmission combined. As we are looking
at adjustment in both directions we have a total of 66 possible results. In the case of Basmati Rice
for horizontal transmission we find no significant relationship among the 3 market pairs. Out of the
remaining 30 market pairs, in 16 the relationship is unidirectional and in 14 it is bidirectional, i.e. a
total of 44 relationships with significant causality.
127 Mahniya Zafar, Naved Hamid and Fatima Arshad
Table 31: Summary of Results
Horizontal Transmission
Crop
Degree of Integration
Degree of Adjustment
If Any Market Plays A
Central Role
(as indicated by)
(as indicated by)
Direction*
% of markets
connected**
Speed of
Adjustment+
Coefficient of
Adjustment++
Potato
Unidirectional
Medium
Rapid
Medium
Okara
Onion
Bidirectional
Strong
Rapid
Medium
Khanewal
Mango
Bidirectional
Strong
Medium
Weak
Khanewal
Rice
N.A.
N.A.
N.A.
N.A.
N.A.
Vertical Transmission
Crop
Degree of Integration
Degree of Adjustment
Market(s) Playing a
Central or Dominating
Role
(as indicated by)
(as indicated by)
Direction*
% of markets
connected**
Speed of
Adjustment+
Coefficient of
Adjustment++
Potato
Bidirectional
Strong
Rapid
Medium
Consumer (Lahore)
Onion
Bidirectional
Strong
Rapid
Strong
Producers
Mango
Bidirectional
Strong
Slow
Weak
Producers
Kinnow
Unidirectional
N.A.
Medium
Weak
Producer (Sargodha)
Rice
Unidirectional
Weak
Rapid
Medium
Producers
*Direction is explained through granger causality tests. If p value is significant in both
directions, we say it is bidirectional and if it is significant in one direction we say it is
unidirectional.
**This is determined by looking at what percentage of the total market-pairs in granger
causality tests are significant. If % <40% then Weak, if 40% to 60% then Medium, and if
>60% then Strong
+ If the significant adjustment coefficient is 1 day then Rapid, if 2-3 days then Medium,
and if 4 days then Slow
++ This is determined by looking at the coefficient of the VAR model. Only day 1
significant coefficient sizes are compared. If the significance occurs on a day later than day
1, it is considered as a weak. If the size of the coefficients (i.e. % of price adjustment taking
place on day one) for at least 50% of the sample is <10%, then weak, if 10-20% then
Medium, and if >20% then Strong.
Source: Authors’ calculations.
Another finding of our analysis is that in most cases (potato being the
exception) it is the producer markets that determine the price in the
consumer market. This is not surprising, because in the case of agricultural
products, in the very short run, we can expect market prices to be supply
driven. What is interesting is that this is not the case for potatoes, where the
price in the producer markets is determined by the consumer market.
The explanation probably lies in the nature of the different crops: In
the case of onions and mangoes, because of the short shelf-life, the fact that
Technical Efficiency in Punjab's Agricultural Markets
128
producer markets determine the price in the consumer markets is not a
reflection of producers’ market-power, but the result of the harvest size more
or less simultaneously determining prices in both producer and consumer
markets. In the case of kinnow and basmati rice, because of the longer shelf-
life and the outside option of the export market, it is an indicator of the
producers’ market-power as the sellers probably decide how much to sell in
the local market on any given day, based on the international prices and
projected demand. Finally, potatoes are somewhere in-between in the sense
that because of the use of cold-storage they have a longer shelf-life but
storage is costly and export options are few, therefore, while producers have
some market power but it is limited i.e. daily prices in the producer
markets are responsive to prices in the consumer market.
5. A Reality Check
As a test of the medium-term impact of technical efficiency of
agriculture markets in Pakistan we look at changes in cropping patterns
to see if these are in line with the changes in market demand and
government price interventions. An important determinant of demand
for agricultural products in a country is income levels and distribution:
according to the World Bank (2016) [i]n Pakistan, the reduction in
poverty led to an increase in dietary diversity for all income groups. For
the poorest, the share of expenditure devoted to milk and milk products,
chicken, eggs and fish rose, as did the share devoted to vegetables and
fruits. In contrast, the share of cereals and pulses, which provide the
cheapest calories, declined steadily between FY02 and FY14”.
24
Thus, if
markets are efficient in transmitting price signals, the changing pattern of
demand should impact cropping patterns in the medium term.
25
.
Trends in cropping patterns for the period 2000 to 2014 are presented
in Table 32. It is seen that the share of the area under vegetables and fruits
increased by over 30 percent during this period, while that of pulses declined
by over 10 percent. Also, during this period, the share of area under maize,
which is the main ingredient in animal feed, particularly in the poultry
industry, increased by 8 percent and, because of the rapid adoption of hybrid
seeds, its production increased by 130 percent (Agriculture Statistics of
24
Pakistan Development Update: Making growth matter, World Bank, November 2016, pages 34-35.
25
Ignoring international trade for the moment.
129 Mahniya Zafar, Naved Hamid and Fatima Arshad
Pakistan, 2016-17).
26
Thus the changing pattern of demand has had a strong
impact on the cropping pattern.
However, contrary to what we expected, the share of the area under
wheat, the main cereal consumed in Pakistan, increased by about 5 percent
during this period. The reason probably was that the Peoples Party
Government (2008-2013) significantly increased the support price of wheat
and since then subsidies have been provided for exporting the surpluses -
wheat exports increased from 0.5 million tons per annum (1.8 percent of the
output) in 2000-2004 to 1.0 million tons per annum (3.9 percent of the output)
in 2010-2014
27
. The share of area under rice, the other important cereal crop,
also increased during this period, but rice is a major export crop and its
exports more than doubled from on average 1.6 million tons per annum (6.9
percent of the output) in 2000-2004 to 3.6 million tons per annum (12 percent
of the output) in 2010-2014
28
.
In brief, medium-term trends in cropping patterns in the post-
2000 period were in accordance with the changing pattern of domestic
(and international) demand, except in the case of wheat where the effect
of government interventions dominates. This supports the results of our
analysis that agricultural markets in the Punjab (and probably in
Pakistan) are well integrated and price signals are transmitted efficiently
between markets.
26
In the period 2000-2004 to 2010-2014 maize yields increased by a phenomenal 120 percent,
which may be compared with increases in wheat and rice yields of 17 percent and 19 percent
respectively during this period.
27
https://www.indexmundi.com/agriculture/?country=pk&commodity=wheat&graph=exports
28
https://www.indexmundi.com/agriculture/?country=pk&commodity=milled-rice&graph=exports
Technical Efficiency in Punjab's Agricultural Markets
130
Table 32: Trends in Cropping Pattern in Pakistan’s Agriculture – 2000 to
2014
(Average % share of the total cropped)
Crops
2000-2004
2005-2009
2010-2014
Wheat
36.67
36.60
38.80
Cotton
13.26
13.00
12.46
Rice
10.47
11.24
11.24
Maize
4.24
4.38
4.56
Sugarcane
4.61
4.40
4.41
Pulses
6.27
6.27
5.55
Vegetables & Fruits
4.72
6.12
6.31
Oilseeds
2.64
3.17
2.81
Fodder
11.19
10.34
9.49
Other crops
5.94
4.48
4.36
Source: Author's calculations using, Agriculture Statistics of Pakistan 2017-18, 2019, Ministry of
National Food Security and Research. http://www.mnfsr.gov.pk/pubDetails.aspx
6. Conclusion
For the development of a dynamic agriculture sector, efficiency of
agriculture markets is critical. We posited that market efficiency is best
thought of as having two elements, i.e. technical efficiency which is
measured by the extent of integration of agricultural markets and
economic efficiency for which marketing margins on aggregate, i.e. the
percentage difference in the price paid by the consumers and that
received by the farmers are the appropriate measure. We know that
economic efficiency of agricultural markets in the Punjab is probably
quite low because agricultural produce passes through many different
hands before reaching the consumer and marketing margins at each point
in the chain are high; but, due to the lack of any data on farm gate prices
we are unable to test for it. Therefore, the focus of our research has been
on determining the technical efficiency of the agricultural marketing
system in the Punjab.
The concepts of market integration and price transmission, where
market integration describes the extent to which different markets are
connected to one another, have been used in many studies to measure
market efficiency which we call technical efficiency. To determine the
extent of market integration we used the Agriculture Management
Information System (AMIS) dataset that has daily wholesale market
131 Mahniya Zafar, Naved Hamid and Fatima Arshad
prices for most crops in the Punjab for the years 2010-2017. As we had
argued that non-traditional crops are more likely to be adversely affected
by an outdated agricultural marketing system we selected four vegetable
and fruit crops (potato, onion, mango, kinnow) and one cereal (basmati
rice
29
) for analysis.
To test for strength and speed of price transmission between
agricultural markets, Granger causality tests and Vector Auto Regressive
(VAR) models were used. We found that potato, onion and mango
markets are well integrated both horizontally and vertically, with the
speed of price adjustment in most cases (mango is the exception) being
very rapid. Therefore, we can say that these three markets are technically
efficient. It is difficult, however, to categorically say that kinnow and
basmati rice markets are technically efficient, because although both
markets are vertically fairly well integrated we are unable to satisfactorily
measure the extent of horizontal integration as price data were not
available for a number of important producer markets. Also, as a reality
check, we looked at trends in the cropping pattern over the period 2000 to
2014 and found that they are in line with the changing market demand
and government price interventions.
In conclusion, we want to highlight that market integration (or
what we call technical efficiency) is not sufficient for “producers and
consumers [to] realize their potential gains” as has been argued by some
researchers (Reddy, 2012, Varela et al., 2012) and what we call economic
efficiency, is a necessary condition for these gains to be fully realized. More
research is needed in the area of economic efficiency of agriculture
markets, and in the case of Pakistan such research would be timely because
the Punjab Agriculture Marketing Regulatory Authority (PAMRA) Act
2020 has fundamentally reformed the legal and regulatory framework for
agricultural marketing in the Punjab and such a study could provide a
baseline for measuring the economic impact of the legal reforms.
29
It would have been preferable to have included wheat as the cereal crop but, because of extensive
government intervention in the market, meaningful analysis of the wheat market was not possible.
Technical Efficiency in Punjab's Agricultural Markets
132
References
Agriculture Marketing Information Service, Directorate of Agriculture.
(2004). Marketing of onion, problem and prospects. Retrieved
November 20, 2019, from http://www.amis.pk/pdf/Pulications/
Onion%20Report.pdf
Ahsan, M. (2018). Commissions and omissions: Agricultural producer
markets in Pakistan. Policy and Institutional Reforms to Improve
Horticultural Markets in Pakistan (Working Paper No. 01/18).
Retrieved October 7, 2019, from https://www.monash.edu/__data/
assets/pdf_file/0010/1428769/WP1_Rana_complete.pdf
Arain, G. (n.d.). Potato cultivation in Pakistan. Retrieved April 10, 2019,
from http://www.valleyirrigationpakistan.com/wp-content/uploads
/2012/09/Potato-Cultivation-in-Pakistan.pdf
Baffes, J. (1991). Some further evidence on the law of one price: The law of
one price still holds. American Journal of Agricultural
Economics, 73(4), 12641273. doi:10.2307/1242454
Bakus, Z., Ferto, I., and Falkowski, J. (2013). What causes asymmetric
price transmission in agro-food sector? Meta-analysis perspective.
Discussion paper. Institute of Economics, Research Centre for
Economic and Regional Studies, Hungarian Academy of Sciences
Baloch, M. K., Bibi, F., and Jilani, M. S. (2011). Quality and shelf life of
mango (Mangiferaindica L.) fruit: As affected by cooling at
harvest time. Scientia Horticulturae,130(3), 642-646. doi:10.1016/j.
scienta.2011.08.022
Barrett, C. B., & Li, J. R. (2002). Distinguishing between equilibrium and
integration in spatial price analysis. American Journal of
Agricultural Economics, 84(2), 292307. doi:10.1111/1467-8276.00298
Basmati Rice the True Pride of Pakistan. (n.d.). Retrieved January, 2020,
from https://reap.com.pk/index.php
133 Mahniya Zafar, Naved Hamid and Fatima Arshad
Baulch, B. (1997). Testing for food market integration revisited. Journal of
Development Studies,33(4), 512-534. https://doi.org/10.1080/
00220389708422479
Baulch, B., Hansen, H., Trung, L., and Tam, T. (2008). The spatial
integration of paddy markets in Vietnam. Journal of Agricultural
Economics,59(2), 271-295. https://doi.org/10.1111/j.1477-9552.2007.
00148.x
Beag, F., and Singla, N. (2014). Cointegration, causality and impulse
response analysis in major apple markets of India. Agricultural
Economics Research Review, 27(2), 289. https://doi.org/10.5958/0974-
0279.2014.00032.9
Blank, S., and Schmiesing, B. (1988). Modeling of agricultural markets and
prices using causality and path analysis. North Central Journal of
Agricultural Economics, 10(1), 35. https://doi.org/10.2307/1349234
Dawson, P. J., and Dey, P. K. (2002). Testing for the law of one price: Rice
market integration in Bangladesh. Journal of International
Development, 14(4), 473484. doi:10.1002/jid.888
Directorate of Agriculture, Crop Reporting Services. (2017). Kharif Crop
Final Estimates Data Book 2016-17. Retrieved March 9, 2018, from
https://crsagri7.pitb.gov.pk/system/files/BOOK%20RABI%20FINA
L%202016-17.pdf#overlay-context=reports
Elliott, G., Rothenberg, T. J., and Stock, J. H. (1996). Efficient tests for an
autoregressive unit root. Econometrica, 64(4), 813836.
doi:10.2307/2171846
Gardner, B. L., and Brooks, K. M. (1994). Food prices and market
integration in Russia: 199293. American Journal of Agricultural
Economics,76(3), 641646. doi:10.2307/1243681
Goletti, F., Ahmed, R., and Farid, N. (2007). Structural determinants of
market integration: the case of rice markets in Bangladesh. The
Developing Economies, 33(2), 196-198. https://doi.org/10.1111/j.1746-
1049.1995.tb00713.x
Technical Efficiency in Punjab's Agricultural Markets
134
Goodwin, B. (2006). Spatial and vertical price transmission in meat
markets. Presentation, University of Kentucky, North Carolina.
Ghosh, M. (2003). Spatial integration of wheat markets in India: Evidence
from cointegration tests. Oxford Development Studies, 31(2), 159
171. doi:10.1080/13600810307426
Haider, S.-A. T.-A., Ahmad, S., Khan, A. S., and Basra, S. M. A. (2017).
Comparison of different fruit coatings to enhance the shelf life of
kinnow mandarin. Pakistan Journal of Agricultural Sciences,54(01),
3544. doi:10.21162/pakjas/17.5317
International Labor Organization (2019). Employment in agriculture (% of
total employment). Retrieved May 1, 2020, from
https://data.worldbank.org/indicator/SL.AGR.EMPL.ZS
Jena, P. K. (2016). Commodity market integration and price transmission:
Empirical evidence from India. Theoretical and Applied
Economics,23(3), 283306.
Kaabia, B., Gill, M., Maria, J., and Boshnjaku, L. (2002). Price transmission
asymmetries in the Spanish lamb sector. Presentation. XthEAAE
Congress, Spain. doi:10.22004/ag.econ.24908
KC, G., and Rajalaxmi, A. (2019). Spatial price integration and price
transition in major markets of onion in India. Journal of
Pharmacognosy and Phytochemistry,2, 1058-1063.
Khan, M. A., and Khan, S. L. (2010). Potential markets of potato. Trade
Development Authority of Pakistan. Retrieved 11 January, 2020,
from https://www.tdap.gov.pk/doc_reports/tdap_report_on_
potential_markets_of_potatoes.pdf
Kharin, S., Lajdova, Z., and Bielik, P. (2017). Price transmission on the
Slovak dairy market. Studies in Agricultural Economics, 119(3), 148
155. doi:10.7896/j.1711
Kinnucan, H. W., and Forker, O. D. (1987). Asymmetry in farmretail
price transmission for major dairy products. American Journal of
Agricultural Economics, 69(2), 285292. doi:10.2307/1242278
135 Mahniya Zafar, Naved Hamid and Fatima Arshad
Listorti, G., and Esposti, R. (2016). Horizontal price transmission in
agricultural markets: Fundamental concepts and open empirical
issues. Bio-Based and Applied Economics, 1(1), 8196.
Li, G., and Janssen, W. (2020). Modernizing Punjab's farming to benefit
farmers and consumers. Retrieved June, 2020, from
https://blogs.worldbank.org/endpovertyinsouthasia/modernizing-
punjabs-farming-benefit-farmers-and-consumers
Maitra, S. (2019). Time series analysis using Granger’s causality and VAR model:
An example with python code. Medium Towards Data Science.
Retrieved 4 May 2020, from https://towardsdatascience.com/
granger-causality-and-vector-auto-regressive-model-for-time-series-
forecasting-3226a64889a6.
Makama, S. A., and Amruthat, T. J. (2016). Spatial market integration of
rice between India and Nigeria: A co-integration approach. Journal
of Agriculture and Veterinary Science, 9(4), 17.
Ministry of Finance. (1980-2016). Pakistan Economic Survey. Retrieved
February 23, 2018, from http://www.finance.gov.pk/rebuttals1.
html
Ministry of National Food Security and Research. (2019A). Fruit,
vegetables and condiments statistics of Pakistan 2017-18. Retrieved
June 24, 2019, from http://www.mnfsr.gov.pk/frmDetails.aspx
Ministry of National Food Security and Research. (2019). Agricultural
statistics of Pakistan 2017-18. Retrieved December 20, 2019, from
http://www.mnfsr.gov.pk/frmDetails.aspx
Nazlioglu, S. (2011). World Oil and agricultural commodity prices:
Evidence from nonlinear causality. Energy Policy, 39(5), 2935-2943.
https://doi.org/10.1016/j.enpol.2011.03.001
Nisbet, R., Elder, J., and Miner, G. (2009). Chapter 4 - Data Understanding
and Preparation. In Handbook of Statistical analysis and data mining
applications. 4975. doi: https://doi.org/10.1016/B978-0-12-374765-
5.00004-8
Technical Efficiency in Punjab's Agricultural Markets
136
Ozturk, O. (2020). Market integration and spatial price transmission in
grain markets of Turkey. Applied Economics, 52(18), 1936-1948.
https://doi.org/10.1080/00036846.2020.1726862
Pakistan Bureau of Statistics. (2016). Crop calendar for potatoes and onions.
Retrieved March 15, 2018, from http://www.pbs.gov.pk/sites/
default/files//tables/Table-14 Crop_Potato & Onion Calendar.pdf
Pakistan Bureau of Statistics. (2018). District wise census results census 2017.
Retrieved December 15, 2019, from http://www.pbs.gov.pk/sites/
default/files//tables/Table-14 Crop_Potato & Onion Calendar.pdf
Parsley, D., and Wei, S.-J. (1996). Convergence to the law of one price
without trade barriers or currency fluctuations. The Quarterly
Journal of Economics, 111(4), 12111236. doi:10.3386/w5654
Pardhi, R. M. (2016). Market integration and price forecasting of mango:
An empirical analysis. Ph.D. Thesis, Department of Agricultural
Economics, Institute of Agricultural Sciences, Banaras Hindu
University, Varanasi.
Paul, R.K., Saxena, R., Chaurasia, S., Zeeshan, M., and Rana, S. (2015).
Examining export volatility, structural breaks in price volatility
and linkages between domestic and export prices of onion in
India. Agricultural Economics Research Review, 28 (conf.), 101116.
doi:10.5958/0974-0279.2015.00026.9
Paul, U.K., Das, G., Debnath, A., and Mathur, T. (2016). Market
integration and price leadership in India’s onion market. Review of
Market Integration, 8(1-2), 4964. doi:10.1177/0974929216687881
Ramadas, S., Chandrasekaran, S., P, V., and Thomas, L. (2014). Testing
market integration and convergence to the law of one price in
Indian onions. African Journal of Agricultural Research, 9(40), 2975-
2984. https://doi.org/10.5897/ajar2013.8037
Rapsomanikis, G., Hallam, D., and Conforti, P. (2004). Market integration
and price transmission in selected food and cash crop markets of
developing countries: Review and applications. Commodity Market
Review, 5175.
137 Mahniya Zafar, Naved Hamid and Fatima Arshad
Ravallion, M. (1986). Testing market integration. American Journal of
Agricultural Economics, 68(1), 102109. doi:10.2307/1241654
Reddy, A. (2011). Market integration of grain legumes in India: The case
of the chickpea market. SAARC Journal of Agriculture, 10(2), 1129.
doi:10.3329/sja.v10i2.18320
Shahzad, M. A., Razzaq, A., and Qing, P. (2019). On the wheat support
policy in Pakistan. Journal of Economic Impact, 1(3), 8086.
Tadesse, M. (2016). Transaction costs and spatial integration of vegetable
and fruit market in Ethiopia. Ethiopian Journal of Economics, 24(1).
The World Bank. (2016). Pakistan development update- making growth matter.
Retrieved May 15, 2020, from http://documents.worldbank.org/
curated/en/935241478612633044/pdf/109961-WP-PUBLIC-disclosed-
11-9-16-5-pm-Pakistan-Development-Update-Fall-2016-with-
compressed-pics.pdf
The Punjab agricultural marketing regulatory authority (amendment)
ordinance. (2019). Retrieved October 7, 2019, from
http://www.amis.pk/Laws/Ordinance/PAMRA.pdf
Usman, M., and Haile, M. (2017). Producer to retailer price transmission
in cereal markets of Ethiopia. Food Security,9(4), 815-829.
https://doi.org/10.1007/s12571-017-0692-0
United States Department of Agriculture. (n.d.). Pakistan milled rice exports
by year. Retrieved June, 20, from https://www.indexmundi.com/
agriculture/?country=pk&commodity=milled-rice&graph=exports
United States Department of Agriculture. (n.d.). Pakistan wheat exports by
year. Retrieved June, 20, from https://www.indexmundi.com/
agriculture/?country=pk&commodity=wheat&graph=exports
Varela, G., Carroll, E., and Iacovone, L. (2013). Determinants of market
integration and price transmission in Indonesia. ASEAN Economic
Bulletin, 30(1), 19. https://doi.org/10.1355/ae30-1b
Technical Efficiency in Punjab's Agricultural Markets
138
Wohlgenant, M. (2001). Chapter 16 marketing margins: Empirical
analysis. Handbook of Agricultural Economics, 1, 933970.
doi:10.1016/s1574-0072(01)01007-6
World Development Indicators: The World Bank. (2020, April). Retrieved
May 1, 2020, from http://wdi.worldbank.org/table/4.2
The World Bank. (2016, November). Pakistan development update. 34-35.
Retrieved April 28, 2020, from http://documents.worldbank.org/
curated/en/935241478612633044/pdf/109961-WP-PUBLIC-
disclosed-11-9-16-5-pm-Pakistan-Development-Update-Fall-2016-
with-compressed-pics.pdf
UN Comtrade. International Trade Statistics Database. (n.d.). Retrieved
October 10, 2019, from https://comtrade.un.org/
United States Department of Agriculture. (2018). Global agricultural
information network (GAIN). Foreign Agriculture Service. Retrieved
20 April, 2020, from https://www.fas.usda.gov/databases/global-
agricultural-information-network-gain
Yang, J., Bessler, D. A., and Leatham, D. J. (2015). The law of one price:
Developed and developing country market integration. Journal of
Agricultural and Applied Economics. Retrieved October 28, 2019, from
https://www.cambridge.org/core/journals/journal-of-agricultural-
and-applied-economics/article/div-classtitlethe-law-of-one-price-
developed-and-developing-country-market-integrationdiv/
CE8FC05322063F13AC605137592831D2
Yazgan, E., and Yilmazkuda, H. (2011). Price-level convergence: New
evidence from U.S. Cities. Economics Letters, 110 (2),76-78
Zahid, M., Qayyum, A., and Malik, W. (2007). Dynamics of wheat market
integration in northern Punjab, Pakistan. The Pakistan Development
Review, 46(4II), 817-830. https://doi.org/10.30541/v46i4ii
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Wheat is Pakistan's main food and strategic crop. Currently, the government controls wheat prices through a minimum support price (MSP) policy to encourage production. However, despite the increase in wheat production, input costs and output prices have been increasing over the years. This paper aims to analyse the impact of wheat support price policies. We use data from different government sources to estimate the financial implications of MSP and compare the support price policies of India and Pakistan. We find that Pakistan's current minimum support price policy encourages farmers to produce larger quantities of wheat, but this places a heavy financial burden on the country's finances. Our results indicate that the higher MSP of wheat has made the country lose its competitiveness in the international market. Besides, we found that the cost of wheat production in Pakistan is much higher than in India. These higher production costs force the government to raise the MSP to maintain farmers' profitability. The high MSP is guaranteed by subsidizing the procurement and release of wheat, which imposes a heavy financial burden on government finances. In addition, the rise in wheat prices in recent years has also hurt consumers. Policymakers can redistribute subsidies by subsidizing wheat inputs, especially fertilizers and seeds, to reduce production costs. To this end, the best policy intervention may be to provide input subsidies rather than subsidies on purchase prices. A reduction in input costs will correspondingly reduce output prices, which will increase farmers' profitability, consumer surplus and the international competitiveness of Pakistani wheat.
Article
Full-text available
Despite being the second largest producer in the world, the price of onion witnesses a significant degree of temporal and spatial variation in India. Addressing this issue has become a major challenge for the policymakers of the country. The present study intends to identify the most influential onion markets in India. Johansen cointegration and vector error correction model (VECM) techniques are used to analyse the monthly wholesale price between 2005 and 2015 in 10 major markets. The results show that Hubli is the most dominant market jointly followed by Kolkata and Lasalgaon. The results exhibit the predominant role played by the non-producing markets.
Article
Full-text available
‘Kinnow’ mandarin is an important citrus cultivar, which ranks first among all the fruit crops in Pakistan. Being a nonclimacteric fruit, it has low shelflife even in cold storage, may lose its quality because of some physiochemical changes. Wax coating has been proved as an important strategy to maintain the fruit quality under storage. The present study was planned to compare the different types of waxes which include commercial waxes used by the traders and lab. prepared coatings (PHRC wax and benzaldehyde coating) by the scientists of AARI and UAF, respectively. The results revealed that sharine and paraffin wax exhibited minimum fruit rot (2%) with lower weight loss (11.87%) after 90 days of storage and PHRC wax performed statistically at par to commercial waxes regarding weight loss (12%) and fruit rot (3%). The minimum changes in juice weight, TSS/TA ratio, peel/pulp ratio, ascorbic acid and sugar contents were recorded in coated fruits except benzaldehyde coating. Wax coatings also maintained higher levels of total phenolics (240.7 mg GAE 100g⁻¹), antioxidant (63.4%) and enzyme activities of catalase (17.6), peroxidase (0.71), and superoxide dismutase (137.6 U mg⁻¹ protein) during storage. It is concluded that PHRC wax can be recommended for long term storage of citrus fruits as it performed almost at par to commercial waxes and maintained the fruit quality as well as bioactive juice components.
Article
Full-text available
In the era of globalization markets are connected to each other not only to compete domestic products with international products to increase economic stand but also export the required products. Therefore, in the recent years markets are more vibrant and transparent in integration with each other. Over the last couple of years, it has been seen that commodity prices have increased and remained highly volatile not only in international market but in India. Therefore, this paper examines the market integration and price transmission in India using the time series techniques of cointegration and vector error correction models. The study finds that there is both short-run and long-run relationship between domestic commodity index (all commodity index, energy index and metal index) prices and international commodity index (all commodity index, energy index and metal index) prices. But, there is no relationship between domestic agriculture index price and international agriculture index price. The study also finds that except the agriculture index price, the change in domestic commodity index (all commodity index, energy index and metal index) prices are due to the change in international commodity index (all commodity index, energy index and metal index) prices.
Article
This paper examines whether the food market of Turkey is cointegrated with the world food market. Using an error correction model, we analyze the response of producer prices of wheat, barley, maize, soybean and rice to changes in world market prices. Results show that the rice market of Turkey is not cointegrated with the world rice market, while the other commodity markets are weakly cointegrated. Results also show that pass-through of changes in the world prices to the domestic prices is relatively low both in the short run and in the long run, and that adjustment to the new equilibrium following a shock is slow. Government intervention policies both at the border and as domestic price supports seem to be underlying causes that have weakened the linkage between domestic and international markets. Fewer protectionist policies at the border and lower levels of government support policies are necessary to increase the domestic market integration with the international market.
Article
This paper investigates price transmission between producer and retail prices of teff, wheat, and maize in Amhara and Oromia, the two major cereal markets in Ethiopia. Market and cereal specific asymmetric error correction models were estimated to analyze producer-retail price transmission using monthly data from 2001 to 2011. For seven out of eight crop-market combinations, with the exception of the Amhara wheat market, we found no evidence of asymmetric price transmission from producer to retail prices in the long run. Neither contemporaneous nor long-run price transmission asymmetry was found in either the Amhara or Oromia teff markets. This was also the case for the maize market except that there existed a short-run asymmetric price transmission that disappeared in the long run in Oromia. We therefore conclude that there is no strong empirical evidence to support the purported ‘market power’ or ‘inventory holding’ behaviour in the Ethiopian cereal markets to explain asymmetric vertical price transmission in the long run. The evidence of asymmetric price transmission for the Amhara wheat market, unlike the wheat market in Oromia, may indicate some differential in the quality of infrastructure and the length and complexity of wheat value chains between these two markets. Symmetric price adjustments in these cereal markets suggest that input price changes may have positive long run implications for food security and welfare of the poor in Ethiopia.
Article
The study has investigated market integration across five major wholesale apple markets, viz. Ahmedabad, Bengaluru, Delhi, Hyderabad and Kolkata, of the country by adopting Johansen's multivariate cointegration approach. The study has confirmed the presence of cointegration, implying the long-run price association among the markets. To get the additional evidence as to whether and in which direction price transmission is occurring between the market pairs, Granger causality test has been used, which has confirmed Delhi to be the price-determining market. Hyderabad has been found comparatively more efficient as it has depicted most bidirectional causal relations with other markets. The market pairs: Ahmedabad — Kolkata and Bengaluru — Kolkata have not shown any causal relation between them. The impulse response functions were also conducted which have confirmed the results of cointegration and Granger causality, but the magnitude of price transmission has been found relatively low in some market pairs that are spatially integrated. The major implication of the study is for the designing of a network of agricultural wholesale markets across the country at almost equal distance from each other to enhance the market integration and better price transmission among them.