EMPIRICAL ANALYSIS OF LIVESTOCK PRODUCTIVITY THROUGH IMPROVED BREEDING IN
PUNJAB, PAKISTAN
M. Ashfaq1, R. Kousar*1, M. S. A. Makhdum 2 and J. Nasir1
1Institute of
Agricultural and Resource Economics, University of Agriculture Faisalabad
2Department of Economics, Government
College University Faisalabad
*Corresponding
author Email: rakhshanda.kousar@uaf.edu.pk
ABSTRACT
In
Pakistan, the productivity of the livestock sector is lower than its capacity
due to lack of breeds with high productive potential, nutrition deficiency and
poor disease control facilities. This study explores the potential of the
livestock sector to move towards more productive breeds of cattle. Primary data
were collected from 340 livestock farmers of Punjab Province through stratified
random sampling technique. Endogenous Switching Regression (ESR) is employed to
identify factors influencing the farmer’s decision to adopt superior/exotic
breeds of buffaloes and cows and impact of adoption on the gross margins. The
results of adoption equation reveal that farmer’s decision to adopt exotic breed
depends positively on the proximity of his house to livestock market, years of
schooling, operational land, access to hired labour and contact with extension
agents. Results also show that different socio-economic factors have
differential impact on the gross margins of both adopters and non-adopters.
There is need of policy reforms to eliminate the constraints of shifting
towards high productive breeds.
Keywords: Improved Livestock
Breed, Constraints and Impact Analysis, Endogenous Switching Regression (ESR), Punjab-Pakistan.
https://doi.org/10.36899/JAPS.2020.6.0185
Published online August 03,2020
The
livestock sector performs a key role in rural settings of Pakistan as
livelihoods of small and landless farmers depend largely on this sector. For
instance, more than 8 million families of Pakistan are involved in raising
livestock (Anonymous, 2019). In Punjab province, which is the main producer of
dairy products, more than 40 percent of the total household income comes from
dairy. Most of farm families are involved in livestock sector as small
ruminants and large animals contribute to the food basket of household (Ashfaq et
al., 2014).
Given
the vital importance of the livestock sector for economic, social, and rural
development of Pakistan, it requires a good deal of attention in the national
policy framework. Unfortunately, the importance of livestock has neither been
fully realized nor reflected in agricultural or economic policy of the country.
Besides other limiting factors, the productivity of livestock sector also
depends on genetic potential of livestock heads. It is well known that some
world-famous breeds of cattle and buffaloes are in Pakistan. The buffalo breeds
of Nili-Ravi, Aza Kheli and Kundi are also popular for their high milk
production. Nili Ravi is the best buffalo breed in world and is famous as Black
Gold of Pakistan (Iqbal et al., 2015). The main dairy cattle breeds in
the country are Sahiwal and Red Sindhi which are internationally recognized
breeds. Other than well-defined cattle breeds, there are many mixed and
crossbred cattle in this country.
Besides
our local and crossbred breeds, exotic breeds which are imported from
Australia, America and Sweden are also the important part of dairy sector in
Pakistan. These breeds are famous for high milk production and breeding
efficiency. However, they require great care. Holstein Friesian, Jersey and Friesian
X Sahiwal are some common breeds in Pakistan (Khan et al., 2008).
Although
Pakistan is endowed with rich cattle genetic diversity and resources. However,
efficient utilization and management of these resources are lacking due to the
lack of awareness and institutional weaknesses. In addition to this,
improvement in tangible breed and its utilization is limited. Only a limited
number of genetic improvement programs could get traction in government policy
because these programs traditionally take long to complete, and due to
political reasons, local governments in Pakistan are generally interested in
programs which yield quick and tangible results (Afzal and Naqvi, 2004).
Several
previous empirical studies have focused on different aspects of livestock
productivity, health and genetic potential like: animal health and dairy
productivity (Hall et al., 2004; Rushton, 2013), acceptance of breeding
and production practices and future of dairy industry (Oltenacu and Algers,
2005), sustainable diets and genetic diversity of animals (Hoffmann and
Baumung, 2013), levels of protein consumption and production of milk (Habib et
al., 2019), factors affecting livestock productivity (Lamy et al.,
2012), livestock diseases (Modisane, 2009; Otufale and Adekoya, 2012). However,
there is limited literature on the impact analysis of genetic potential on the
productivity of livestock by considering the selection bias due to observable
and unobservable factors. Failure to distinguish between the causal effects of
breed adoption and unobservable heterogeneity may lead to biased estimates and
misleading policy implications. Based on the existing research gap and within
this backdrop, the present study has focused on exploring the impact of
adopting productive animal breeds on productivity and identifying the key
constraints of adoption by taking into consideration of endogeneity and
selectivity bias.
MATERIALS AND METHODS
This
study is basically primary data based. Three districts (Sahiwal, Jhang, and
Sargodha) from Punjab province were selected based on the highest total population
of buffaloes and cattle in Punjab (PBS, 2006). A total of 340 livestock farmers
were selected by multistage random technique. The district Sahiwal has 2
tehsils, district Jhang has 4 tehsils, and district Sargodha has 7 tehsils. To
account for the geographical variations, all tehsils from districts of Sahiwal
and Jhang were selected, while 4 out of 7 tehsils were randomly selected from
district Sargodha. After the selection of tehsils, 3 villages were randomly
selected from each tehsil of district Sargodha and Jhang, while 5 villages were
randomly selected from each tehsil of district Sahiwal. The reason for
selecting more villages from district Sahiwal was the division of such a large
district into two tehsils only. A list of all of the livestock farmers in a
selected village was prepared, and 10 farmers were selected from each village
by systematic random sampling. The location of the sampling districts in Punjab
province is given in Figure 1.
Figure 1. Map of
Punjab province and the sampling districts
Livestock
farmers were categorized as small, medium, or large farmers based on their
possession of adult buffaloes and cattle. Farmers having 1-3 adult animals were
categorized as small, 4-6 as medium and those having more than 6 adult animals
were categorized as large farmers. The categorization of farmers in such a way
is also found in Moaeen and Babar (2006) and Ashfaq et al. (2014).
The
socioeconomic characteristics of farmers such as age, livestock farming
experience, education, and family size etc., are provided in Table 1.
Livestock
farms were selected by proportionate random sampling technique as most of
farmers are small (about 55 percent) in the study area. It implies that farming
community in these districts consists of small, subsistence farmers. It is also
worth noting that many farmers, in the small farm category, were landless.
The
animal inventory of farmers is presented in Table 3. Overall, farmers possessed
more buffaloes (2.31 buffaloes per farm) than adult cows (1.78 cows per farm).
The other animals are included for more context. Another interesting result is
the possession of very few bulls by all type of farms. Low possession of bulls
is because farmer consider it non-productive as its role to produce bullocks
for work has been replaces by mechanization over the years.
Nili-Ravi
is considered a superior breed of buffalo and is recognized internationally. It
is more productive, and more expensive. Kundi is another good breed of buffalo.
However, the Kundi buffaloes are not usually present in Punjab (Khan et al.,
2007). Reference to previous empirical study (Khan et al., 2007), Nili
and Ravi as separate breeds and Nile-Ravi is considered as mixed breed while no
crossing is reported between Kundi and Nili-Ravi. The Nili-Ravi breed was about
one third of total adult buffalo populations on the farms (about 34 percent),
while the Kundi breed was about one percent of the total adult buffaloes in the
Faisalabad. The rest of the buffalo population consisted of ‘mixed’. Table 4
also indicates such a wide gap of population for two breeds. Farmers of the
study area attribute such a wide gap to challenges in rearing exotic breed in
terms of their acclimatization to Punjab’s environment, poor carriage capacity
and unavailability of pedigree bulls.
The
breed diversity of cattle at livestock farms was higher than buffaloes. There
were all types of breeds ranging from indigenous to crosses of exotic breeds.
However, about 60 percent of farmers had indigenous breeds at their farms which
are less productive than exotic breeds, such as Jersey, Friesian, and crosses
of these breeds. About 30 percent of the cattle population was described as
‘non-descriptive’ breed which is an issue to be addressed. The population of
pure exotic breeds e.g. Holstein-Friesian, and Jersey was only 20 percent at
livestock farms. The population of exotic cross breeds was about 20 percent.
Farmers
were using both natural breeding and artificial insemination for breeding
purposes (Table 6). However, natural breeding was found more prevalent in
buffaloes, and the practice of artificial insemination was more prevalent for
cows. This might be due to easy availability of buffalo sires in the villages.
Also, it was revealed during the field survey that farmers were not much
concerned about breed selection for buffaloes. In case of buffaloes, only about
30 percent of farmers were using artificial insemination as a breeding source.
In the case of cattle, artificial insemination was being done by above 80
percent of farmers. However, one notable point is that the large farmers
usually went for expensive semen, and a wide range of semen quality was being
provided at various prices.
Definition
of Variables and Descriptive Statistics: The dependent variables used in
this study is the adoption of exotic breed and gross margins. The explanatory
variables include level of education, household size, labor, water frequency,
extension services, operational land, off-farm income, distance of household
from livestock market and locational dummies. The definitions and sample
statistics of the variables used in the analysis are given in Table 7.
Estimation
of Gross Margins of Livestock Farmers: The gross margins are calculated by using
the following traditional formula:
Gross Margins =
Total Revenue – Total Variable Cost (1)
Empirical
Specification: We
employed endogenous switching (ESR) model in this study in order to account for
selection bias. ESR is the extension of Heckman selection model with the
exception that Heckman method assumes outcome function would differ only by
unobservable factors between the adopters and non-adopters. ESR is a parametric
approach and accounts for selection bias caused by observable and unobservable
factors, influencing outcome. The underlying model has two separate regimes for
adopters and non-adopters. Inverse mill ratio is calculated to control for
selection bias and plugged into the outcome equations.
In
this study, farmers have two choices: to adopt exotic breeds or to adopt
indigenous breeds (not to adopt exotic breeds). They weigh their choices on the
basis of utility. The binary choice decision based on the utility from adoption
or no adoption can be specified as:
(2)
Here
depends on a
vector of variables and error term which is assumed
to have zero mean and equal variance. The error term is composed of measurement
errors and unobserved factors. This model assumes two regimes: 1, to adopt and
2, not to adopt. is the vector of
parameters to be estimated.
The
outcome equations can be formulated as:
(3a)
(3b)
Here
and represent gross
margins for adopters and non-adopters respectively. represents a
vector of individual, household and locational characteristics of respondent i.
, , and are parameters to
be estimated and andare the error
terms. and represent the
adoption decision of the farmer to adopt and not
to adopt respectively. For the ESR model to be correctly specified, Z contains
the same variables as plus at least one
suitable instrument that is correlated with the adoption decision but
uncorrelated with the outcome.
Finally,
the error terms are assumed to have a trivariate normal distribution, with zero
mean and non-singular covariance matrix expressed as:
(4)
Where,
corresponds to the variance of
the error term in the adoption equation and , represent the variance
of the error terms in the outcome equations.
According
to Maddala (1983), when there are unobservable factors associated with
selection bias, the important implication of the error structure is that as the
error term of the adoption
equation (2) is correlated with the error terms of the outcome
functions (3a) and (3b), the expected values of conditional on
the sample selection are non-zero, giving rise to selection bias. To solve this
problem, we estimated inverse mills ratio (also called non-selection hazard) as
under:
(5a)
(5b)
where and represent
probability density and cumulative distribution functions of the standard
normal distribution respectively. The ratio of and evaluated at , as represented
by and in equations (5a)
and (5b) are inverse mills ratio (IMR) which denote selection bias terms.
Two
stage method have been used to estimate ESR in previous studies () whereas the
underlying study employs a single stage Full-Information Maximum Likelihood
(FIML) method proposed by Lokshin and Sajaia (2004). The model estimates
determinants of adoption and impact of adoption on adopters and nonadopters
simultaneously thus minimizing the calculation error. The functional form is described
as:
(6)
The
signs of the correlation coefficients anddescribe important
economic explanations. In case, the signs of and are alternate, the
farmers tend to adopt exotic breeds on the basis of their comparative advantage.
Additionally, it also underlines that adopters have above average returns from
adoption and nonadopters have above-average returns from non-adoption. On contrary,
if the coefficients have the similar signs, it is the evidence of hierarchical
sorting: whether they decide to adopt or not, adopters have above average
returns but they are better off adopting, whereas nonadopters have
below-average returns in either case, but they are better off not adopting.
The
average variable costs and the gross margins per animal are provided in Table 8.
Overall, the gross margins per animal are Rs. 24,515. Surprisingly, the gross
margins of small farmers are higher than the large farmers. This is due to a
variety of reasons. One is the higher revenues earned by small farmers selling
of animals. This could be due to more animals sold by small farmers in the
survey year to fulfill any urgent cash needs. Secondly, small farmers have a
low percentage of unproductive (dry) animals at their farms. These gross
margins are calculated by dividing the costs and revenues by ‘total number of
adult animals’, not counting whether some of those adult animals could also be
dry for some period of survey year. The large farmers had 21 percent dry adult
animals at their farms as compared to small farmers having only 6 percent dry
adult animals. Because the small farmers are more likely to be cash starved,
therefore, they probably could not afford having a dry animal for longer
periods. They either sell it or replace it with milking animal.
On
the costs side, the major cost components were fodder and concentrate costs.
The labor cost for large farmers was 4 times that of the small farmers.
However, the total variable cost per animals was almost same for small and
large farmers.
The
Potential of Livestock Sector in Shifting toward High Producing Breeds and Key
Constraints regarding this Shift: One of the objectives of this study was to
explore the potential of livestock farmers to shift toward high milk producing
breeds, and to identify the key constraints in breeding and adopting of exotic
breeds. The data on milk productivity of exotic/superior breeds shows promising
results. The comparison of milk productivities of various buffalo and cattle
breeds is provided in Tables 9 and Table 10. Two facts about milk productivity
can be spotted easily by looking at these two tables. First, milk productivity
increases as we move across farm sizes, from small to large farmers (except for
some cases in cattle). Secondly, the productivity of superior/exotic breeds is
significantly higher than that of other breeds.
Similarly,
the average productivity of exotic and exotic cross breeds of cattle is higher
than that of indigenous breeds (Table 10). Our productivity results showed that
the superior breeds of animals perform better, even in the existing conditions
of undernourishment, disease, and poor breeding practices.
Factors
affecting the Adoption of Exotic Breeds of Animals, and Gross Margins of
Adopters & Non-adopters: To introduce the policy reforms to improve
the genetic potential at farms, it is necessary to understand the behavior of
adopters and non-adopters of exotic breeds of animals. In order to analyze the
driving forces behind the farmers’ decision to adopt an exotic breed of buffalo
and/or cow, we employed an ESR, as it provides the advantage of controlling for
observable and unobservable selection bias. FIML estimates of ESR for equation
(6) are provided in Table 11. The third column of Table 11 presents the
estimated coefficients of the adoption equation on exotic breed, whereas the
last two columns of the Table 11 present the impact of different factors on
gross margins of adopters and non-adopters respectively. The empirical results
of probability of exotic breed adoption show that the education of the
household head significantly increases the probability of choosing better
breeds and making better decisions (Huffman, 2001). Educated farmers are likely
to have more access to literature provided by different extension agencies on
the benefits and managements of superior animal breeds. Thus, education is a
powerful source that leads farmers to opt for exotic breeds of buffaloes and
cows.
Another
important factor that drives farmers’ decision to choose exotic breeds is their
ability to afford permanent labor for livestock. Farmers belonging to
households endowed with valuable physical capital, such as landholdings, are
more likely to choose an exotic breed of buffalo and cow. The results show that
farmers having more operational landholdings are more likely to adopt better
breeds. This could also be an indicator of affluence. Thus, affluent farmers
are more likely to adopt exotic breeds. It is observed that exotic animals
require extra care, better food and sufficient space. The coefficients of the
variables representing labor and operational land are positive and statistically
different from zero, implying that farmers endowed with labor and operational
land holding are more likely to adopt exotic animal breeds.
The
variable representing the visits of extension workers also shows positive sign
and is statistically significant, indicating that availability of dairy
extension services also play an important role in farmers’ decision to adopt
exotic breeds of animals. The empirical results suggest that farmers, where
livestock extension workers pay visits have a significantly higher probability
of having exotic breeds. Thus, improving the livestock extension services could
also be an important factor toward improving genetic potential at farms. The
coefficient of off-farm income is positive, indicating that off-income tends to
adopt exotic breed, but this variable is not significantly different from zero.
The
results of the second part of FIML endogenous switching regression model which
represents outcome equation both for adopters and non-adopters are presented in
last two columns of Tables 10. Identification of the model requires at least
one variable which is present in the adoption equation but does not enter into
the gross margin equation. Distance of farmers from livestock market is used as
the identifying instrument. It is evident from the results that distance of
farmers’ residence from livestock market significantly impacts their decision
to choose an exotic breed, but it may not affect the outcome indicating no
direct effect of distance on the gross margin of breeds. The farmers located
closely to the livestock markets are more likely to adopt an exotic breed, and
vice versa, probably due to the fact that market is imperfect, and farmers are reluctant
to trade good breeds there.
The
non-significance of covariance terms in the case of gross margins of adopters
and non-adopters, in the lower panel of the table, shows the absence of
endogenous switching in both cases, indicating that there is no self-selection
due to unobservable factors. Also results show that the covariance terms have
alternate signs with > 0 and < 0, which indicates that adoption of high
yield breed is based on its comparative advantage. It shows that farmers, who
adopted have above average returns from adoption and those who choose not to
adopt have above-average returns from non-adoption.
The
results of the second part of FIML-ESR shows that the coefficient of education
of household head is positive, indicating that level of education tends to
exert positive effect on the gross margins of both adopters and non-adopters.
Thus, education appears to have a key factor of production as it is linked with
better livestock management and economic approaches which is in line with the
human capital theory (Kousar and Abdulai, 2015; Kousar et al., 2018; Kousar
et al., 2019).
The
composition of households, such as household size, seems to have negative
impact on the gross margins of both the adopters and non-adopters. This could
be due to a high proportion of expenditures being diverted toward the increased
requirements of domestic expenditures, leaving less budget to spend on the management
of animals. However, this impact of was not significant in both cases.
Similarly, the watering frequency of dairy animals seems to have a positive
impact on gross margins. This could be due to increased milk productivity with
increased water frequency. As noted by Etgen and Reaves, 1982 and; Ali et al.,
2015, dairy animals watered more frequently or freely produce more milk,
therefore, increased watering frequency or free access to water for dairy
animals could be helpful in increasing gross margins via increased milk
productivity.
Coefficient
of hired labour is positive and significant in the case of adopters, implying
that exotic breed needs extra care. Access to livestock extension services from
various agencies has a significant positive effect on the gross margins of both
the adopters and non-adopters. Farmers, who have access to livestock extension
services, are more likely to obtain higher gross margins. This could be the
positive impact of extension services on the adoption of better livestock
management practices, as well as exotic breeds, as discussed previously.
Operational
landholdings, which also includes rented in shared in land besides owned land,
seems to have a positive and significant impact on the gross margins of both
the adopters and non-adopters. This is perhaps due to better feeding practices
with increased fodder cultivation on plenty of land.
Constraints
related to Improved Breeding and Adoption of Exotic Breeds: To investigate
reasons for low genetic potential at farms, farmers were asked to list the
constraints toward the adoption of exotic breeds and improving their breeding
practices. The results of this constraint analysis are presented as percent
scores (Table 12). The average percent score of constraints faced by all
farmers was 4.35. The constraints having a score of above 5 were designated as
severe constraints, while the constraints which had a percent score above or
equal to the average score, but below 5, were designated as moderate
constraints.
Overall,
there were 13 constraints faced by farmers in the severe and moderate category.
Out of those 13 constraints, 6 were common across all three levels of farmer.
The severe constraints reported by all farmers were the unavailability of
pedigree bulls in their herds, high price of exotic breeds, the extra care
required for exotic breeds, and the high nutrition/feed cost of exotic breeds.
Apart from these constraints, the other severe constraints reported by small
farmers were the high expenses to maintain pedigree bulls in their herds, and
the high price of artificial insemination. However, the percent scores for
these constraints on medium and large farmers were moderate. Large farmers also
reported that they did not adopt exotic breeds because they were highly
vulnerable to diseases, their drought tolerance was low, they involved high
risk, and the price of milk was low which means that return on the expensive
animals was not enough. These were reported as severe constraints. Most of
these constraints were also reported by medium farmers. However, these
constraints had moderate scores for small farmers.
The
moderate level constraints reported by all farmers were the unavailability of
pedigree bulls in the entire area, and the availability of exotic breeds.
Farmers reported that they could not cross breed their animals with superior
bulls because, these bulls were not available, even in the nearby villages, and
one of the constraints in purchasing the exotic breeds of animals was the issue
of their availability. Farmers reported that it was difficult to find a good
animal because the markets were far away, and, in some cases, the owners were
not willing to sell their animals.
Table 1.
Socioeconomic Characteristics of Livestock Farmers.
General
Information
|
Farm
Category
|
Small
|
Medium
|
Large
|
Overall
|
Age
(Years)
|
40.92
|
41.28
|
40.63
|
40.97
|
Livestock
Farming Experience (Years)
|
23.43
|
23.38
|
23.46
|
23.42
|
Schooling
(Years)
|
5.87
|
7.77
|
9.22
|
6.96
|
Family
Size (No.)
|
|
|
|
|
Adult Male
|
2.72
|
3.59
|
3.70
|
3.12
|
Adult Female
|
2.51
|
3.29
|
3.22
|
2.84
|
Children
|
2.33
|
2.76
|
3.10
|
2.58
|
Total
|
7.56
|
9.63
|
10.02
|
8.54
|
Family
Type (%)
|
|
|
|
|
Nuclear
|
42.5
|
18.3
|
15.0
|
75.8
|
Joint
|
4.2
|
5.8
|
13.3
|
23.3
|
Extended
|
0.00
|
0.8
|
0.00
|
0.8
|
Table
2. Number of Farms in the Sample.
Farm Category
|
Frequency
|
Percent
|
Small
|
190
|
55.9
|
Medium
|
90
|
26.5
|
Large
|
60
|
17.6
|
Total
|
340
|
100.0
|
Table
3. Buffaloes and Cattle Inventory.
Animals
|
Farm Category
|
Small
|
Medium
|
Large
|
Overall
|
Buffaloes
|
Adult
|
0.98
|
2.91
|
5.62
|
2.31
|
Heifer
|
0.49
|
1.04
|
2.47
|
0.99
|
Bulls
|
0.04
|
0.14
|
0.47
|
0.14
|
Calves
|
0.62
|
1.79
|
3.18
|
1.38
|
Cattle
|
Adult
|
0.83
|
1.80
|
4.78
|
1.78
|
Heifer
|
0.39
|
0.61
|
1.20
|
0.59
|
Bulls
|
0.19
|
0.67
|
0.38
|
0.35
|
Oxen
|
0.08
|
0.07
|
0.28
|
0.11
|
Calves
|
0.59
|
1.09
|
2.97
|
1.14
|
Table 4. Diversity
of Buffalo Breeds at Livestock Farms.
Farmer Category
|
Buffalo Breeds (%)
|
Nili Ravi
|
Kundi*
|
Non-descriptive/mixed
|
|
Small
|
27
|
2
|
71
|
100
|
Medium
|
39
|
0
|
61
|
100
|
Large
|
36
|
0
|
64
|
100
|
Overall
|
34
|
1
|
65
|
100
|
*Kundi is not the
common breed in study area but 2 kundi reported here was gift received by
farmer.
**Based on adult
animal.
Table
5. Diversity of Cattle Breeds at Livestock Farms (Percent).
Cattle Breed Type
|
Farm Category
|
Overall
|
Small
|
Medium
|
Large
|
Indigenous Breeds
|
Sahiwal
|
26
|
25
|
26
|
25
|
Cholistani
|
1
|
3
|
2
|
2
|
Achai
|
0
|
0
|
1
|
0.33
|
Dhanni
|
1
|
3
|
0
|
2
|
Sahiwal x Cholistani
|
1
|
0
|
0
|
1
|
Non-descriptive
|
34
|
31
|
23
|
30
|
Total
|
65
|
62
|
52
|
61
|
Exotic Breeds
|
Holstein-Friesian
|
9
|
13
|
27
|
15
|
Jersey
|
5
|
7
|
5
|
6
|
Total
|
14
|
20
|
31
|
20
|
Cross
with Exotic Breeds
|
Friesian x Sahiwal
|
14
|
16
|
14
|
14
|
Jersey x Sahiwal
|
7
|
3
|
2
|
4
|
Total
|
21
|
19
|
16
|
18
|
**Based on adult
animals
Table
6. Source of Breeding Service (Percent).
Source
|
Farm Category
|
Total
|
Small
|
Medium
|
Large
|
Buffaloes
|
Artificial
|
27.68
|
35.44
|
22.64
|
29.10
|
Natural
|
72.32
|
64.56
|
77.36
|
70.90
|
Cattle
|
Artificial
|
84.44
|
77.61
|
87.50
|
83.10
|
Natural
|
15.56
|
22.39
|
12.50
|
16.90
|
Table 7. Variable
Names, Definitions and Descriptive Statistics.
Variables Names
|
Definition of
variables
|
Sample Mean
|
Standard
Deviation
|
Dependent
variables: Adoption
of Exotic Breed and Gross Margins
|
Explanatory
variables
|
Education
|
Years of
schooling of the head of the household
|
5.89
|
4.54
|
HHSize
|
Number of
members in a household
|
5.12
|
3.20
|
Labor
|
1 if farmer has
hired permanent labor, 0 otherwise
|
0.76
|
0.44
|
WaterFreq
|
Number of times,
water served to animals in a day
|
1.02
|
0.99
|
Extension
|
1 if livestock extension
worker visits, 0 otherwise
|
0.83
|
0.43
|
OffIncome
|
Income from
sources other than farming (Rs.)
|
65452.45
|
5214.21
|
OperatLand
|
Area of land
used for cultivation (Acres)
|
12.98
|
11.92
|
Distance
|
Distance of HH
from livestock market (Km)
|
0.66
|
0.79
|
Locational
Dummies
|
Sahiwal
|
1 if Sahiwal
district, 0 otherwise
|
0.67
|
0.45
|
Sargodha
|
1 if Sargodha
district, 0 otherwise
|
0.85
|
0.32
|
|
|
|
|
Source: Survey data
Table 8. Gross
Margins per Animal (Rs.)
Average
Cost per Animal
|
Small Farmers
|
Medium Farmers
|
Large Farmers
|
Overall
|
Fodder cost
|
25,724
|
22,718
|
23,368
|
24,513
|
Concentrate cost
|
24,297
|
20,959
|
22,860
|
23,160
|
Labor cost
|
1,185
|
3,548
|
4,756
|
2,441
|
Health care cost
|
2,973
|
2,481
|
2,453
|
2,751
|
Breeding cost
|
339
|
346
|
415
|
354
|
Total Variable cost
|
54,518
|
50,051
|
53,853
|
53,218
|
Average
Value of Output per Milking Animal per Year (Rupees)
|
Milk
|
50,445
|
54,955
|
52,439
|
51,991
|
Selling of animals
|
30,174
|
21,397
|
16,677
|
25,469
|
Byproduct Revenue
|
95
|
385
|
676
|
274
|
Total
|
80,714
|
76,737
|
69,791
|
77,734
|
Gross Margin
|
26,196
|
26,685
|
15,938
|
24,515
|
Table 9.
Productivity of different Buffalo Breeds (Liters).
Farmer Category
|
Breed Type
|
Nili
Ravi
|
Kundi
|
Non-descriptive
/Mixed
|
Overall
|
Small
|
6.5
|
5.2
|
4.7
|
5.46
|
Medium
|
6.9
|
-
|
5
|
5.95
|
Large
|
7.8
|
-
|
5.2
|
6.5
|
Total
|
7.06
|
5.2
|
4.96
|
5.97
|
Table 10.
Productivity of different Cattle Breeds (Liters).
Breed Type
|
Farm Category
|
Indigenous
Breeds
|
Sahiwal
|
18.23
|
Cholistani
|
11.79
|
Dhanni
|
6.67
|
Sahiwal x Cholistani
|
12.00
|
Non-descriptive
|
4.25
|
Average
|
10.59
|
Exotic
Breeds
|
Holstein-Friesian
|
30.14
|
Jersey
|
24.16
|
Average
|
27.15
|
Exotic
Cross
|
Friesian x Sahiwal
|
25.04
|
Jersey x Sahiwal
|
14.50
|
Average
|
19.77
|
Overall
|
20
|
Table
11. Full information maximum likelihood estimates of the endogenous switching
regression.
Dependent variable: Adoption of Exotic
Breed and Gross Margins
|
Variables
|
Description
|
FIML Endogenous Switching Regression
|
Adoption
Decision
1/0
|
Adoption=1
Adopters
|
Adoption=0
Non-adopters
|
Education
|
Years
of schooling of HH head
|
0.033(0.01)***
|
999(442)
|
760(3081)
|
HHSize
|
No.
of HH members
|
-0.007(.01)
|
-2573(3527)
|
-1768(2456)
|
Labor
|
1
if farmer has hired permanent labor, 0 otherwise
|
0.772(0.17)***
|
292993(65397)***
|
-204356(45554)***
|
WaterFreq
|
No.
of times providing water to animals in a day
|
0.029(0.08)
|
11110(31437)
|
7606(21899)
|
Extension
|
1
if livestock extension worker visits, 0 otherwise
|
0.188(0.10)**
|
71734(36860)**
|
49517(25663)**
|
OffIncome
|
Income
from sources other than farming (Rs.)
|
0.000(0.00)
|
0.06(0.05)
|
-0.04(0.04)
|
OperatLand
|
Acres
of land used for cultivation
|
0.127(0.06)**
|
48382(21122)**
|
33832(14717)**
|
Sahiwal
|
1
if Sahiwal district, 0 otherwise
|
-0.036(0.13)
|
-13759(47503)
|
9958(33094)
|
Sargodha
|
1
if Sargodha district, 0 otherwise
|
-0.027(0.11)
|
-10057(41656)
|
7520(29016)
|
Distance
|
Distance
of HH from livestock market (Km)
|
0.000(0.00)***
|
|
|
Constant
|
|
0.022(0.22)
|
7454(84678)
|
-6251(58290)
|
|
|
|
379834(1.348)***
|
264576(0.909)***
|
|
|
|
0.380(0.234)
|
-0.840(0.306)
|
Note: Significance
of t-statistics of mean difference is at the *10%, **5%
and ***1% levels.
denotes the square root of the
variance of the error terms in the outcome equations, denotes correlation coefficient b/w the
error term of selection equation and error term of outcome equation.
Table
12. Constraints in Improved Breeding and Adoption of Exotic Breeds (percent
score).
Constraints
|
Farm Category
|
Overall
|
Small
|
Medium
|
Large
|
Severe
Constraints
|
|
|
|
|
Unavailability
of pedigree bulls in the herds
|
5.06
|
5.16
|
5.02
|
5.08
|
Pedigree
bulls are expensive to maintain
|
5.01
|
-
|
-
|
-
|
High
price of artificial insemination
|
5.04
|
-
|
-
|
-
|
High
price of exotic breeds
|
5.17
|
5.38
|
5.04
|
5.20
|
High
vulnerability of exotic breeds
|
-
|
-
|
5.16
|
-
|
Exotic
breeds demand extra care
|
5.08
|
5.25
|
5.24
|
5.15
|
Low
drought tolerance of exotic breeds
|
-
|
5.08
|
5.30
|
5.03
|
Exotic
breeds involve high risk
|
-
|
5.01
|
5.04
|
-
|
High
nutrition/feed cost of exotic breeds
|
5.05
|
5.04
|
5.06
|
5.05
|
Low
price of milk--
|
-
|
-
|
5.04
|
-
|
Moderate
constraints
|
|
|
|
|
Unavailability
of pedigree bulls in the whole area
|
4.72
|
4.76
|
4.94
|
4.77
|
Pedigree
bulls are expensive to maintain
|
-
|
4.96
|
4.96
|
4.99
|
High
price of cross breeding with pedigree bulls
|
4.37
|
4.36
|
-
|
4.35
|
High
price of artificial insemination
|
-
|
4.96
|
4.85
|
4.99
|
High
vulnerability of exotic breeds
|
4.92
|
4.96
|
-
|
4.97
|
Low
drought tolerance of exotic breeds
|
4.92
|
-
|
-
|
-
|
Exotic
breeds are not easily available
|
4.65
|
4.71
|
4.81
|
4.69
|
Exotic
breeds involve high risk
|
4.94
|
-
|
-
|
4.98
|
Low
price of milk
|
4.63
|
4.61
|
-
|
4.96
|
Livestock
market does not exist
|
-
|
-
|
4.51
|
-
|
Conclusions and
Policy Options: This
study aims at identifying the key constraints in the improvement of genetic
potential at farms and exploring the potential of traditional livestock sector
in shifting toward a high productivity sector. For this purpose, data were
collected from 340 livestock farmers from the Punjab, Pakistan.
Study
shows that about 70 percent of the adult population of buffaloes is comprised
of low genetic potential breed, and about 60 percent of the adult cattle
population can be described as indigenous population. Most of the buffaloes are
being bred by bulls, while artificial insemination is more prevalent for
cattle. Given the present composition of livestock herds, small farmers earn
higher gross margins on per animal basis. The empirical results of ESR showed
that the education level of farmers, the availability of labor, livestock
extension services, operational landholding, and distance from livestock market
are the main factors that affect the farmers’ decision to adopt an exotic breed
of animal. Most of these factors also had a positive impact on the gross
margins of adopters of exotic breeds. The severe constraints toward improved
breeding/adoption of exotic breeds were the unavailability of pedigree bulls in
their herds, the high price of exotic breeds, the extra care required for
exotic breeds, and the high nutrition/feed cost of exotic breeds. The moderate
level constraints in this regard were the high price of artificial
insemination, high prices of cross breeding with pedigree bulls, scarce
availability of exotic breeds, and low price of milk.
A
solid policy action is required to eliminate the constraints facing this kind
of productivity shift. Farmers do not have pedigree bulls of the superior
indigenous breeds. Pedigree bulls on a cooperative level could be provided to
improve genetic potential. New policy reforms regarding artificial insemination
could be introduced to control prices and ensure quality.
REFERENCES
- Afzal, M. and A. N. Naqvi (2004). Livestock resources of Pakistan: Present status and future trends. Qua. Sci. Vis. 9(1): 1-2.
- Ali, A., I. Musatafa., M. Q. Bilal., G. Muhammad., M. Lateef, and S. Ullah (2015). Effect of watering frequency on feed intake, milk production and composition in Sahiwal cattle during summer. The J. Anim. Plant. Sci. 25(1): 19-22.
- Anonymous. (2019). Economic Survey, Economic Affairs Division, Govt. Pakistan., Islamabad.
- Ashfaq, M., G. Muhammad., Shamsheer-ul-Haq, and A. Razzaq (2014). Effects of livestock diseases on dairy production and incomes in district Faisalabad, Punjab, Pakistan. Working paper No. 023, International Food Policy Research Institute (IFPRI), Washington DC, USA.
- Etgen, W. M. and P. M. Reaves (1982). Dairy cattle feeding and management. 6th Ed. John Wiley and Sons.
- Habib, G., A. Hameed, and M. Saleem (2019). Effect of dietary levels of protein on milk production, milk-urea and nitrogen use efficiency in peri-urban milking buffaloes. The J. Anim. Plant. Sci. 29(6): 1558-156.
- Hall, D. C., S. K. Ehui, and B. I. Shapiro (2004). Economic analysis of the impact of adopting herd health control programs on smallholder dairy farms in Central Thailand. Agri. Eco. 31(2-3): 335-342.
- Hoffmann, I. and R. Baumung (2013). The Role of livestock and livestock diversity in sustainable diets. In: Fanzo, J., Hunter, D., Borelli, T. and F. Matteri (eds.) Diversifying food and diets: Using agricultural biodiversity to improve nutrition and health issues in agricultural biodiversity. Routledge. ISBN: 1136461469, 9781136461460.
- Huffman, W. E. (2001). Human capital: Education and agriculture. Handbook of agricultural economics. 1: 333-381.
- Iqbal, M. A., A. Iqbal., N. Akbar., H. Z. Khan, and R. N. Abbas (2015). A study on feed stuffs role in enhancing the productivity of milch animals in Pakistan-Existing scenario and future prospect. Glob. Vete. 14(1): 23-33.
- Khan, M. S., N. Ahmad, and M. A. Khan (2007). Genetic resources and diversity in dairy buffaloes of Pakistan. Pak. Vet. J. 27(4): 201.
- Khan, M. S., Z. Rehman., M. A. Khan, and S. Ahmad (2008). Genetic resources and diversity in Pakistani cattle. Pak Vet. J. 28(2): 95-102.
- Kousar, R. and A. Abdulai (2015). Off-farm work, land tnancy contracts and investment in soil conservation measure in rural Pakistan. The Aus. J. Agri. and Res. Eco. 59(3):119.
- Kousar, R., M. S. A. Makhdum, and Q. Abbas (2018). Secured land rights, household welfare and agricultural productivity: evidence from rural Pakistan. Pak. J. Agri. Sci. 55(1): 243-247.
- Kousar, R., M. S. A. Makhdum., A. Abbas, and J. Nasir (2019). Issues and impacts of the apricot value chain on the upland farmers in the Himalayan range of Pakistan. Sustainability. 11(16): 4482.
- Lamy, E., S. van Harten., E. Sales-Baptista., M.M.M. Guerra, and A.M. de Almeida (2012). Factors influencing livestock productivity. In Env. Stres. and Ame. in Live. Prod. 19-51. Springer, Berlin, Heidelberg.
- Lokshin, M. and Z. Sajaia (2004). Maximum likelihood estimation of endogenous switching regression models. Sta. J. 4(3): 282–289.
- Moaeen-ud-Din, M. and M. E. Babar (2006). Livestock farming in peri-urban areas of Faisalabad, Pakistan. Live. Res. Rur. Deve. 18(1): 89-94.
- Modisane, B. M. (2009). Field services: Eradication and control of animal diseases. Onder. J. of vet. Res. 76(1): 115-121.
- Maddala, G. S. (1983). Limited dependent and qualitative variables in econometrics. Cambridge University Press, New York.
- Oltenacu, P. A. and B. Algers (2005). Selection for increased production and the welfare of dairy cows: Are new breeding goals needed? Ambio. 34(4): 311-315.
- Otufale, G. A. and M. I. Adekoya (2012). Traditional perception of cattle herdsmen: Control of ticks invasion on cattle in Ikorodu local area of Lagos State. J. of Agri. Sci. 3(2): 95-102.
- PBS. (2006). Agriculture Census. Pakistan Bureau of Statistics, Government of Pakistan.
- Rushton, J. (2013). An overview of analysis of costs and benefits of government control policy options. Proceedings of the OECD Conference on livestock disease policies: Building bridges between animal sciences and economics, OECD, Paris.
|