Hedonic Analysis of Price Expectations of Goat Producers in Afghanistan: Implications for Production and Marketing Decisions

Abstract

The authors describe the goat markets in Afghanistan by analyzing goat producers’ price expectations and by identifying the factors that determine these price expectations. Data on expected prices for goats transacted were collected from 280 goat producers from Baghlan and Nangarhar provinces, along with information on factors anticipated to influence the price expectation from May 2008 to April 2009. A price expectation model was built and analyzed using a general linear model.
Results indicated that goat producers adjusted expected prices for marketing day (Saturday and Thursday), location of sales (district and provincial markets), live weight, and goat producers’ market network. However, goat producers did not expect a premium for goat attributes like breed and age. The implications of the study are that goat producers can expect more when they plan their goat sales based on live weight, market day, marketing place and sex of goat. [Econlit Citations: Q130; Q120, C100].

Full text

Hedonic Analysis of Price Expectations of Goat Producers
in Afghanistan: Implications for Production and Marketing
Decisions
Tav va S ri nivas1
International Center for Agricultural Research in the Dry Areas (ICARDA), Afghanistan Research
Program, P.O. 1355, Kabul, Afghanistan. E-mail: s.tavva@cgiar.org
Aden Aw-Hassan
International Center for Agricultural Research in the Dry Areas (ICARDA), Social, Economic and
Policy Research (SEPR) Program, P.O. Box 5466, Aleppo, Syria. E-mail: A.Aw-Hassan@cgiar.org
Barbara Rischkowsky
International Center for Agricultural Research in the Dry Areas (ICARDA), Diversification &
Sustainable Intensification of Production Systems (DSIPS) Program, P.O. Box 5466, Aleppo, Syria.
E-mail: B.Rischkowsky@cgiar.org
Markos Tibbo
Food and Agriculture Organization of the United Nations, Regional Office for the Near East. 11 Al
Eslah El Zerai St., Dokki, PO Box 2223, Cairo, Egypt. E-mail: Markos.Tibbo@fao.org
Javed Rizvi
International Center for Agricultural Research in the Dry Areas (ICARDA), Afghanistan Research
Program, P.O. 1355, Kabul, Afghanistan. E-mail: j.rizvi@cgiar.org
Abdul Halim Naseri
International Center for Agricultural Research in the Dry Areas (ICARDA)–Afghanistan Research
Program. E-mail: abdulhalim.nasery@gmail.com
ABSTRACT
The authors describe the goat markets in Afghanistan by analyzing goat producers’ price expectations
and by identifying the factors that determine these price expectations. Data on expected prices for goats
transacted were collected from 280 goat producers from Baghlan and Nangarhar provinces, along with
information on factors anticipated to influence the price expectation from May 2008 to April 2009. A
price expectation model was built and analyzed using a general linear model.
Results indicated that goat producers adjusted expected prices for marketing day (Saturday and Thurs-
day), location of sales (district and provincial markets), live weight, and goat producers’ market network.
However, goat producers did not expect a premium for goat attributes like breed and age. The implica-
tions of the study are that goat producers can expect more when they plan their goat sales based on live
weight, market day, marketing place and sex of goat. [Econlit Citations: Q130; Q120, C100]. C2012 Wiley
Periodicals, Inc.
1. INTRODUCTION
Goats in Afghanistan are multipurpose animals providing meat and milk for direct home
consumption, providing income through sales, and serving as a living asset that can be liquidated
when needed. With a national herd of 7.3 million goats and an average holding size of 2.4 animals
(Food and Agriculture Organisation, 2003), goats are an essential element in the farming
systems and the livelihoods of rural communities. A quarter of total goats in Afghanistan are
1Presently working as Senior Scientist, Agricultural Economics, Section of Social Sciences, Central Tuber Crops
Research Institute, Thiruvananthapuram 695 017, Kerala, India. E-mail: srinictcri@yahoo.com. This article forms part
of the work done during Dr Tavva Srinivas’s tenure as Visiting Scientist (Agricultural Economics) with International
Centre for Agricultural Research in the Dry Areas (ICARDA)-Afghanistan Research Program, Kabul during 2008-09.
Agribusiness, Vol. 29 (2) 133–146 (2013) C2012 Wiley Periodicals, Inc.
Published online in Wiley Online Library (wileyonlinelibrary.com/journal/agr). DOI: 10.1002/agr.21313
133

134 SRINIVAS ET AL.
Figure 1 Distribution of Goats in Different Provinces of Afghanistan.
in the Herat, Helmand, and Nuristan provinces and the highest number of goats per family is
in Nuristan (23.5 goats), followed by Kunar (8.2) in eastern Afghanistan (Fig. 1).
Goat producers sell live goats directly themselves or through middlemen. As goats are com-
monly traded for cash, every goat producer naturally will have some price expectation on the
basis of which he or she initiates the bargaining process. The price agreed by producers and
buyers depends on their knowledge of market supply and demand, coupled with their skills in
assessing animal condition (and weight) as goats are not weighed before purchase (Bett et al.,
2011); and their market information including their knowledge of different attributes of goats
preferred by different buyers (Francis, 1990). Market price information generally in livestock
markets in developing economies is mainly limited to personal interactions between market
agents due to poor market intelligence systems (Rodriguez et al., 1995). In a conflict-affected
country like Afghanistan, one cannot expect to have an effective market intelligence system.
Under such circumstances, it is likely that producers’ production and marketing decisions may
not yield high returns. It is therefore necessary to understand factors influencing price ex-
pectations of market agents, complexities within the market price determination mechanisms,
and the importance of transparent market information that can be used by goat producers in
formulating better strategies for production and marketing of goats.
This study was conducted to evaluate the factors that determine price expectations of goat
producers in Baghlan and Nangarhar provinces having 3.28 and 3.24% of total goats in the
country with 1.75 and 2.2 goats per family, respectively. This information provides important
insights into the management and marketing practices that influence expected prices and will
indicate how producers can better tailor their goat sales to increase profitability. The hypothesis
of this study was that price expectations are influenced by the different attributes of goats
and by farmer access to information networks. The specific objectives were to develop a price
expectation model for live goats and to identify different factors influencing goat producers’
price expectations.
2. METHODOLOGY
A hedonic model of prices is one that decomposes the price of an item into separate compo-
nents that determine the price (Mart´
ınez-Garmendia, 2010). According to Lancaster (1966,
1976, 1979), goods are seen as bundles of quality characteristics and the “marginal value”
consumers’ attribute to each of the characteristics explains the variation in prices of goods.
Rosen (1974) introduced a market-based approach for deriving a hedonic price function, where
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 135
utility-maximizing buyers and sellers interact to establish the market value for a given attribute.
A differentiated product can therefore be completely described by the vector of objectively
measured characteristics of the product such that the observed price will be a composite of the
coefficients of the embedded attributes. This technique can be used in relating the price per
animal to its various attributes and characteristics (Jabbar & Diedhiou, 2003). The analysis
of covariance (AnCov) technique is a combination of linear regression and analysis of vari-
ance (ANOVA). In the AnCov technique, the results are adjusted for the linear relationship
between the dependent variable and the factors (qualitative variables) and covariates (quan-
titative variables; Singh-Knights et al., 2005). In principle, both linear regression and AnCov
techniques perform the same function except that the AnCov technique allows for more di-
rect interpretation and comparison of differences between categories of a factor (Gujarati,
1988).
This technique of hedonic pricing has been previously applied to investigate factors affecting
selling prices of small ruminants in Nigeria (Francis, 1990; Jabbar, 1998), in Pakistan (Rodriguez
et al., 1995), in Philippines (Orden et al., 2005), and in Ethiopia (Andargachew & Brokken,
1993; Ayele et al., 2006). Turner & Williams (2002) examined the factors influencing prices
received by livestock producers at the level of primary markets in rural villages and found that
price formation is socially biased by gender, wealth, and location, reflecting differential access
and powers within local markets.
A hedonic price model (a modification of the Andargachew & Brokken, 1993, model) was used
in our analysis to assess the extent to which the goat, producer, market ,and other characteristics
affect the goat producers’ price expectations.
2.1 Model Specification
The price expectation model developed to identify factors determining the producers’ price
expectations was
Wprice =β1+β2Lwt +β3E+iβiSeasoni+iβiBRi+iβiSi+iβiPSi
+iβiMNWi+iβiBCi+iβiCi+iβiMDi+iβiMPi+iβiβi
+iβiAi+iβiLwtSeasoni+iβiLwtBRi+iβiLwtSi+iβiLwtPSi
+iβiLwtM NW i+iβiLwtBCi+iβiLwtCi+iβiLwtMDi
+iβiLwtMPi+iβiLwtBi+iβiLwtAi+ei
Table 1 shows different variables that are included in this model. Due to wide variations in
per head live weight of goats transacted (from 7 kg to 42 kg), goat producers price expectations
per head adjusted to live weight, i.e., expected price per kg live weight, was used as a dependent
variable (Wprice). Sets of dummy variables were used for season of transaction, breed (BR),
sex (S), production system (PS), access to market network (MNW), body condition (BC), goat
carrier to market (C), marketing day (MD), market place (MP), and buyer (B). The variable
specification included three season dummies, three breed dummies, female and male goats for
sex, good and bad body condition, three categories of goat age, two categories of goat carrier to
market, four categories of marketing days, three categories of market place, and three categories
of buyers. Interactions of live weight with season, breed, sex, body condition, marketing day,
marketing place, buyer, carrier, production system, access to market network, and age were also
included.
Agribusiness
DOI 10.1002/agr

136 SRINIVAS ET AL.
TABLE 1. Variables in the Goat-Price Expectation Model
Name of Explanation Expected
variable of variable effects
Dependent
Expected price of goat per kg live weight
(Wprice)
Expected price of goat
transacted is divided with the
live weight
Independent (Quantitative)
Live weight (Lwt) Live weight of goat transacted Negative
Experience in goat rearing (E) Number of years of experience
in rearing goats
Positive
Independent (Qualitative)
Season dummies (other seasons as base) If the goat is transacted during
the corresponding season=1;
otherwise=0
Negative in Winter
Positive in summerSummer (April, May, June)
Winter (Oct, Nov, Dec)
Goat breed dummies (BR) (other breeds
as base)
If the goat transacted is
respective breed=1;
otherwise=0
Positive for Watani and
Gujry
Watani dummy
Gujry dummy
Sex of goat dummy (S) If the goat transacted is Male
=1; Otherwise =0
Positive for male
Production system dummy (PS) If the goat transacted is from
irrigated production system
=1; Otherwise =0
Positive for irrigated
Market net work dummy (MNW) If the goat producer with access
to market network =1;
Otherwise =0
Positive for goat producer
with access to market
network
Body condition of goat dummy (BC) If the body condition of
condition goat transacted is
Good =1;Otherwise =0
Positive for good body
Person carrying goat to market dummy
(C)
If husband carries goat to
market =1; Otherwise =0
Marketing day dummies (MDi) (other
days as base)
If the goat is transacted on the
respective week day =1;
Otherwise =0
Positive for Thursday,
Friday, and Saturday
Thursday
Friday
Saturday
Market place dummies (MPi) (Village
market sales as base)
If the goat is sold in respective
market =1; Otherwise =0
Positive for district and
provincial markets
District
Provincial
Buyer of goats dummies (Bi)(Other
buyers as base)
If the buyer is corresponding to
the market functionary
dummy =1;
Positive for butcher
Wholesaler Otherwise =0
Butcher
Age of goat dummies (Ai)(>2yearas
base)
If the goat transacted is in the
corresponding age group =1;
Otherwise =0
Positive for 1–2 year dummy
<1year
1–2 year
As more than half goat sale transactions were in the winter and summer compared to fall
and spring, three season dummies considered were for winter, summer, and other seasons’
(combined sales during fall and spring) sales.
Three categories for goat breed dummies (Watani, Gujry, and other breeds) were also intro-
duced because Watani and Gujry breeds dominated the goats sold and the aim is to see if they
yield significantly different price expectations from other less-traded goats.
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 137
TABLE 2. Mean and Percentage of Cases of Independent Variables
Var iab le M# Cases % Cases in sample
Expected price per kg live weight 178 100
Live weight 24 100
Experience in goat rearing 11 100
Season dummy 1 (Summer) – 15
Season dummy 2 (Winter) – 41
Season dummy 3 (Spring and fall seasons) – 44
Watani – 57
Gujry – 19
Other breeds (Asmari, Chily, and Tedipk) – 24
Goat sex (Male) – 78
Goat sex (Female) – 22
Production system dummy (Irrigated) – 50
Production system dummy (Rainfed) – 50
Access to market network dummy (With access) – 50
Access to market network dummy (Without access) – 50
Body condition (Good) – 91
Body condition (Bad) – 9
Goat carrier dummy (Husband) – 74
Goat carrier dummy (Others) – 26
Market day dummy 1 (Thursday) – 24
Market day dummy 2 (Friday) – 31
Market day dummy 3 (Saturday) – 14
Market day dummy 4 (Other days) – 31
Market place dummy 1 (District) – 63
Market place dummy 2 (Province) – 7
Market place dummy 3 (Other places) – 30
Buyer dummy 1 (Wholesaler) – 16
Buyer dummy 2 (Butcher) – 60
Buyer dummy 3 (Others) – 24
Age dummy 1 (<1 year) – 8
Age dummy 2 (1–2 years) – 84
Age dummy 3 (>2 years) – 8
Marketing on Thursdays, Fridays, and Saturdays were included as marketing day dummies
because goat producers sold 24%, 31%, and 13% of goats on Thursday, Friday, and Sat-
urday, respectively, in Baghlan and Nagarhar provinces. As district and provincial markets
in both the provinces and village markets in Nangarhar province are held on these market
days, the majority of goat producers carried out goat sales on these days. Hence, transactions
done on Sunday, Monday, Tuesday, and Wednesday were placed into one category as “other
days” under market day dummy variables, thus making only four categories of marketing
days.
The betas represent the structural parameters in the equation. Mean and percentage of cases
of variables included in the price expectation model are presented in Table 2. The model was
fitted using the general linear model procedure in SPSS package.
2.2 Data
Two-hundred eighty goat producers were randomly selected in equal proportions for rainfed
and irrigated systems from 28 villages in four districts (Baghlan-e-Sannhati, Pul-I-Kumiri, Dar-
e-Noor, and Achin) in Baghlan and Nangarhar provinces (Fig. 2). The districts were purposively
Agribusiness
DOI 10.1002/agr

138 SRINIVAS ET AL.
Figure 2 Districts Surveyed in the Baghlan and Nangarhar Provinces in Afghanistan. The Afghanistan
province and district boundaries shown on this map are provisional based on the United Nations Popu-
lation Fund Central Statistics Organization Household Listing (UNFPA/CSO HHL) Project fieldwork
and it does not signify official acceptance by the United Nations.
selected to represent areas where development activities under the “Goats for Women Project”2
were implemented and to include others without project activities. Seven villages from each
district and 10 households from each village were selected randomly.
Data on production, goat market structure, expected price, live weight, sex, age, breed, body
condition, carrier (person who takes the animal to market), month and day of marketing,
location of market place, access to market network and buyer of the latest live goat transactions
during the previous year (May 2008–April 2009) were collected from goat producers using a
structured questionnaire. In the absence of any records on goat transactions, producers were
asked to give prices they expected to receive for their goats during their latest goat transactions.
3. RESULTS AND DISCUSSION
3.1 Price Expectation Model for Goats
The parameters estimated for goat price expectations are shown in Table 3. The adjusted R2
of the model was about 55%, indicating the percentage of variation in goat producers’ price
2To improve the skills and knowledge of rural women in raising dairy goats, processing and marketing surplus
products, and improving the use of natural resources and their access to technologies, the International Centre for
Agricultural Research in the Dry Areas (ICARDA) implemented an International Fund for Agricultural Development
(IFAD)-cofunded pilot research program, “Rehabilitation of Agricultural Livelihoods of Women in Marginal and
Post-conflict Areas of Afghanistan” in the Nangarhar (Dar-e-Noor district) and Baghlan (Baghlan-e-Sannhati district)
provinces of Afghanistan.
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 139
TABLE 3. Price Expectation Model for Goats in the Baghlan and Nangarhar Provinces in Afghanistan
Parameter Coefficient (βi)SE
Intercept 539.89∗∗ 87.27
Live weight −15.13∗∗ 3.80
Experience in goat rearing 0.19 0.24
Season dummy 1 (Summer=1; Otherwise=0) 19.36 19.48
Season dummy 2 (Winter =1; Otherwise =0) 20.34 15.69
Breed dummy 1 (Watani =1; Otherwise =0) 3.56 26.71
Breed dummy 2 (Gujry =1; Otherwise =0) −36.84 35.60
Sex (male =1, female =0) 42.03∗17.72
Production system (Irrigated =1; Otherwise =0) −4.98 12.21
Access to market network (With access =1; Otherwise =0) −29.93∗14.83
Body condition (Good =1; Otherwise =0) −0.77 20.12
Goat carrier (Husband =1; Otherwise =0) 10.12 15.80
Marketing day dummy 1 (Thursday =1; Otherwise =0) 36.97∗20.11
Marketing day dummy 2 (Friday =1; Otherwise =0) 18.00 20.55
Marketing day dummy 3 (Saturday =1; Otherwise =0) 60.79∗24.81
Market place dummy 1 (District =1; Otherwise =0) 64.34∗∗ 18.05
Market place dummy 2 (Province =1; Otherwise =0) 154.02∗∗ 29.61
Buyer dummy 1 (Wholesaler =1; Otherwise =0) 4.29 23.40
Buyer dummy 2 (Butcher =1; Otherwise =0) −6.40 15.86
Age dummy 1 (<1year=1; Otherwise =0) 42.97 40.48
Age dummy 2 (1–2 years =1; Otherwise =0) 46.75 31.62
Live weight ×Season dummy 1 (Summer =1; Otherwise =0) −1.12 0.86
Live weight ×Season dummy 2 (Winter =1; Otherwise =0) −0.78 0.62
Live weight ×Breed dummy 1 (Watani =1; Otherwise =0) −0.13 1.14
Live weight ×Breed dummy 2 (Gujry =1; Otherwise =0) 1.35 1.33
Live weight ×Sex (Male =1; Otherwise =0) −1.83∗0.73
Live weight ×Production system dummy (Irrigated =1; Otherwise =0) 0.33 0.50
Live weight ×Market network (With access =1; Otherwise =0) 1.12∗0.60
Live weight ×Body condition (Good =1; Otherwise =0) 0.40 0.98
Live weight ×Goat carrier (Husband =1; Otherwise =0) −0.25 0.61
Live weight ×Marketing day dummy 1 (Thursday =1; Otherwise =0) −1.64∗0.99
Live weight ×Marketing day dummy 2 (Friday =1; Otherwise =0) −0.74 0.87
Live weight ×Marketing day dummy 3 (Saturday =1; Otherwise =0) −1.97∗0.98
Live weight ×Market place dummy 1 (District =1; Otherwise =0) −2.27∗∗ 0.70
Live weight ×Market place dummy 2 (Province =1; Otherwise =0) −6.22∗∗ 1.57
Live weight ×Buyer dummy 1 (Wholesaler =1; Otherwise =0) −0.01 0.94
Live weight ×Buyer dummy 2 (Butcher =1; Otherwise =0) 0.42 0.62
Live weight ×Age dummy 1 (Age <1year=1; Otherwise =0) −2.00 1.78
Live weight ×Age dummy 2 (Age 1–2 year =1; Otherwise =0) −1.95∗1.09
Adjusted R20.55
∗p<0.05. ∗∗p<0.01.
expectations that the model explains. An analogous model for expected headage price per goat
had adjusted R2of 82%.
Expected prices and per kg live weight were negatively related. With one unit increase in
the live weight of goat, expected price decreased by Afs 153per kg. Rodriguez et al., (1995)
also reported findings similar to our results that per kg live weight was negatively related to
the expected prices. This is true because in the survey data, per kg price expectation for a goat
weighing 7 kg was Afs 286 and for a goat weighing 42 kg, it was Afs 167. As the animal weight
increases, price expectation did not increase proportionately. Thirty-four percent of goats sold
were less than 20 kg in weight. Hence, the inverse relationship is as expected.
3Afs is the abbreviation for the Afghanistan currency, Afghani. One US $ =Afs 48 in 2012.
Agribusiness
DOI 10.1002/agr

140 SRINIVAS ET AL.
Goat sex was determined to have a significant role in price determination. When male goats
or bucks were taken to market, the expected unit price increased by Afs 42 per kg live weight,
which is significant. From the observed prices received by goat producers, it was clearly evident.
Male goats fetched higher prices (Afs 147 per kg) than females (Afs 139 per kg). However, the
difference was not statistically significant. Thus, goat producers were expecting a high price
for male goats or bucks than for does in Baghlan and Nangarhar provinces. Interestingly,
when live weight interacted with sex, price expected for male goat decreased by Afs 18.3 per
10 kg live weight. This indicates that goat producers expect more for relatively heavier female
goats.
Marketing day was also an influencing factor on the price expectations for goats. Expected
prices were high when goats were sold on market days such as Saturdays and Thursdays in the
week. Also, if goats were marketed on Friday, the expected price was high, but not significant.
Butchers indicated that in general demand for goat meat was high on Thursdays (Baghlan,
Nangarhar, and Kabul). Also, goat producers were expecting high prices on Thursdays (Afs
195 per kg live weight) and Fridays (Afs 172 per kg live weight) compared to other marketing
days (Afs 168 per kg live weight) as observed from the survey data.
The nature of the market (village, district, or province) was another important factor having
significant influence on expected prices. Consistent with other studies (Emuron et al., 2010;
Mlozi et al., 2003; Moges et al., 2010; Williams et al., 2006), expected as well as observed prices
were higher in provincial markets than in district markets. Expected prices were high when goat
producers sell in a provincial and district market than in a village market itself. Data collected
from butchers has confirmed that prices were higher in provincial markets than in district mar-
kets by Afs 15 per kg weight. However when live weight interacted with market place, expected
prices decreased more in provincial market (Afs 62 per 10 kg weight) than in district market (Afs
23 per 10 kg weight). This suggests that provincial and district markets do not prefer heavier
goats.
Consistent with the findings by Dossa et al., (2008) in Southern Benin, and Okali & Upton
(1985) in Southern Nigeria, and by Rodriguez et al., (1995) in Pakistan, breed also had no
significant effect on price expectations in the present study. The Gujry breed was top ranked by
goat producers for meat purposes. However, this was not translated significantly in the model
results. It may be due to the fact that buyers gave preference to live weight rather than to breed.
Negative correlation between expected price and live weight for female Gujry goats transacted
from Nangarhar province also supports the view that live weight is an important factor in
expected goat prices.
Similarly, goat body condition did not have a significant influence on price expectations as
was also the case by Bett et al., (2011) in Kenya. The season of sale also did not have significant
influence on goat price expectations as indicated by the seasonal coefficients obtained from the
model.
Price expectations of goat producers with access to an information network decreased by Afs
30 per kg live weight. This may be due to the fact that goat producers’ market network was not
well developed and were poorly informed about the prices leading to high price expectations.
However, the price expectations of goat producers with access to a market network when
interacted with live weight have increased by Afs 11 per 10 kg live weight. Similar to the
study by Bett et al., (2011), poor market network/information of goat producers limited their
marketing decisions and consequently affected production and sales decisions. Upton (2000)
acknowledges that the lack of information results in poor integration of spatially dispersed
markets and cyclical fluctuations in production and prices. However, the relatively high margins
for the intermediaries reflect opportunities present at the market (Aklilu et al., 2007). Therefore,
these opportunities can be effectively utilized by the farmers if they are provided with adequate
and reliable information as well as through group marketing. The goat production system is
not having significant influence on the price expectations, even with the interaction of live
weight.
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 141
3.2 Live Weights and Prices of Goats Traded
In the price expectation model, the live weight of goat has played an important role among
different variables considered. Therefore,an attempt is made in the following sections to explain
the goat marketing in Baghlan and Nangarhar provinces taking into account the variations in
the live weight and prices of goats with respect to breed, sex, production system, province, etc.
The prices and live weights of goats correspond to the actual transactions reported by goat
producers.
3.2.1 Goat sales by breed, sex and province.
Watani breed dominated goat sales. Out of goats
sold by surveyed producers, Watani was the most frequent (57%) followed by Gujry (19%),
Tedipk (19%), Asmari (4%) and Chily (1%). Though goat producers indicated Gujry as the
preferred breed for meat purposes, large sales of the Watani breed are due to its larger share
in the population. Also, 90% of Gujry goats transacted were from Nangarhar as it was the
dominant breed among Nangarhar producers.
Male goats dominated the sales volume in all breeds. Seventy-eight percent of goats transacted
by goat producers were males and 22% were females. Males sold at a higher percentage in
Nangarhar province (83%) than in Baghlan (77%). Male goats of Watani were marketed at a
lower weight than were females, indicating that they were marketed at a younger age due to
their high market demand, while females were kept for breeding. In the case of the Gujry breed,
males were heavier than females. As Gujry is the most preferred breed for meat purposes, males
were sold at a later stage after attaining good weight, coinciding with Eid Al Adha, a religious
holiday, as 50% of them were sold in December. The average live weight of male and female
goats of Nangarhar province was almost the same (28 kg each), but female goats were heavier
than the male goats (19 kg) of Baghlan province.
Overall, males fetched higher prices than females. This is also the case in all breeds except
for Gujry and Tedipk. Price per kg live weight of males was higher in Baghlan, whereas in
Nangarhar province it was higher for females. The high prices for females need to be further
probed especially in Nangarhar province, as the total number of female Gujry and Tedipk goats
sold were only eight and five, respectively, in the current survey sample.
The high price per kg and live weight correlations obtained indicated that prices offered were
proportional to live weights. This also suggests that goats were mainly purchased for slaughter
and prices were arrived based on live weight of the goat. The low correlation coefficient (0.21)
of price and live weight for overall female Gujry goats may be due to poor condition of these
goats as is evident from the negative correlation for the same in Nangarhar province (Table 4).
3.2.2 Weights and prices by production system.
Prices and live weights of goats were also
analyzed by production system (Table 5). There was no significant difference in the weight of
goats sold between irrigated (23 kg) and rainfed (24 kg) production systems. Price per kg live
weight was more in irrigated production system (Afs 146) than in rainfed (Afs 141). The high
price-live weight correlation coefficient obtained for both production systems indicates that
prices offered for animals were proportional to live weights except in the Pul-I-Kumiri district.
3.3 Live Weight and Seasonality of Goat Sales
Differences in the live weight of goats transacted in different seasons between production
systems and provinces are shown in Figure 3. The live weight per goat was high during the
spring and summer in an irrigated production system, whereas it was high in the winter in
a rainfed production system. Goat producers spend more on supplementary feed in rainfed
production systems than the producers from irrigated production system do to fatten the goats
in winter. This high expenditure on supplementary feed may be the reason for more goat
weight during winter for the goats from a rainfed production system. The live weight of goats
marketed from the Nangarhar province was higher compared to animals from the Baghlan
province. This was due to early age sales of goats in Baghlan (1.5 years) than in Nangarhar
Agribusiness
DOI 10.1002/agr

142 SRINIVAS ET AL.
TABLE 4. Live Weights, Prices per Head and per kg and Live Weight-Price Correlations for Goats in the Baghlan and Nangarhar
provinces (Breed and Sex)
Mean goat Mean live Mean live Correlation
price weight (kg) weight price coefficient (price-
Province Breed Sex (Afs/head) (Afs/kg) live weight) (r)
Baghlan province Watani Male 2514 18 142 0.57
Female 2477 19 133 0.46
Gujry Male 1850 17 109 1.00
Female 4000 30 133 -
Asmari Male 2875 22 132 0.69
Female 4000 33 121 -
Tedipk Male 3183 19 172 0.82
Female 3500 25 143 -
Nangarhar province Watani Male 4170 28 148 0.85
Female 4783 34 142 0.88
Gujry Male 4737 34 138 0.79
Female 4633 31 151 −0.91
Asmari Male 3100 21 145 0.33
Female 3800 24 158 -
Chili Male 5100 30 170 1.00
Tedipk Male 3714 24 152 0.69
Female 4025 26 158 0.80
Overall Watani Male 2947 20 144 0.77
Female 2909 21 136 0.82
Gujry Male 4542 33 137 0.85
Female 4533 31 145 0.21
Asmari Male 2971 22 138 0.46
Female 3900 29 137 1.00
Chili Male 5100 30 170 1.00
Tedipk Male 3631 23 155 0.80
Female 3900 25 158 0.82
Baghlan All male 2588 18 151 0.62
All female 2610 20 133 0.58
Nangarhar All male 4203 29 146 0.78
All female 4494 30 150 0.81
Overall Male 3423 24 147 0.82
Female 3277 24 139 0.81
Note. Afs is the abbreviation for the Afghanistan currency, Afghani. One US $ =Afs 48 in 2012.
(1.8–2 years). Overall, the live weight of goats marketed was higher during the spring and fall
through the year. High prices in the fall and more live weights of goats sold in the fall indicate
that goats were priced mostly for their body condition and weight. The high price in the fall
season despite a large supply and sale of goats, was due to high demand for goat meat during
festivals (Eid Al Fitr and Ramadan) in this season during 2008. Similarly, other studies noted
the existence of the same scenario relating to the price and the sociocultural events especially
the Christian and Muslim festivals (Aklilu et al., 2007; Bett et al., 2011; Emuron et al., 2010;
Halima et al., 2007; Moges et al., 2010). Therefore, the goat producers should take advantage
of the demand shifts resulting from such festivals.
3.4 The Differences Between Expected and Observed Prices
The price differential between anticipated and observed indicates the bargaining power of
goat producers. Irrespective of the production systems, the difference between expected and
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 143
TABLE 5. Live Weights, Prices per Head and per kg and Price/Live Weight Correlations forGoats (Production System and Province)
Mean price Mean price per Correlation
per head Mean live kg live coefficient(r)
Category District (Afs) weight (kg) weight (Afs) (Price- Liveweight)
Irrigated production
system
Baghlan-e-sannhati 2467 17 142 0.82
Pul-i-kumiri 2545 17 153 0.35
Dar-i-noor 4562 32 143 0.66
Achin 3763 25 150 0.62
All 3396 23 146 0.86
Rainfed production
system
Baghlan-e-sannhati 2566 19 137 0.84
Pul-i-kumiri 2815 21 131 0.29
Dar-i-noor 4664 33 142 0.93
Achin 3668 24 154 0.69
All 3418 24 141 0.82
Baghlan province Baghlan-e-sannhati 2515 18 140 0.83
Pul-i-kumiri 2692 19 140 0.33
All 2515 18 140 0.83
Nangarhar province Dar-i-noor 4608 32 143 0.80
Achin 3712 24 152 0.66
All 4251 29 146 0.83
Overall 3407 24 143 0.84
Note. Afs is the abbreviation for the Afghanistan currency, Afghani. One US $ =Afs 48 in 2012.
observed goat prices was higher in summer and winter (Fig. 4). These price differences were
statistically significant. The expected price was more during the summer, whereas the observed
price per kg live weight was more during the fall in both the production systems. Similar results
of greater difference in the expected price for goats in the summer were obtained by Rodriguez
et al., (1995) in Pakistan. Large sales of goats during the summer (as crosschecked with traders),
caused a decline in observed prices; hence, the large difference between expected and observed
prices in summer. The high observed price during the fall might be due to Ramadan and Eid
Al Fitr festival demand in 2008.
When the difference between anticipated and observed prices of live goats were examined
by province, as represented in Figure 5, it was clear that the difference was high during the
fall and winter in the Baghlan province and during the spring and fall in the Nangarhar
province. Overall, the difference was high in the Baghlan compared to the Nangarhar province.
Figure 3 Live Weight of Goats Marketed Based on Production System, Province, and Season.
Agribusiness
DOI 10.1002/agr

144 SRINIVAS ET AL.
Figure 4 Goat Producer Prices and Deviations FromExpected Prices per kg Live Weight Under Irrigated
and Rainfed Production Systems.
Figure 5 Goat Producers’ Expected and Observed Live Weight Prices in the Baghlan and Nangarhar
Provinces.
The average price differential for goats marketed in Baghlan represented 26% of the average
expected price per kg live weight with a coefficient of variation of 10%. In contrast, the price
difference for goats marketed in Nangarhar represented only 7% of the price expected per kg
live weight with a coefficient of variation of 79%. This indicates that Nangarhar goat markets
have more room for bargaining than the Baghlan goat markets.
4. CONCLUSIONS
The foregoing analysis identified live weight, sex, marketing day, market place, and access to
a market network as important factors influencing goat producers’ price expectations in the
Baghlan and Nangarhar provinces in Afghanistan. The study also further indicates that live
weight has more influence (40%) on goat producers’ price expectations than the breed or age
of goats. The study only analyzed the components of goat producers’ price expectations and
it provided a baseline for goat producers to plan their sales. Goat producers can expect more
when they plan their goat sales based on live weight, market day, marketing place, and goat sex.
Improving grazing sources such as green fodders, crop residues, etc., in the countryside can be
of help to goat producers in reducing their feed cost for goat production, as live weight is an
important determinant of prices.
Agribusiness
DOI 10.1002/agr

PRICE EXPECTATIONS OF GOAT PRODUCERS IN AFGHANISTAN 145
ACKNOWLEDGMENTS
The authors are grateful to the International Fund for Agricultural Development (IFAD)
for their financial support to the project “Rehabilitating Agricultural Livelihood of Women
in Marginal and Post-Conflict Areas of Afghanistan and Pakistan.” The authors gratefully
acknowledge the hard work of the ICARDA team based in Afghanistan in a very difficult and
insecure environment. Sincere thanks are due to the Ministry of Agriculture, Irrigation and
Livestock (MAIL) of Afghanistan, and its provincial Directorates in the two target provinces.
Without the full cooperation and support received from the Ministry of Women Affairs, trading
and farming communities, “Shuras” and “Village Elders,” and security updates / assistance
from providing agencies, it would not have been possible to conduct this study.
REFERENCES
Aklilu, H., Almekinders, C., Udo, H., & van der Zijpp, A. (2007). Village poultry consumption and market-
ing in relation to gender, religious festivals and market access. Tropical Animal Health and Production, 39,
165–177.
Andargachew, K., & Brokken, R. (1993). Intra-annual sheep price patterns and factors underlying price variations in
the central highlands of Ethiopia. Agricultural Economics, 8, 125–138.
Ayele, G., Jabbar, M.A., Teklewold, H., Mulugeta, E., & Kebede, G. (2006). Seasonal and inter-market differences in
prices of small ruminants in Ethiopia. Journal of Food Products Marketing, 12, 59–77.
Bett, H.K., Peters, K.J., & Bokelmann, W. (2011). Hedonic price analysis to guide in breeding and production of
indigenous chicken in Kenya. Livestock Research for Rural Development, 23, 6. Retrieved October 19, 2011, from
http://www.lrrd.org/lrrd23/6/bett 23142.htm
Dossa, L.H., Rischkowsky, B., Birner, R., & Wollny, C. (2008). Market potential for the local Djallonke goat in
Southern Benin: Empirical evidence from two rural markets. Livestock Research for Rural Development, 20,
Article 7. Retrieved January 5, 2010, from http://www.lrrd.org/lr rd20/5/doss20078.htm
Emuron, N., Magala, H., Kyazze, F.B., Kugonza, D. R., & Kyarisiima, C.C. (2010). Factors influencing the trade
of local chickens in Kampala city markets. Livestock Research for Rural Development, 22, Article 76. Retrieved
February 21, 2011, from http://www.lrrd.org/lrrd22 /4/emur22076.htm
Food and Agriculture Organisation. (2003). Livestock census of Afghanistan, Kabul: Author.
Francis, P.A. (1990). Small-ruminant marketing in Southwest Nigeria. Agricultural Economics, 4, 193–208.
Gujarati, D.N. (1995). Basic econometrics (3rd ed.). New York: McGraw Hill.
Halima, H., Neser, F., van Marle-Koster, E., & De Kock, A. (2007). Village-based indigenous chicken production
system in north-west Ethiopia. Tropical Animal Health and Production, 39, 189–197.
Jabbar, M.A. (1998). Buyer preferences for sheep and goats in southern Nigeria: A hedonic price analysis. Agricultural
Economics, 18, 21–30.
Jabbar, M.A., & Diedhiou, M.L. (2003). Does breed matter to cattle farmers and buyers? Evidence from West Africa.
Ecological Economics, 45, 461–472.
Lancaster, K. J. (1966). A new approach to consumer theory. The Journal of Political Economy, 74, 132–157.
Lancaster, K.J. (1976). Hierarchies in goods-characteristics analysis. Advances in Consumer Research, 3, 348–352.
Lancaster, K.J. (1979). Variety, equity and efficiency. New York: Columbia University Press.
Mart´
ınez-Garmendia, J. (2010). Application of hedonic price modelling to consumer packaged goods using store
scanner data. Journal of Business Research, 63, 690–696.
Mlozi, M.R.S., Kakengi, A.V.M., Minga, U.M., Mtambo, A.M., & Olsen, J.E. (2003). Marketing of free range local
chickens in Morogoro and Kilosa urban markets, Tanzania. Livestock Research for Rural Development, 15(2).
Retrieved February 3, 2011, from http://www.lrrd.org /lrrd15/2/mloz152.htm
Moges, F., Azage, T., & Tadelle, D. (2010). Indigenous chicken production and marketing systems in Ethiopia:
Characteristics and opportunities for market-oriented development [IPMS (Improving Productivity and Market
Success) of Ethiopian Farmers Project Working Paper 24]. Nairobi, Kenya: International Livestock Research
Institute.
Okali, C., & Upton, M. (1985). The market potential for increased small ruminant production in Southwest Nigeria.
In J.E. Sumberg & K.A. Cassaday (Eds.), Sheep and goats in humid West Africa. In Proceedings of the Workshop
on Small Ruminant Production Systems in the Humid Zone of West Africa (pp. 68–74). Addis Ababa, Ethiopia:
ILCA.
Orden, M.E.M., Jamandre, W.E., Brown, E.O., Orden, E.A., Cruz, E.M., Alo, A.P., et al. (2005). Traders’ pref-
erence for goat characteristics in selected markets of Pangasinan, Philippines. Animal Science Journal, 76,
179–185.
Rodrigues, A., Ali, I., Afzal, M., Shah, N.A., & Mustafa, U. (1995). Price expectations of sheep and goats by producers
and intermediaries in Quetta market, Pakistan. Agricultural Economics, 12, 79–90.
Rosen, S. (1974). Hedonic prices and implicit markets: Product differentiation in pure competition. The Journal of
Political Economy, 82, 34–55.
Agribusiness
DOI 10.1002/agr

146 SRINIVAS ET AL.
Singh-Knights, D., Smith, D.K., & Knights, M. (2005, July 24–27). A hedonic analysis of sheep and goat prices in a
changing environment: The role of consumers and implications for management. Paper presented at the American
Agricultural Economics Association Annual Meeting, Providence, RI.
Turner, M.D., & Williams, T.O. (2002). Livestock market dynamics and local vulnerabilities in the Sahel. World
Development, 30, 683–705.
Upton, M. (2000). The ‘livestock revolution’ – implications forsmallholder agriculture: A case study of milk and poultry
production in Kenya. Food and Agriculture Organization (FAO) Livestock Information and Policy Branch, AGAL.
Retrieved from http://www.fao.org/ag/againfo/resources/en/ publications/sector_discuss/PP_Nr1_Final.pdf.
Accessed on July 10, 2011.
Williams, T.O., Okike, I., & Spycher, B. (2006, August). A hedonic analysis of cattle prices in the central corridor of
West Africa: Implications for production and marketing decisions. Paper presented at the International Association
of Agricultural Economists Conference, Gold Coast, Australia. Retrieved from http://purl.umn.edu/25423
Dr Tavva Srinivas is currently Senior Scientist (Agricultural Economics) at the Central Tuber Crops Re-
search Institute of Indian Council of Agricultural Research. He earned his M.Sc (Ag.) (1991) in Agricultural
Economics at Andhra Pradesh Agricultural University and Ph.D degree (1995) in Agricultural Economics
at Banaras Hindu University, India with a specialization in agricultural production economics. His research
focuses on impact assessment of agricultural technologies, techno-economic feasibility studies, participatory
evaluation of improved varieties, improving supply chains of agricultural produce. He has experience in impact
assessment of technologies, marketing studies, resource use efficiency and yield gap analysis.
Dr Aden Aw-Hassan is the Program Director of Social, Economic and Policy Research Program (SERP)
at the International Center for Agricultural Research in Dry Areas (ICARDA). He received M.Sc in
Agricultural Economics at Utah State University in 1988 and Ph.D in Agricultural Economics at Oklahoma
State University in 1992. He has over 20 years experience in agricultural research and development in
developing countries. His research focuses on impact assessment, technology evaluation, market access,
micro-finance, natural resource economics, participatory research, gender analysis, poverty and livelihoods.
He has also practical field experience in technology transfer and in micro-finance.
Dr. Barbara Rischkowsky is a senior livestock scientist at the International Center for Agricultural Research
in Dry Areas. She earned her Master degree (1989) and Ph.D degree (1996) in Agricultural Sciences at the
Justus-Liebig University in Giessen/Germany with a specialization in livestock production. Her expertise is
in analysis and management of ruminant production systems and animal genetic resources. She has extensive
international experience in leading livestock research and development projects in developing countries.
Dr. Markos Tibbo is currently Animal Production and Health Officer at the Regional Office for the Near
East of the Food and Agriculture Organization of the United Nations. He is an Animal Breeder, Geneticist
and Veterinarian by training. He received a DVM degree in Veterinary Medicine at Addis Ababa University
(1993, Addis Ababa, Ethiopia) and a Ph.D in Animal Science (especially in genetics and breeding) at the
Swedish University of Agricultural Sciences (2006, Uppsala, Sweden). His research focuses on improving
livestock production and health under low-input production systems. He has experience in epidemiological
studies, disease diagnosis and control, genetics of disease resistance, breed characterization and genetic
improvement and management.
Dr Javed Rizvi is the Country Program Manager, International Center for Agricultural Research in the Dry
Areas (ICARDA) – Afghanistan Research Program. He earned his Ph.D in Plant Physiology (1981) at
Gorakhpur University, India. He has more than 25 years of research experience, including a long experience on
conflict-post conflict research for development issues. His research interests include crop-livestock integration
research, food security, transfer of technology, alternative livelihoods and on-farm management.
Abdul Halim Naseri is the National Coordinator, Goat Project, International Center for Agricultural Re-
search in the Dry Areas (ICARDA) – Afghanistan Research Program. He earned his DVM degree (1993)
at Nangarhar University, Afghanistan. His current topics of research interest include Small ruminants and
forages in dry lands.
Agribusiness
DOI 10.1002/agr