Rahman 2020

Perceived Risk and Risk Management
Strategies in Pond Aquaculture

Md Takibur Rahman, Patuakhali Science and Technology University and University of Copenhagen;
Rasmus Nielsen, University of Copenhagen; Md Akhtaruzzaman Khan, Bangladesh Agricultural
University; Dewan Ahsan, University of Southern Denmark
ABSTRACT
The growing aquaculture industry is facing several challenges including risks and uncertainties. Studies explor-
ing farmers’ risk perceptions and risk management strategies are, however, limited within pond aquaculture,
although they are well described within the field of agriculture. This study explores farmers’ risk perceptions
and risk management strategies in pond aquaculture and the relationship of risk perceptions and risk manage-
ment strategies with farm and farmers’ characteristics. The data are analyzed using principal component anal-
ysis and multivariate regression. The results show that price variability and financial risks are perceived as the
most influential risk factors. Farm management and financing are perceived as the most effective risk manage-
ment strategies. In most cases, farmers need to focus on more than one risk management strategy to address a
particular type of risk. This study provides knowledge of farmers’ risk perceptions and strategies, which can be
used to develop better and more focused management strategies.
The authors would like to thank three anonymous reviewers for their suggestions and comments that
helped substantially to improve the paper. The authors would also like to thank the Danish International Devel-
opment Agency (DANIDA) for its financial support for the project “Upgrading Pangas and Tilapia Value Chains
in Bangladesh,” project number F38A26778, under which this research was conducted.
Key words: Aquaculture, management strategies, risk perception.

JEL codes: C01, D22, Q12, Q14, Q22.
IN TRODUCTIO N
Aquaculture is growing fast in developing countries (Belton, Bush, and Little 2018). The fast
growth is in many cases driven by an uncontrolled expansion of small-scale subsistence farmers,
who in most cases have limited knowledge of management practices and lack managerial abilities
(Rahman et al. 2020). Because of the lack of knowledge and absence of spatial planning, these
farmers may also be more exposed to some of the risks and uncertainties within the aquaculture
farming industry. These risks include yield loss due to diseases and poor management, increasing
costs on inputs due to increasing demand, and fluctuations in prices on outputs due to increased
supply (Lebel, Lebel, and Chuah 2018; Kobayashi et al. 2015; Le and Cheong 2010). Furthermore,
Md Takibur Rahman is a PhD student, University of Copenhagen, Denmark, and Patuakhali Science and Technology Uni-
versity, Department of Accounting and Information Systems, Patuakhali Science and Technology University, Dumki, Patuakhali
8602, Bangladesh (email: [email protected]). Rasmus Nielsen is an associate professor, Department of Food and Resource Econom-
ics, Faculty of Science, University of Copenhagen, Rolighedsvej 25, DK-1958 Frederiksberg C, Denmark (email: [email protected]).
Md Akhtaruzzaman Khan is a professor, Department of Agricultural Finance, Bangladesh Agricultural University, Mymensingh,
Bangladesh (email: [email protected]). Dewan Ahsan is an associate professor, Department of Sociology, Environmental, and
Business Economics, University of Southern Denmark, Niels Bohrs Vej 9, DK-6700 Esbjerg, Denmark (email: [email protected]).
Received December 30, 2019; Accepted July 8, 2020; Published online December 4, 2020. https://doi.org/10.1086/711066
Marine Resource Economics, volume 36, number 1. © 2020 MRE Foundation, Inc. All rights reserved.
0738-1360/2021/3601-0003$10.00
44 | MARINE RESOURCE ECONOMICS | VOLUME 36 NUMBER 1 2021
small-scale subsistence farmers are often financially constrained (Mitra, Khan, and Nielsen 2019)
and may also be confronted with regulatory and environmental issues (Asche 2015). Thus, many
types of risks are associated with fish farming in terms of production, marketing, financial, human,
and institutional risks.
The expansion and consolidation of the aquaculture sector in developing countries is impor-
tant because the sector provides food security and income for the rural poor in many of these
countries (Belton, Bush, and Little 2018). Being able to identify the major risk factors and pro-
vide tools to mitigate these are important for the further development of this sector, with the aim
to increase food security and alleviate poverty. However, only a limited amount of empirical
work addresses risks and uncertainties in pond aquaculture (production to market). Therefore,
more attention is needed to identify the factors that contribute to risks and uncertainties and
which methods can be used to mitigate these (Tisdell et al. 2012).
The objectives of this study are (1) to identify pond aquaculture farmers’ perceptions of risk
(subjective risks) and perceived risk management strategies based on ranking of sources by risk
level, and (2) to assess the impacts of farm characteristics and farmers’ sociodemographic char-
acteristics on farmers’ perception of risk and choice of risk management strategies. The insights
provided will improve knowledge on how small-scale pond farmers cope with risks. Further-
more, a better understanding of farmers’ perceived risks and management strategies may help
policy makers, financial institutions, input providers, and fish traders to adjust their future strat-
egies, adding to a more productive and efficient farming sector for the benefit of the whole value
chain, which will in turn improve food security and alleviate poverty in developing countries.
MOTI VATI ON OF T HE ST UDY

A literature survey on risks in aquaculture indicated that most of the existing research has been
conducted on assessing ecological risks, disease-related risk, and risks of climate change on aqua-
culture (Kabir et al. 2017; Kabir, Alauddin, and Crimp 2017; Chitmanat et al. 2016; Lebel, Lebel,
and Lebel 2016). In contrast, several studies have focused on risk related to productivity and eco-
nomic efficiency (Khan, Guttormsen, and Roll 2018; Sarker et al. 2016; Tveteras 1999) and price
risk related to profitability (Asche, Sikveland, and Zhang 2018; Guttormsen 1999). Studies related
to risk management strategies include Watson et al. (2018), Quagraine, Kuethe, and Engle (2007),
and Bergfjord (2007), looking at contracts and insurance as a risk mitigating tool. Though aqua-
culture farmers’ risk perceptions play an important role in decision-making processes, only a few
studies (Ahsan 2011; Le and Cheong 2010; Bergfjord 2009) have comprehensively analyzed
sources of risk, perceptions of risks, and risk management strategies in aquaculture. Furthermore,
only Ahsan (2011) and Le and Cheong (2010) focus on risk perception in aquaculture within a
developing-country context.
This study differs from other studies of risk in pond aquaculture in the sense that it focuses
on extensive inland freshwater pond aquaculture producing pangas and tilapia mainly aimed for
consumption in the domestic market. In contrast, Ahsan (2011) and Le and Cheong (2010) fo-
cused on shrimp and catfish produced in intensive production systems that use more sophisti-
cated inputs (feed, seeds, and medicine) and that aim products to export markets. Furthermore,
regional heterogeneity in perceived risks and management strategies is included while estimating
effects on farming attitudes, sociodemographic characteristics, and past experiences of perceived
risks and management strategies.
Perceived Risk and Risk Management Strategies in Pond Aquaculture | 45
To advise farmers and policy makers on risk mitigating strategies, it is important to under-
stand in depth farmers’ risk-taking behavior to be able to design strategies and policies that sup-
port farmers’ needs.
This study is structured as follows: the next section introduces the Bangladesh pond aqua-
culture industry. The third and fourth sections report survey data and provide a description of
the method. The fifth and sixth sections describe and discuss the results. The seventh section
concludes the paper.
B AN GL ADE SH POND A QUAC U L TU R E IN DU STR Y

Bangladesh is the world’s fourth-largest aquaculture-producing country (FAO 2018). Aquacul-
ture production has been steadily growing during the last decades with increasing contributions
from pond-raised species like pangas and tilapia. Figure 1 shows the development of aquaculture
production from 2005 to 2018, including the contribution from tilapia and pangas, covering
44% of pond production in 2018. The two species have been successful because they are easy
to grow, have relatively low production costs, and are resistant to disease, and because there is
a large domestic market for low-priced fish. The farms are mainly small homestead earthen ponds
(80%), and the production is consumed domestically (Rahman, Nielsen, and Khan 2019).
Some of the difficulties that farmers face are relatively high input prices (such as for seeds,
feeds, fertilizers, disinfectants, and labor), poor quality of inputs, restricted access to markets and
to formal credit systems, poor transport facilities, and lack of supportive public policies.
Figure 2 shows the real price for tilapia and pangas from 2010 to 2019. Prices have been rela-
tively constant from 2014 to 2018. However, prices have fallen during the period of data collection
Figure 1. Historical Production Trend of Inland Capture, Pond Aquaculture, Total Aquaculture, Marine Cap-
ture, Tilapia (2009–18), and Pangas (2009–18) in Bangladesh. A color version of this figure is available online.
Figure 2. Retail Price Trend of Pangas and Tilapia Fish in Bangladesh (at real price). A color version of this
figure is available online.;
for this study (2017). In figure 3, annual percentage changes in costs of the most important
inputs—feed (80%), fingerlings (12%), and labor (6%)—and output values are shown per hectare
from 2014 to 2018. The figure shows that changes in output values (revenues) are lower than
changes in costs of some of the inputs, especially labor and feed, which can put profit margins un-
der pressure if farms are not continuously able to increase productivity (Rahman et al. 2020).
A few large feed producers dominate the feed markets in Bangladesh and have higher bar-
gaining power than the farmers. Consequently, farmers who are financially and institutionally
credit constrained—that is, most farmers in Bangladesh—are dependent on feed sellers, paying
higher prices for feed than the average market price while buying feed on credit (Islam et al.
Figure 3. Annual Percentage Changes in Input Costs (Feed, Fingerlings, and Labor) and Output Values at
the Farm Level (per hectare at 2014 constant). A color version of this figure is available online.
2020). Furthermore, farmers often receive cash credits from feed sellers and wholesale fish buy-
ers on the condition that they will buy feed from and sell fish to them, respectively, affecting the
farmers’ profitability and productivity (Mitra, Khan, and Nielsen 2019).
Furthermore, lack of proper aquaculture training and farm management knowledge often
lead the farmers to follow improper farm management practices (Rahman et al. 2020; Watson
et al. 2018; Chitmanat et al. 2016), which cannot efficiently mitigate business risks.
DATA
Before conducting the survey, a literature survey covering scientific articles and information
from governmental published reports was conducted to add to the understanding of farmers’ risk
perceptions and management strategies within the context of Bangladesh. After constructing the
interview schedule based on the knowledge obtained, preliminary interviews with 10 farmers
were conducted to validate the schedule and obtain information on the potential risks and risk
management strategies that might have been overlooked. Finally, a pilot survey of 20 farmers
was conducted in order to further validate the relevance of the questions and to identify ambig-
uous or missing questions. For instance, questions addressing the issues of water quality, soil qual-
ity, natural conditions, and some other environmental factors were dropped following responses
and views at this stage, since most farmers do not know the water and soil quality and because they
have neither the knowledge nor the instruments to check it regularly. Furthermore, farmers are
less concerned about environmental change/impacts on farming. Most of the questions were de-
signed as closed questions, mainly in the form of five-point Likert scales and dichotomous ques-
tions with the possible responses of yes or no. The questionnaire included questions regarding the
following issues: (1) farming attitudes, farm characteristics, the farmers’ sociodemographic vari-
ables, and the economic performance of the farm; (2) farmers’ perceptions of risk (including dif-
ferent sources of risks); and (3) farmers’ perceptions of various risk management strategies.
The data were collected from aquaculture pond farmers engaged in pangas and tilapia pro-
duction in Bangladesh. The data were collected using face-to-face interviews. A cluster sampling
technique was used, because it saves time and is more convenient when collecting data from a
huge, geographically scattered area (Levy and Lemeshow 1991). However, there can be problems
due to sample homogeneity. Nevertheless, in this case, the data seem reasonably heterogeneous
within the clusters because of differences in socioeconomic status and other variables (e.g., dif-
ferent education levels, income levels, trained/nontrained farmers, and experience). Therefore,
considering all the variables collected, it is reasonable to believe that statistical accuracy and valid-
ity are in order. In total, 645 farms were randomly selected for the survey. The seven selected
districts cover 57% of the tilapia and 82% of the pangas production in Bangladesh. Figure 4 shows
the districts from which the data were collected.
In total, 27 sources of risks and 36 risk management strategies found to be relevant to pond
aquaculture were included in the questionnaire. Farmers were asked to score each source of risk
on a Likert scale from 1 (no or very low impact) to 5 (very high or severe impact) to express how
significant they considered each source of risk to be in terms of its potential impact on economic
performance. Furthermore, the likelihood of occurrence of each risky event was reported on a
Likert scale from 1 (never) to 5 (very often). Thus, the perceived risk scores were calculated by
multiplying the associated scores of perceived probability (likelihood of occurrences) and impact
of the various sources of risk. The farmers were also asked to indicate their perceived importance
Figure 4. Study Areas. A color version of this figure is available online.
of each risk management strategy on a Likert scale from 1 (no or very negligible effect) to 5 (very
significant effect).
Table 1 includes the mean values of farm and farmers’ characteristics, highlighting the socio-
economic profiles of the farmers. These variables are later used as independent variables in a
multivariate analysis to see how they affect farmers’ perceptions of risks and risk management
strategies.
ME THOD
Rowe (1977, 24) defines risk as “the potential for realization of unwanted negative consequences
of an event.” The characteristics of a risk event are, therefore, a choice of action, a perceived
magnitude of loss, and the chance of realizing the loss.
Every risk has two domains: objective risk and subjective risk. An objective risk is a situation in
which an individual knows the probability of an event and the consequences of that probability
Table 1. Characteristics of Farms and Farmers in Bangladesh
Variable Name Farms in Survey

Characteristics (Description of Variables) (Independent Variables) (n p 645)
Personally owned farms (%) OwnerF 90

Personally owned land (%) OwnerL 46
Equity financed farms (%) SourceF 94
Agricultural land converted to aquaculture farms (%) LandCon 83
Age of farm operator (years) Age 40
Full-time fish farmers (%) OccM 91
Years of education (mean) Edu 10
Years of experience (mean) Exp 11
Trained farmers (%) Train 55
Large farms (area ≥ 2.0 hectares)a DummyL 27
Medium farms (area ≥ 0.5 ! 2.0 hectares)a DummyM 52
Small farms (area ! 0.5 hectare)a DummyS 22
Family size (mean)b FamilyS 6
Household income (mean) (in USD; 1 BDT p 80 USD)c LnIncome 12,717.49
Note: a Farm size refers to the total land area of ponds under harvest in the relevant year. b Family size refers to
a single family living in the same house and does not include extended family. c Household income refers to the
yearly net income of the family and is expressed in USD.
(either positive or negative; Wolff, Larsen, and Øgaard 2019). A risk is subjective if the probability
of a risky event and the consequences are measured based on the judgment of an individual
(Sjoberg, Moen, and Rundmo 2004; Sjoberg 1998). Risk perception (individuals’ judgment) is af-
fected by a large number of factors, including heuristics, socioeconomic factors (e.g., age, educa-
tion, income, and gender), the voluntariness of the risk, the expectation of control, the severity of
the consequences, the equity of distribution of risk and benefits, and the perceived benefit itself
(Van Winsen et al. 2016; Pennings and Garcia 2001; Willock, Deary, and McGregor 1999).
Empirical evidence has shown that in a risky situation, an individual does not always behave
according to the key assumptions of expected utility theory (Bocqueho, Jacquet, and Reynaud
2014) and rather makes decisions based on a subjective estimation using a personal probability
of loss or gain (Slovic 1987; Sitkin and Pablo 1992). Thus, “subjective” risk estimates cannot be
ignored or underestimated in risk management, as it is the individual who makes the decision of
being a farmer and investing in the farm.
Psychometric models (Slovic 1987; Sjoberg 1998; Ahsan 2011) have been widely used to es-
timate risk perception and the influence of perceived risks in management strategies in several
research studies, including behavioral economics and business management. In this model, dif-
ferences in risk attitudes across domains can be attributed to a different perception of risk (We-
ber, Blais, and Betz 2002; Weber and Milliman 1997). An individual’s economic behaviors are
determined by the perceived economic environment (Lien et al. 2006), which changes over do-
main and time. Furthermore, if variables representing the economic environment (such as farmer
and farm characteristics) are excluded from an objective model, it may lead to inappropriate pol-
icy recommendations (Bishu et al. 2018). In this paper, a psychometric model is used to ascertain
risks and risk management strategies of pangas and tilapia aquaculture farmers in Bangladesh.
A NALYTI C AL ME THODS
There are 27 sources of risks and 36 management strategies in the sample collected. The farmers’
perceptions of risks and risk management strategies are measured based on the impacts and
efficacies of risk sources and management strategies, respectively. The impacts of risks are ranked
by the mean values of risk level calculated by multiplying Likert scale responses to consequences
of an individual risk source with the corresponding responses to likelihood of occurrence (prob-
ability), in order to evaluate the perceived importance of various sources of risks. Similarly, the
risk management strategies are ranked to evaluate the perceived importance of an individual risk
management strategy. Standard deviation is used to evaluate the variability in farmers’ percep-
tions of both risk and risk management strategies.
There are different approaches to reduce dimensionality of variables and to investigate the re-
lationship of reduced variables with other variables, preferably partial least square regression
(PLS) and principal component regression (PCR). In PCR, the analysis follows a two-step pro-
cedure. In the first step, the observed variables are transformed into smaller numbers of variables
(commonly known as principal components or factors) using a principal component analysis
(PCA) technique. In the second step, the principal components derived from the PCA are used
as dependent variables in a multiple regression model (D’Ambra and Sarnacchiaro 2010). So,
in PCR, instead of regressing the dependent variable on the explanatory variables directly, the
principal components of the explanatory variables are used as regressors (Shaw 2003).
It should be noted that from a computational point, the mean and standard deviation of prin-
cipal components are 0 and 1, respectively (Duinen et al. 2015; Everitt and Hothorn 2011, 74).
Both PCR and PLS techniques help in solving multicollinearity problems, whereas PLS uses
more degrees of freedom than PCR (Krämer and Sugiyama 2011).
Thus, the reasons for using PCA in this study for extracting principal components and using
them in multiple regression models are that the correlations between independent variables (farm
and farmers’ characteristics) in this study are very low and the numbers of independent variables
are few (Jolliffe and Cadima 2016; Sawatsky, Clyde, and Meek 2015; Abdi 2003). The model pro-
vides robust results and performs at least as well as PLS. Furthermore, there are many empirical
studies in agriculture and aquaculture that follow similar methodology; these include Hayran
(2019); Cullen et al. (2018); Asravor (2018); Bishu et al. (2018); Duinen et al. (2015); Ahsan
(2011); and Le and Cheong (2010).
Accordingly, because of the large number of risk sources and management strategies, PCA
was employed to group the factors into smaller numbers of factors (e.g., categories of risk sources
and types of risk management strategies for this research) for subsequent analysis that can more
easily be interpreted and evaluated empirically. This has been done based on the latent root cri-
terion (eigenvalue ≥ 1).1 Furthermore, in order to have the most relevant and parsimonious set
of factors, PCA is repeated with varying numbers of factors (Hair et al. 2006).
To ensure maximum independence of the resulting factors, the orthogonal (varimax) rota-
tion extraction method was used. Missing responses were replaced with mean values of obtained
valid responses following Lien et al. (2006) and Flaten et al. (2005). The Kaiser-Meyer-Olkin
(KMO)2 test is used to check the suitability of data for principal component analysis (Cerny and
1. The eigenvalue indicates how much of the variance of the observed variables (27 sources of risk and 36 risk management
strategies) that a factor (categories of risk sources and types of risk management strategies) can explain. Any factor (risk category /
management strategy) with an eigenvalue ≥ 1 explains more variance than a single observed variable.
2. The formula for the KMO test is given by the following:
oi≠j rij
2
KMOj p ,
oi≠j ij 1 oi≠j uij
2
r
where rij is an element in the correlation martrix, and uij, is an element in the partial covariance matrix.
Kaiser 1977). A high KMO value (maximum 1.0) generally indicates that the PCA method is useful
when aiming to reduce the number of factors. In PCA, factors with factor loadings (which indi-
cate the relationships of each variable with the underlying factors) greater than 0.30 are considered
to be significant (Le and Cheong 2010; Ahsan 2011).
Principal component regression was used to identify relationships between farm character-
istics and farmers’ sociodemographic variables, risk perceptions, and risk management strategies.
In the regression model for risk perception, categories of risk sources (principal components/
factors extracted from 27 risk sources) are taken as dependent variables (table 6) and farm and
farmers’ characteristics are taken as independent variables. On the other hand, in the regression
for perception of risk management strategies, types of risk management strategies are taken as
dependent variables (i.e., the principal components/factors extracted from 36 management strat-
egies by PCA), and both farm and farmers’ characteristics including categories of risk sources are
taken as independent variables.
To account for regional heterogeneity, regional dummies are included in both regression
models, taking Mymensingh as the reference region since it is considered to be the cradle of aqua-
culture production in Bangladesh (Rahman, Nielsen, and Khan 2018). If heteroscedasticity is pre-
sent, robust regression (maximum likelihood method) is applied. The correlation coefficients
between all the pairs of independent variables have been estimated. However, all correlation
coefficients were low. All variance inflation factor (VIF) values are less than or equal to 2.0, in-
dicating that there are no multicollinearity problems within the studied variables. The regression
models are constructed as follows:
RF i p b0 1 b1 OwnerF 1 b2 OwnerL 1 b3 LandCon 1 b4 Age 1 b5 OccM
1 b6 Edu 1 b7 Exp 1 b8 Train 1 b9 DummyL 1 b10 DummyM (1)
1 b11 FamilyS 1 b12 LnIncome 1 bDi 1 ε,
RMSj p b0 1 b1 OwnerF 1 b2 OwnerL 1 b3 LandCon 1 b4 Age 1 b5 OccM
1 b6 Edu 1 b7 Exp 1 b8 Train 1 b9 DummyL 1 b10 DummyM (2)
1 b11 FamilyS 1 b12 LnIncome 1 bRF i 1 bDi 1 ε,
where the following applies:
RFi: Standardized principal components scores defined as categories of risk sources (i p

production risk, financial risk, market-related risk, and institutional risk). The principal
components/factors and their scores are derived by extracting observed 27 risk
sources using the PCA technique.
RMSj: Standardized principal components scores defined as types of risk management
strategies ( j p financial support, farm management, quality control, marketing and
logistics, farmers’ cooperation, extension and collaboration, diversification, disease
management, and input supply). The principal components/factors and their scores
are derived by extracting observed 36 risk management strategies using the PCA
technique.
Di3: Regional dummies (i p Mymensingh [reference region], Bogura, Comilla, Khulna,

Jessore, Chittagong, and Bhola).
RESULTS
The data are analyzed at four stages: (1) descriptive statistics of farm and farmers’ characteris-
tics, (2) risk factors and risk management strategies, (3) principal component analysis technique
extracting principal components, and (4) multiple regression, where the extracted principal
components are used as dependent variables and farm and farmers’ characteristics are used
as independent variables.
According to the hypotheses stated in the introduction, farmers’ socioeconomic profiles may
influence their perceptions of risk and risk management. Thus the economic profile—in terms
of farm ownership, primary occupation (OccM), level of training (Train), and the state of the
land before it was converted into a fish farm—is taken into consideration to identify farming
attitudes and commitments to fish farming. Two pond types are used: homestead ponds (used
for fish farming and household purposes) and commercial ponds (used only for fish farming).
Table 1 highlights the socioeconomic profiles of the farmers. Ninety percent of the farms are
personally owned, and 46% of the farms use personally owned land for fish farming. Farm own-
ership (OwnerF) and pond land ownership (OwnerL) have statistically significant influences on
perception of risk and risk management strategies. Eighty-three percent of the aquaculture
ponds in the sample are developed by converting agricultural land. Interviews under the field
study indicate that a common motivation for this conversion is higher returns for fish farming
than for agriculture. The main concerns when converting farmland to fish ponds are liquidity
and limited access to credit (Khan, Guttormsen, and Roll 2018) needed to be able to buy finger-
lings, feed, and medicine.
F ARM ERS ’ P ER C EPT I ON S O F R IS K S OUR C E S

Table 2 presents the 27 risk sources and their ranking, mean scores, and standard deviations.
The source of risk that the farmers found most important (highest mean) is ranked as number 1,
and the subsequent risk factors are presented in descending order. The standard deviations for all
the risk factors are approximately 1.0, indicating a high level of consensus among the respondents.
Price variability has a mean value of 3.76 and is considered the greatest risk to pond aqua-
culture in Bangladesh; this finding is similar to the findings for catfish farming in Vietnam (Le and
Cheong 2010) and shrimp in Bangladesh (Ahsan 2011). Interviewed farmers reported that in the
last few years (especially in the year of data collection, 2017), they have experienced losses due to
fluctuating prices of fish and at the same time increasing costs of the main inputs (e.g., feed, fin-
gerlings, and labor). This situation often motivates farmers to partially or temporarily withdraw
from production.
3. Mymensingh is considered the cradle of fish farming because of the availability of inputs including seeds, labor, feed, fertile
soil, and key research institutions; however, the region lacks access to fresh water. Bogura is known for housing many small and
medium-size industries including feed mills and hatcheries; it has good road connectivity but lacks fresh water. Comilla is popular
for its road connectivity with the larger cities and housing export processing zones, making labor scarcer. Chittagong and Khulna
are closer to the coast; water is abundant and inputs are more available since Chittagong is a business capital. Two seaports are also
located in these two regions. Bhola is surrounded by larger rivers with access to open water; however, it lacks access to markets and
quality seeds and feeds. Jessore is a small city and popular for hatcheries, but it has limited access to open water.
Table 2. Means, Standard Deviations, and Ranks of Risk Consequences
Risk Sources Rank Mean Std. Dev.
Fish price variability 1 3.763 1.065

High death rate due to disease 2 3.743 1.090
Low quality of fingerlings (not healthy) 3 3.606 1.040
Low quality of feed 4 3.575 1.128
Timely supply and price of fingerlings 5 3.543 1.047
Lack of own capital 6 3.530 0.983
Fingerlings infected by diseases 7 3.509 1.126
Price variations of feed and chemicals 8 3.439 1.086
High cost of operating inputs/farm equipment 9 3.409 1.047
Lack of credit from banks/financial institutions 10 3.384 1.085
Lack of storage and transportation facilities 11 3.364 1.117
Low credit availability 12 3.361 1.043
Inappropriate size of fish harvest 13 3.326 1.060
Low awareness of disease prevention by farmers 14 3.240 0.984
Use of undersized fingerlings 15 3.222 0.992
Limited knowledge about the usage of chemicals and medicines 16 3.216 1.012
Using chemicals and medicines improperly 17 3.206 0.986
Overfeeding, causing pollution and waste accumulation 18 3.203 1.081
Over (density) stocking fingerlings 19 3.177 0.989
High interest rate for loans 20 3.161 1.052
Changes in consumer preferences 21 3.141 0.946
Fingerlings with unknown origin 22 3.135 0.984
Supply of microcredit from NGOs 23 3.084 1.012
Supply of private capital (debt, equity) 24 3.051 0.973
Exploitation from middlemen 25 2.949 0.939
Weak enforcement in conducting sales contracts with processors 26 2.887 1.002
Farms have no reserved areas for waste management 27 2.808 1.089
Other factors that receive a high score are high death rate due to diseases (2), low quality of
fingerlings (3), and fingerlings infected by diseases (7), which all stress that disease problems are
major risk factors. Furthermore, low quality of feed (4) also seems to affect farms. In extensive
pond farming, the two primary cost-return drivers are feed and fingerlings because the future
possible harvest quantity is closely related to the quality and management of these two inputs.
Anwar (2011) also found that fish price variability and fish diseases were common risk factors
for pangas and tilapia fish farmers in Bangladesh.
Another issue that seems to be highly relevant is the availability of capital or credits. Lack of
own capital (6), lack of credit from banks/financial institutions (10), and low credit availability
(12) are found to be major challenges for farmers. This supports the findings of Ahmed, Alam,
and Hasan (2010) that indicate that farmers may use low-quality feeds (4) and fingerlings (3, 5,
and 7) because of financial constraints.
Inappropriate size of fish harvest (13) is another source of risk. This finding is supported by
information that farmers provided during the interviews; they claimed that they have to sell un-
dersized fish because of financial constraints.
Furthermore, overfeeding (ranked 18) and fingerling density (ranked 19) are also identified
as major sources of risk. It seems that farmers who are not suffering from credit constraints are
facing problems of overfeeding and overstocking due to the lack of knowledge of input manage-
ment (Alam, Khan, and Huq 2012). This finding is also supported by Khan, Guttormsen, and
Roll’s (2018) study of production risk in pangas farming.
To provide a deeper understanding of the farmers’ perceptions of risk, the principal compo-
nent extraction method in combination with the varimax rotation method has been applied. Us-
ing these methods, the 27 sources of risks are reduced to four main risk factors (termed later as
categories of risk sources): production-related risk, financial risk, market-related risk, and insti-
tutional risk, shown in table 3. An individual source of risk is assigned to a category of risk source
based on its corresponding factor loading value (highest). These four categories of risk sources
have eigenvalues exceeding 1.0 and total variance of 60%. The applicability of PCA is verified us-
ing the KMO test of sample adequacy and Bartlett’s test of sphericity. The KMO measure of ad-
equacy is 0.96, and Bartlett’s test of sphericity is statistically significant at 1% (table 3). All of the
tests indicate that PCA is a good fit for this dataset. Sources of risk having factor loadings above
0.30 are considered to be important (Flaten et al. 2005). Production-related risk shows high load-
ing for most factors related to production inputs, such as fingerlings, feed, chemicals, and med-
icines. Financial risk includes factors like supply of credit and capital, with the highest loading
including changes in consumer preferences for fish, among others. Marketing risk includes fac-
tors like fish price variability. Institutional risk includes factors like weak enforcement of sales
contracts with producers and waste management with highest factor loading.
F ARM ERS ’ PERCEPTI ONS OF RISK MA NAG EMENT ST RAT EGI ES

Farmers rated the efficacy of 36 risk mitigating strategies on a five-point Likert scale, with 1 as a
negligible effect and 5 as a very significant effect. Table 4 presents the average scores of the ef-
ficacies and their ranks in descending order. The risk management strategies have mean scores
between 3.1 and 4.0. In general, inputs of good quality, such as feed, fingerlings, and water, are
perceived as the most effective risk mitigating strategies, and the choice of a good brand of feed is
considered the best risk management strategy. These findings are consistent with the findings of
Alam, Khan, and Huq (2012), who point out that farms using commercial feed are more efficient,
and Ahmed (2007), who points towards commercial feed as the most effective way of cultivating
fish if efficiently managed. However, most small and medium-sized farms use homemade feed
(traditional) because it is less costly, and feed constitutes the most important input in terms of
costs (Alam, Guttormsen, and Roll 2019; Ali, Rahman, et al. 2018).
Selecting good-quality fingerlings and buying from reliable sources are ranked as the 2nd
and 4th most effective strategies. In Bangladesh, fingerlings are most often provided from three
sources: own production, small traders of fingerlings, or larger fingerling agents (Hernandez et al.
2018). Farmers with small and medium-sized farms mostly depend on small fingerling traders
that provide low quality but at a relatively low cost, negatively affecting productivity (Hernandez
et al. 2018).
Available market information and an appropriate price policy (price floor, taxes, etc.) are
ranked 6th and 8th. There is no formal arrangement to provide market information, which could
benefit the farmers. Attending workshops or training is ranked 10th and is provided by the Fish-
eries Department, but resources for this extension service are limited. Fifty-five percent of the
surveyed farmers have received training, but this is often limited to one time and is quite short
(one day to one week), which means that there is no continuous training and updating of skills
and knowledge at the farmers’ level. This leads farmers to follow “farming by seeing” and “learning
by doing,” which affects farm productivity and profitability (Rahman, Nielsen, and Khan 2018).
The study by Hernandez et al. (2018) notes that improving aquaculture-supportive infrastructure,
Table 3. Varimax Rotated Factor Loadings for Risk Sources
Principal Components / Factorsa

(Categories of Risk Sources)
Production Financial Market- Institutional

Risk Risk Related Risk Risk
Risk Sources Communalitiesb (1) (2) (3) (4)
Low quality of fingerlings (not healthy) 0.648 0.682 0.240 0.351 –0.032
Fingerlings with unknown origin 0.622 0.607 0.368 0.116 0.324
Timely supply and price of fingerlings 0.571 0.706 0.129 0.117 0.205
Fingerlings infected by diseases 0.600 0.692 0.269 0.217 0.047
Over (density) stocking fingerlings 0.639 0.704 0.319 0.012 0.203
Use of undersized fingerlings 0.525 0.548 0.166 0.261 0.359
Low quality of feed 0.609 0.654 0.255 0.339 0.035
Overfeeding, causing pollution and
waste accumulation 0.669 0.714 0.280 0.110 0.260
Using chemicals and medicines improperly 0.515 0.503 0.339 0.268 0.275
Price variations of feed and chemicals 0.690 0.619 0.124 0.493 0.221
Limited knowledge about the usage of
chemicals and medicines 0.540 0.477 0.183 0.259 0.460
High death rate due to disease 0.628 0.554 0.140 0.552 0.076
Low awareness of disease prevention
by farmers 0.569 0.473 0.255 0.471 0.388
Farms have no reserved areas for waste
management 0.701 0.212 0.143 –0.013 0.797
Inappropriate size of fish harvest 0.533 0.228 0.466 0.386 0.337
Fish price variability 0.682 0.263 0.236 0.737 0.117
Weak enforcement in conducting sales
contracts with processors 0.714 0.104 0.069 0.032 0.835
Exploitation from middlemen 0.387 0.139 0.344 0.208 0.454
Lack of storage and transportation facilities 0.556 0.520 0.340 0.355 0.209
High cost of operating inputs/ farm
equipment 0.514 0.434 0.537 0.438 0.105
Changes in consumer preferences 0.504 0.206 0.606 0.136 0.275
Lack of own capital 0.590 0.145 0.727 0.270 –0.024
Lack of credit from bank/institutions 0.593 0.263 0.685 0.222 0.071
Supply of private capital (debt, equity) 0.630 0.345 0.706 0.092 0.066
Supply of microcredit from NGOs 0.683 0.260 0.771 0.133 0.049
High interest rate for loans 0.568 0.162 0.611 0.343 0.225
Low credit availability 0.594 0.218 0.610 0.378 0.178
Percentage of total variance explained 44.021 6.152 5.528 3.822

Percentage of total cumulative variance
explained 44.021 50.173 55.701 59.523
KMO test 0.961

Bartlett’s test of sphericity chi-square value at 1% with df 351 9,506
Note: a Categories of risk sources are production, financial, marketing, and institutional. The categorizations
are based on the factor loading (loadings 1 0.30) of individual risk sources. The highest loading values are given in
bold. b Indication of how much variance is explained by the corresponding source of risk out of the total variance.
such as establishing storage facilities (9th) and improving transportation facilities (20th), may im-
prove the contribution of aquaculture to the national economy of Bangladesh.
Similar to the sources of risk, the numbers of risk mitigating strategies are reduced by PCA.
Principal component extraction in combination with the varimax rotation method is applied.
Table 4. Efficacies of Risk Management Strategies: Ranks, Means, and Standard Deviations
Risk Management Strategies Rank Mean Std. Dev.
Choose a good brand of feed 1 3.957 0.935

Select good-quality fingerlings 2 3.927 0.870
Maintain a well-managed water environment in pond 3 3.816 0.961
Buy fingerlings from reliable sources 4 3.741 0.817
Carefully check the fingerlings when buying 5 3.732 0.946
Acquire available market information 6 3.722 0.947
Obtain timely supply of fingerlings 7 3.690 0.786
Request appropriate price policy 8 3.661 0.911
Establish storage facilities 9 3.611 0.857
Attend workshops or training 10 3.594 0.919
Apply medicines, chemicals to prevent disease 11 3.575 0.868
Keep cash on hand for farming 12 3.566 0.894
Produce at lowest possible costs 13 3.564 0.911
Prevent disease by regularly checking and observing pond 14 3.550 0.934
Strictly follow the recommended guide 15 3.526 0.891
Use large-size fingerlings 16 3.487 0.906
Buy fingerlings only from certified producers 17 3.474 0.839
Secure bank loans 18 3.461 0.954
Enter into sales contract with middlemen/processors 19 3.454 0.927
Improve transportation facilities 20 3.447 0.895
Involve production contract with predetermined size 21 3.442 0.846
Collaborate with association of fish farmers 22 3.434 0.938
Reduce farm size to appropriate scale 23 3.431 0.909
Use only factory-made (pellet) feed 24 3.428 0.869
Remove influence of middlemen 25 3.409 0.878
Reduce density of fingerling stocking 26 3.402 0.852
Secure available microcredit 27 3.371 0.913
Collect information on consumer preferences 28 3.366 0.809
Involve off-farm work 29 3.363 0.884
Use laborers with aquaculture knowledge 30 3.305 0.873
Check prohibited substances (hormones, chemicals) 31 3.293 0.886
Use economic consulting services 32 3.259 0.950
Diversify products 33 3.257 0.875
Engage in farmers’ cooperative association 34 3.251 0.850
Increase cooperative marketing 35 3.223 0.942
Share ownership of equipment/partnership 36 3.164 0.903
The KMO is 0.80, and Bartlett’s sphericity is statistically significant at the 1% level (table 5). A
total of nine factor4 solutions (types of risk management strategies) are selected, accounting for
51% of the total variance. Based on the factor loadings, the nine types of risk management strat-
egies and their respective loadings are presented in table 5. The risk management strategies are
financial support, farm management, quality control, marketing and logistics, farmers’ cooper-
ation, extension and collaboration, diversification, disease management, and input supply. The
most appropriate risk mitigating strategies have the highest factor loadings (table 5).
Financial support has the highest loadings for keeping cash on hand for farming (0.718) and
share ownership of equipment/partnership (0.564), both of which make perfect sense as risk
4. The statistical package by default suggested 12 factors with a total variance of 58%. However, for ease of interpretation, the
package directed a nine-factor solution. The cost is only the loss of 7% (58% – 51%) of the total variance.
mitigating strategies. Other strategies with loadings related to financial support are assurance of
available microcredit (0.488), off-farm work (0.406), and assurance of bank loans (0.392).
Farm management is mostly related to the control of inputs such as choosing a good brand of
feed (0.687), selecting good-quality fingerlings (0.631), buying fingerlings from reliable sources
(0.558), using large-size fingerlings (0.434), and having a well-managed water environment in
the pond (0.421).
Quality control is related to activities controlling the quality of inputs, such as using laborers
with aquaculture knowledge (0.702), using only factory-made feed (0.523), strictly following the
recommended guide (0.433), and checking chemicals (0.425). Similarly, all the other risk man-
agement strategies are named based on the factors (risk management strategies) with higher fac-
tor loadings (see table 5).
RELATI ONSHIP S A MONG PERCEPTI ONS OF RISKS A ND FARM

CHARACT ERI STI C S
The relationships among the farmers’ risk perceptions, farm characteristics, and the farmers’ so-
cioeconomic characteristics are analyzed using multiple regression (equation 1). The results are
presented in table 6. Four models are regressing farm and farmers’ characteristics with each cat-
egory of risk source. Regional dummies are included as well to see whether the risk perceptions
significantly vary between regions. All the models are statistically significant at the 1% level and
are homoscedastic; only the model for market-related risk shows signs of heteroscedasticity. Thus,
this model is reestimated using the robust regression maximum likelihood method (Fox and
Weisberg 2013). All of the models have low adjusted R2, indicating that some of the variables that
may explain the farmers’ perceptions of risk are not included in the models. However, earlier
studies on risk perception in both agriculture and aquaculture, including Hayran (2019), Cullen
et al. (2018), Asravor (2018), Patrick and Musser (1997), Meuwissen, Hurine, and Hardker (2001),
Flaten et al. (2005), Le and Cheong (2010), and Ahsan (2011), also observed very low R2 values.
Durbin-Watson statistics for all four models are estimated following earlier studies (Le and
Cheong 2010) and are between 1.04 to 1.18 (table 6), suggesting that autocorrelation is not a prob-
lem. For the test of multicollinearity, the variance inflation factor is used. No multicollinearities
are found since all the values of VIF are approximately 2.00.
The results reveal that the perceived production risks are higher for farms that are made from
converting agricultural land (commercial ponds) into fish ponds. That means homestead ponds
have lower production-related risks such as diseases and so on. Furthermore, the volume of pro-
duction in homestead ponds is lower than in the commercial ponds. They are less sensitive to
production risk, since the lower the volume of production, the lower the potential loss when
something bad occurs at a farm (Le and Cheong 2010).
Perceived financial risk is higher for the personally owned pond land and farms than for farms
that have rented land and shared ownership. Larger and medium-sized farmers also assume higher
financial risks than the small farmers do, perhaps due to higher capital requirements. It is also higher
for trained farmers. Furthermore, farm operators who are more educated perceive financial risks
to be lower than the less educated farm operators do.
Perceived marketing risk is higher for trained farmers. However, perceived marketing risk is
lower for personally owned and self-financed farms than for shared and debt-financed farms.
Furthermore, perceived institutional risk is higher for personally owned and self-financed farms
and lower for trained farm operators. Interestingly, the household incomes of farm operators have
Table 5. Varimax Rotated Factor Loadings for Risk Management Strategies
Principal Components / Factorsa (Types of Risk Management Strategies)
Financial Farm Quality Marketing Farmers’ Extension and Disease Input

Support Management Control and Logistics Cooperation Collaboration Diversification Management Supply
Risk Management Strategies Communalities (1) (2) (3) (4) (5) (6) (7) (8) (9)
Select good-quality fingerlings 0.515 0.004 0.631 –0.008 0.170 –0.100 0.134 0.085 0.150 0.173
Buy fingerlings from reliable sources 0.446 0.153 0.558 –0.037 –0.014 0.201 –0.080 0.139 0.064 0.199
Buy fingerlings only from certified
producers 0.423 0.164 0.124 0.042 0.067 0.040 –0.069 0.540 –0.072 0.267
Obtain timely supply of fingerlings 0.525 0.179 0.428 0.101 –0.036 0.040 0.050 0.006 –0.008 0.542
58
Carefully check the fingerlings when
buying 0.522 –0.012 0.211 0.090 0.107 0.043 0.003 0.174 0.028 0.651
Strictly follow the recommended guide 0.523 0.253 0.051 0.433 –0.351 0.146 0.105 0.159 0.266 0.133
Reduce density of fingerling stocking 0.507 0.168 0.204 0.496 –0.097 0.260 –0.148 0.491 0.094 0.018
Use large-size fingerlings 0.476 0.037 0.434 0.473 0.133 –0.010 –0.139 0.212 0.222 –0.127
Choose a good brand of feed 0.559 –0.018 0.687 0.128 0.250 0.071 0.016 0.036 0.014 –0.031
Check prohibited substances
(hormones, chemicals) 0.465 –0.004 0.197 0.425 0.114 0.417 –0.014 0.136 –0.103 –0.196
Use only factory-made (pellet) feed 0.419 0.162 –0.044 0.523 0.223 –0.011 0.154 0.122 0.151 0.079
Use laborers with aquaculture knowledge 0.528 0.036 0.056 0.702 0.111 –0.033 0.104 0.077 0.020 –0.017
Request appropriate price policy 0.561 0.185 0.117 0.410 0.502 –0.051 0.061 0.069 –0.021 0.284
Maintain a well-managed water
environment in pond 0.487 0.073 0.421 0.096 0.374 –0.108 –0.039 –0.068 0.365 –0.065
Apply medicines, chemicals to prevent
disease 0.667 0.038 0.153 0.101 –0.059 0.024 0.092 –0.110 0.774 0.091
Prevent disease by regularly checking
and observing pond 0.660 0.006 0.051 0.028 0.001 0.189 –0.168 0.186 0.747 –0.008
Attend workshops or training 0.499 0.561 –0.028 0.154 0.310 0.064 –0.132 –0.071 0.115 0.155
Table 5. (Continued)
Principal Components / Factorsa (Types of Risk Management Strategies)
Financial Farm Quality Marketing Farmers’ Extension and Disease Input

Support Management Control and Logistics Cooperation Collaboration Diversification Management Supply
Risk Management Strategies Communalities (1) (2) (3) (4) (5) (6) (7) (8) (9)
Involve production contract with

predetermined size 0.384 0.335 0.057 0.030 0.150 –0.026 0.294 –0.053 0.393 –0.017
Enter into sales contract with middlemen/
processors 0.506 –0.084 0.016 –0.031 –0.055 0.469 0.473 0.102 0.170 –0.111
Collaborate with association of fish
farmers 0.549 –0.138 0.066 0.081 0.082 0.594 0.209 0.265 0.212 –0.021
Acquire available market information 0.550 0.308 0.203 –0.019 0.331 0.507 0.103 –0.129 0.050 0.133
Increase cooperative marketing 0.638 0.048 –0.127 0.068 –0.024 0.782 –0.012 0.042 –0.017 0.006
Engage in farmers’ cooperative association 0.521 –0.023 –0.135 0.097 0.022 0.107 0.683 0.098 0.021 –0.064
Remove influence of middlemen 0.375 0.094 –0.301 0.047 0.156 0.223 0.250 0.325 0.123 0.126
Establish storage facilities 0.540 0.211 0.136 0.042 0.642 0.221 –0.069 –0.043 –0.079 –0.035
Improve transportation facilities 0.525 –0.048 0.088 0.008 0.546 0.026 0.319 0.309 0.094 0.102
Produce at lowest possible costs 0.448 0.072 0.178 0.189 0.607 0.019 0.012 0.071 0.034 –0.015
Collect information on consumer
preferences 0.476 0.231 0.197 0.154 0.615 0.101 0.160 0.178 –0.055 –0.538
Diversify products 0.525 –0.052 0.063 0.129 0.093 0.113 0.302 0.614 –0.064 –0.089
Reduce farm size to appropriate scale 0.319 0.172 0.411 0.058 –0.236 0.095 0.401 –0.063 0.104 –0.034
Secure bank loans 0.427 0.392 0.067 0.155 0.057 –0.082 0.100 0.321 0.249 –0.244
59
Involve off-farm work 0.417 0.406 0.204 0.072 0.003 0.036 0.288 0.239 –0.076 –0.242
Secure available microcredit 0.571 0.488 –0.018 0.048 –0.002 –0.019 –0.082 0.501 0.189 –0.192
Use economic consultant services 0.499 0.256 0.024 0.036 0.168 –0.068 0.600 –0.039 –0.180 0.063
Keep cash on hand for farming 0.610 0.718 0.111 0.089 0.191 –0.087 0.104 0.130 0.050 –0.006
Share ownership of equipment/
partnership 0.503 0.564 –0.042 0.297 –0.131 0.094 0.222 0.091 –0.102 0.036
Percentage of total variance explained 15.468 6.612 5.491 4.902 4.206 3.716 3.553 3.198 3.056
Percentage of total cumulative variance
explained 15.468 22.08 27.571 32.472 36.678 40.394 43.947 47.145 50.201
KMO test 0.801

Bartlett’s test of sphericity chi-square value
at 1% with df 703 5,298
Note: a The categorizations are based on the factor loading (loadings 1 0.30) of individual risk sources. The highest loading values are given in bold.
Table 6. Results of Multiple Regression for Categories of Risk Sources
Dependent Variables (Categories of Risk Sources That Are Derived from PCA)
Independent Variables Production Risk Financial Risk Market-Related Risk Institutional Risk
Intercept 1.790*** 1.283*** 0.820*** 0.824*

OwnerF –0.150 0.162* –0.151** –0.151
OwnerL –0.047 0.239*** 0.074 0.348***
SourceF 0.130 –0.010 –0.355** 0.300*
LandCon 0.280** 0.156 0.145 0.047
Age 0.003 –0.007* –0.003 0.003
OccM 0.039 –0.038 –0.033 –0.350**
Edu 0.007 –0.019** 0.004 –0.014
Exp 0.004 0.007 0.009 –0.003
Train 0.078 0.010* 0.233*** –0.167**
DummyL 0.181 0.412*** 0.079 0.285
DummyM –0.047 0.221* 0.028 0.221
FamilyS –0.002 –0.033*** 0.016 –0.013
LnIncome –0.181*** –0.121*** –0.141*** –0.063*
Bogura 0.121 –0.004 –0.400 –0.103

Comilla 0.207* 0.207* –0.067 –0.016
Khulna 0.200 –0.047 –0.447 0.296
Jessore –0.108 –0.136 –0.279* 0.044
Chittagong 0.147 0.169 –0.288* –0.021
Bhola 0.213 0.183 –0.243 0.580***
Adjusted R2 0.043 0.040 0.046 0.055

Durbin-Watson 1.042 1.098 1.130 1.181
Note: ***, **, and * indicate statistical significance at 1%, 5%, and 10% levels, respectively.
a statistically significant negative influence on the farmers’ perceptions of all the categories of risk
sources. Perhaps individuals with diversified income sources are less exposed to different types of
risks because of a higher ability to adjust to changing market requirements. For instance, they can
adjust harvest time when prices are low.
The economic implications of the above risk perceptions are that the farmers perceive more
risk, either because they objectively face more risk or because they have a higher subjective per-
ception of the risks. For instance, in the case of financial risks, farmers may not share ownership
because doing so would also mean sharing profit, which they objectively see as the same as facing
higher risks. On the other hand, smaller farms subjectively face lower financial risks. It is worth
mentioning that this paper is about subjective risk assessment. Furthermore, farmers who per-
ceive more risk may decide to reduce production to hedge against possible production, financial,
and marketing risk, or in the long run they may quit the industry. That may cause contraction of
the industry, reducing fish supply in the market.
RELATI ONSHIP S A MONG PERCEPTI ONS OF RISK MA NAG EMENT ST RAT EGI ES
A N D FA R M C H A R A C T ER I ST I C S
Corresponding to the risk sources, the relationships among the perceptions of risk management
strategies and farm characteristics and farmers’ socioeconomic characteristics are determined us-
ing multiple regression including regional dummies. Furthermore, the relationships among four
categories of risks and nine risk management strategies (types of risk management strategies) are
determined. All of the models are statistically significant at the 1% level, with R2 varying from
0.014 to 0.180, which seems to be very low (table 7). All the models are tested for heteroscedasticity
and misspecification by Breusch-Pagan and Ramsey RESET tests; models 1, 2, 3, 4, and 6 were
heteroscedastic and models 5, 7, 8, and 9 were found to be homoscedastic. Thus, models 5, 7, 8,
and 9 were reestimated by robust regression maximum likelihood method.
Land ownership, land condition, farm size, and household income have statistically negative
associations with financial support strategies. This indicates that these types of farmers consider
financial support to be less important. However, only farm ownership has statistically significant
positive associations with financial support strategies.
Farmers’ age and household income have a statistically positive association with farm man-
agement and a negative association of age with input supply. It is not unusual for older farmers
to put more emphasis on management than younger farmers do. On the other hand, younger
people tend to focus on input supply and suppliers, perhaps because younger farmers are gener-
ally more motivated to develop and build their business through strong business relationships
with suppliers and customers and are more open to the external business environment than older
farmers, who tend to emphasize internal farm management (May et al. 2019). This finding is sim-
ilar to that of Van Winsen et al. (2016), that older farmers are less inclined to use modern tech-
nology and inputs than younger farmers are.
Farmers having personal ownership consider quality control, extension and collaboration, and
disease management more important risk minimizing strategies compared with other types of
farm ownership. This may be because activities like quality control, extension and collaboration,
and disease management are more difficult than when the ownership is shared.
Land ownership, sources of funds, and household income have statistically significant positive
associations with extension and collaboration strategies. This indicates that the farmers with per-
sonally owned farmland, own capital, and higher household income consider extension and col-
laboration to be more important than their counterparts do. However, trained farmers consider it
to be less important than the nontrained farmers. This might be due to their higher ability com-
pared with that of the nontrained farmers to explore opportunities.
Medium to large farms put more emphasis on disease management. This may be because the
larger farms follow intensive production systems that are more prone to diseases (Alam, Guttorm-
sen, and Roll 2019; Khan, Guttormsen, and Roll 2018). Such perceptions may motivate the farmers
to follow extensive or semi-intensive farming systems, negatively affecting productivity. Medium-
size and more household income-generating farm operators consider farm diversification to be an
effective tool for risk mitigation.
Overall, knowledge of improved farm management, increased extension services, and supports
to control disease management (access to water, quality feeds and seeds) may improve productiv-
ity and profitability and support industry expansion.
Implications are that since risk-taking behavior influences risk management strategies, a risk-
averse farmer may prefer to act passively by downsizing production or the farm and by saving rather
than investing further. However, a risk-seeking (willing to take more risk) farmer may be proac-
tive through farm and income diversification and by linking to the market to optimize production
and profit.
The differences in perceived risks and management strategies also imply that individually de-
signed risk management strategies should be applied for different regions when national policies
are designed.
Table 7. Results of Multiple Regression for Types of Risk Management Strategies
Dependent Variables (Types of Risk Management Strategies That Are Derived from PCA)
Financial Farm Quality Marketing Farmers’ Extension and Disease

Support Management Control and Logistics Cooperation Collaboration Diversification Management Input Supply
Independent Variables (1) (2) (3) (4) (5) (6) (7) (8) (9)
(Constant) 1.017* –1.670*** –0.916* 0.984* –1.239** –1.278*** –0.358 –0.357 0.827
OwnerF 0.084* 0.046 0.554*** –0.076 0.146 0.214** –0.011 0.349** –0.089
OwnerL –0.089** –0.011 –0.070 –0.040 0.048 0.165* 0.049 –0.178 –0.026
SourceF –0.154 0.196 –0.190 –0.200 0.117 0.307** –0.177 0.168 0.013
LandCon –0.167*** 0.045 0.083 0.036 –0.115 0.045 –0.143 –0.090 –0.133*
Age 0.001 0.004** 0.002 –0.001 0.004 –0.004 –0.002 –0.002 –0.008**
OccM –0.054 0.093 –0.025 –0.160 –0.033 0.021 0.162 0.002 –0.016
Edu 0.009 0.011 –0.002 –0.006 0.008 –0.001 0.003 0.001 –0.007
Exp 0.005 –0.008 0.002 –0.000 –0.006 0.011 –0.001 –0.002 0.005
Train 0.015 –0.100 0.114 0.086 0.006 –0.228*** 0.007 –0.122 0.037
DummyL –0.307*** –0.092 –0.094 0.047 0.115 –0.134* 0.077 0.075** 0.067
DummyM –0.165** 0.059 0.067 0.044 –0.020 –0.026 0.007* 0.122*** 0.106
FamilyS –0.003 –0.008 0.008 –0.004 –0.019 –0.015 0.001 0.017 –0.002
LnIncome –0.073* 0.074*** 0.030 –0.022 0.058* 0.068*** 0.059* –0.002 –0.037
Production risk –0.115*** 0.386*** 0.009 0.072* –0.057 –0.117*** 0.010 –0.032 0.092*
Financial risk 0.203*** 0.118*** 0.152*** 0.003 0.010 0.019 0.199*** –0.004 –0.237***
Market-related risk 0.171*** 0.158*** 0.053 0.323*** –0.115*** –0.026 –0.160*** –0.015 0.207***
Institutional risk 0.032 0.013 0.115*** 0.027 0.073* 0.342*** 0.056 0.017 –0.014
Bogura 0.610*** 0.523** 0.159 –0.182 0.192 0.004 –0.411** –0.242* –0.167
Comilla 0.375*** 0.496** 0.189** –0.090 0.154 0.251* –0.382*** –0.348** –0.007
Khulna 0.349 –0.215 –0.483** –0.409 0.024 0.032 –0.308 –0.719** 0.292
Jessore –0.020 0.236 0.050 –0.514** –0.143 –0.086 –0.704*** –0.485** –0.389**
Chittagong 0.233 –0.118 0.233 –0.292** 0.353** 0.038 –0.612*** –0.557*** 0.016
Bhola 0.223 0.216 –0.116 –0.173 0.032 0.179 –0.540*** –0.365** –0.046
Adjusted R 2 0.145 0.162 0.049 0.120 0.016 0.180 0.053 0.014 0.098
Durbin-Watson 1.651 1.440 1.652 1.592 1.588 1.491 1.703 1.521 1.959
Note: ***, **, and * indicate statistical significance at 1%, 5%, and 10% levels, respectively.
R EGI ONA L HETER OGE NEIT Y IN PER CEIVE D R ISK A ND RI SK MAN AGEM ENT
STRA TEG IES
There are statistically significant differences in perceived sources of risk and management strat-
egies across the regions within the study. Farmers in Comilla perceive production and financial
risks to be more important than do farmers in the reference region (Mymentsingh). Farmers in
Bhola perceive institutional risks to be more important, while they consider marketing risks to
be less critical. Such differences in perceived risks may be due to differences in production en-
vironments, access to important inputs like water, feed, and fingerlings, availability of labor, and
access to markets and supporting infrastructure. Other studies, including Mitra et al. (2019) and
Rahman, Nielsen, and Khan (2018, 2019), indicate that differences in local production environ-
ments and access to inputs and markets influence farm productivity and efficiency.
Similarly, risk management strategies are perceived differently among the regions. Farmers
outside Mymensingh (base region) perceive farm diversification and disease control as less effec-
tive. Farmers in Comilla perceive financial supports, farm management, quality control, and ex-
tension as more effective risk management strategies, whereas farmers in Khulna, Chittagong, and
Jessore perceive quality control, marketing and logistics, and input supply strategies as less effec-
tive than the base region does. Similarly, earlier studies by Mitra et al. (2019) and Rahman, Nielsen,
and Khan (2018) found that there are interregional differences in farm density, access to extension,
and quality inputs that affect farm productivity and profitability.
D I S C U SS I O N
That perceived production risks are higher for farms that are made from converting agricultural
land (commercial ponds) may be due to intensive production systems and poor management of
the water environment in these ponds (Lebel, Lebel, and Chuah 2018).
Perceived financial risk is higher for personally owned farms than for farms that have shared
ownership. It may be that in case of shared farm ownership, risk can be shared among the owners,
which is not possible in cases of personally owned farms. Similarly, medium and large farms are
experiencing higher financial risk than the smaller farms. Investment and thereby risk are larger
for larger-scale farms, a finding reflected in Neira, Engle, and Ngugi (2009) for large-scale com-
mercial tilapia farms, and Khan, Guttormsen, and Roll (2018) for pangas.
More-educated farm operators and larger families perceive financial risks as less important.
More-educated operators may be able to obtain income from other sources and have more
knowledge on how to access other means such as institutional borrowing, whereas farm opera-
tion and financing activities may be shared in larger families.
Perceived marketing risks are lower for those having personal ownership, own-financed farm,
and higher household income because of the lower volume of production and alternative sources
of income. However, for trained farmers marketing risks are higher, which is also found in Ali,
Upraity, et al. (2018). The reason may be that marketing risk becomes more important if the
ownership is shared and if the farmers have debt and lower household income.
Perceived institutional risk is higher for farms in personally owned land and self-financed
farms. This may be due to limited access to institutional support. For example, these types of farm-
ers do not get access to the large processors who offer better prices. By contrast, full-time farm-
ers and trained farmers perceive institutional risk to be less important, perhaps because they are fully
devoted to fish farming.
The higher emphasis that operators of personally owned farms place on financial supports,
quality control, extension and collaboration, and disease management as risk mitigating strat-
egies indicate that shared farming (cooperative farming) can provide an improved solution to
the risk associated with these risk management strategies.
Production-related risks are significantly associated with multiple risk management tools like
financing, farm management, marketing and logistics, and extension and collaboration. It seems
reasonable since quality inputs, better management, and knowledge are commonly known as de-
terminants of higher yields. Since efficient input management can have positive impacts on yield
(Alam, Guttormsen, and Roll 2019; Khan, Guttormsen, and Roll 2018; Alam, Khan, and Huq
2012; Valderrama and Engle 2001), whereas marketing contracts can reduce price risk (Quagraine,
Kuethe, and Engle 2007), long-term strategies may include organizing small-scale farmers in a group,
which will help them to gain the advantages of economies of scale, scope, and flexibility in accessing
services and markets that are mostly limited to large-scale commercial aquaculture farmers.
The negative associations of household income with all of the categories of risk sources indi-
cates that farmers with higher incomes are relatively less dependent on the next harvest and fur-
thermore have the opportunity to invest more in risk mitigation measures, such as training, in-
formation, and specialized inputs, improving economic performance.
In the majority of risk management strategies, the strategies have significant influence on
farmers’ perceived financial risks. This means that the common motive of any risk management
strategy is to minimize financial loss since there are no formal financial risk management tools
like insurance. However, formal financial risk management tools can contribute to reducing pro-
duction, revenue, and environmental risks in aquaculture (Watson et al. 2018).
Marketing risk is considered to be very important. In this regard, formal marketing contracts
can effectively reduce price risk (Quagraine et al. 2007; Bergfjord 2007). Thus, efforts to link trad-
ers or retailers with farmers may reduce concerns over market risks. Institutions, especially the
extension department of the government, can help farmers to link up with retailers and traders
because farmers usually do not have the ability to link up with the downstream value chain.
Only quality control, farmers’ cooperation, and extension and education have significant in-
fluences on mitigating institutional risks. Thus, institutional risks can be reduced by implement-
ing appropriate quality control activities, forming farmers’ cooperatives, and improving exten-
sion and education services to farmers. Ahsan (2011) found that shrimp aquaculture farmers
place more emphasis on organization supports, collaboration, disease control, and diversification
strategies to reduce institutional risks.
Risk management is a dynamic and complex process. Like any other form of aquaculture,
farming of pangas and tilapia is also associated with diverse sources of risks. Therefore, to adopt
one particular risk management strategy is not always enough to minimize risks; rather, it de-
mands a set of multiple risk management strategies (Ahsan 2011; Le and Cheong 2010; Meu-
wissen, Hurine, and Hardker 2001). This study reveals that the farmers are aware of this, indicat-
ing the importance of multiple risk management strategies (table 7). Thus, the multidimensional
relationships between perceived risk sources and management strategies suggest that there are no
one-to-one risk management strategies. However, a particular risk management strategy can be
applied to different sources of risk.
If the fish farmers themselves are not able to respond efficiently to the risks, government agen-
cies could provide support to develop the farmers’ abilities to adopt the appropriate risk manage-
ment tools and strategies. In this, financial institutions play an important role because they can
be a tool to reduce financial risks. Furthermore, the sharing of market risk can be facilitated if
farmers can be linked horizontally and vertically in the value chain.
C O N C L U SI O N S
The objective of this study is to provide empirical insights into pond aquaculture farmers’ percep-
tions of risk sources and their management strategies. Descriptive statistics and a principal com-
ponent analysis technique are used to assess farmers’ perceptions towards sources of risk and risk
management strategies. Furthermore, the relationships of risk perception and risk management
strategies with the farmers’ socioeconomic characteristics are investigated by using a principal
component regression model.
The results suggest that, in general, price variability and quality of inputs are perceived as the
most important sources of risk. The use of quality inputs is perceived as the most important risk
management strategy, whereas strategies focusing on price risk reduction, like sales contracts, are
not considered to the same extent as input quality.
Overall, household income, training, and land ownership have the most significant influences
on farmers’ risk perceptions. Similarly, owners of larger and medium-sized farms consider finan-
cial risk to be more important than do owners of smaller farms. Financial risks are also considered
to be an important risk source by single owners and owners who personally own land. Farmers
having higher household incomes perceive all sources of risk to be less influential. Accordingly,
a variety of risk management strategies are needed to address a specific source of risk. Further-
more, cooperative farming can also help farmers minimize many risks associated with financ-
ing, quality control, extension and collaboration, and disease management. Altogether, the most
important risk management strategies are better access to financial resources and improved farm
management.
Furthermore, the perceived risks and their management strategies statistically significantly
vary across the studied regions, indicating that a common framework for risk management un-
dertaken either at a farm level or at a policy level may not be effective.
This study identifies means of interventions targeting either risk perception or risk response
(strategy). The paper will be a contribution to help policy makers, development NGOs, consult-
ants, and farm operators to identify, guide, and formulate appropriate risk management strate-
gies. A further contribution is to direct future research towards more integrated approaches, taking
into account the perceptions of risk, risk attitudes, and risk responses when looking at managing
aquaculture in developing countries. Another domain of future research could be in relation to
ecological risks since the sector is growing fast, possibly at the cost of biosecurity and environmen-
tal degradation.
Further research could also look at objective risks and how those risks could be shared and re-
duced if small-scale farmers are organized in cooperatives (horizontal integration) or linked with
buyers, retailers, and institutional lenders (vertical integration), which could potentially increase
earnings and improve livelihoods and food security in the growing aquaculture sectors in devel-
oping countries.
REFERENCES
Abdi, H. 2003. “Partial Least Square Regression (PLS Regression).” Encyclopedia for Research Methods for
the Social Sciences 6 (4): 792–95.
Ahmed, N. 2007. “Economics of Aquaculture Feeding Practices: Bangladesh.” In Economics of Aquaculture

Feeding Practices in Selected Asian Countries, edited by M. R. Hasan, 33–64. FAO Fisheries Technical
Paper 505. Rome: FAO.
Ahmed, N., M. F. Alam, and M. R. Hasan. 2010. “The Economics of Sutchi Catfish (Pangasianodon
hypophthalmus) Aquaculture under Three Different Farming Systems in Rural Bangladesh.” Aquaculture
Research 41 (11): 1668–82. https://doi.org/10.1111/j.1365-2109.2010.02549.x.
Ahsan, D. A. 2011. “Farmers’ Motivations, Risk Perceptions and Risk Management Strategies in a Develop-
ing Economy: Bangladesh Experience.” Journal of Risk Research 14 (3): 325–49. https://doi.org/10.1080
/13669877.2010.541558.
Alam, M. A., A. G. Guttormsen, and K. H. Roll. 2019. “Production Risk and Technical Efficiency of Tilapia
Aquaculture in Bangladesh.” Marine Resource Economics 34 (2): 123–41.
Alam, M. F., M. A. Khan, and A. A. S. M. Huq. 2012. “Technical Efficiency in Tilapia Farming of Bangla-
desh: A Stochastic Frontier Production Approach.” Aquaculture International 20:619–34.
Ali, H., M. M. Rahman, K. Murshed-e-Jahan, and G. C. Dhar. 2018. “Production Economics of Striped Cat-
fish (Pangasianodon hypophthalmus, Sauvage, 1878) Farming under Polyculture System in Bangladesh.”
Aquaculture 491:381–90.
Ali, H., V. Upraity, S. Gurung, G. C. Dhar, and B. Belton. 2018. “Making Sense of the Market: Assessing the
Participatory Market Chain Approach to Aquaculture Value Chain Development in Nepal and Bangla-
desh.” Aquaculture 493:395–405.
Anwar, J. 2011. “Market Study on Some Freshwater Farmed Fish: Tilapia and Pangas (Mekong River Cat-
fish).” Final report, USAID and PRICE.
Asche, F. 2015. Aquaculture: Opportunities and Challenges. Geneva: International Centre for Trade and
Sustainable Development.
Asche, F., M. Sikveland, and D. Zhang. 2018. “Profitability in Norwegian Salmon Farming: The Impact of
Firm Size and Price Variability.” Aquaculture Economics and Management. https://doi.org/10.1080
/13657305.2018.1385659.
Asravor, R. 2018. “Smallholder Farmers’ Risk Perceptions and Risk Management Responses.” African Journal
of Economic and Management Studies 9 (3): 367–87.
Belton, B., S. R. Bush, and D. C. Little. 2018. “Not Just for the Wealthy: Rethinking Farmed Fish Consumption
in the Global South.” Global Food Security 16:85–92.
Bergfjord, O. J. 2007. “Is There a Future for Salmon Futures? An Analysis of the Prospects of a Potential
Futures Market for Salmon.” Aquaculture Economics and Management 11 (2): 113–32.
———. 2009. “Risk Perception and Risk Management in Norwegian Aquaculture.” Journal of Risk Re-
search 12 (1): 91–104. https://doi.org/10.1080/13669870802488941.
Bishu, K. G., S. O’Reilly, E. Lahiff, and B. Steiner. 2018. “Cattle Farmers’ Perceptions of Risk and Risk Man-
agement Strategies: Evidence from Northern Ethiopia.” Journal of Risk Research 21 (5): 579–98.
Bocqueho, G., F. Jacquet, and A. Reynaud. 2014. “Expected Utility or Prospect Theory Maximisers? Assess-
ing Farmers’ Risk Behaviour from Field-Experiment Data.” European Review of Agriculture Economics.
https://doi.org/10.1093/erae/jbt006.
Cerny, B. A., and H. F. Kaiser. 1977. “A Study of a Measure of Sampling Adequacy for Factor-Analytic Cor-
relation Matrices.” Multivariate Behavioral Research 12 (1): 43–47.
Chitmanat, C., P. Lebel, N. Whangchai, J. Promya, and L. Lebel. 2016. “Tilapia Diseases and Management
in River-Based Cage Aquaculture in Northern Thailand.” Journal of Applied Aquaculture 28 (1): 9–16.
Cullen, A. C., C. L. Anderson, P. Biscaye, and T. W. Reynolds. 2018. “Variability in Cross-Domain Risk
Perception among Smallholder Farmers in Mali by Gender and Other Demographic and Attitudinal
Characteristics.” Risk Analysis 38 (7): 1361–77.
D’Ambra, A., and P. Sarnacchiaro. 2010. “Some Data Reduction Methods to Analyze the Dependence with
Highly Collinear Variables: A Simulation Study.” Asian Journal of Mathematics and Statistics Science
3 (2): 69–81.
Duinen, R. V., T. Filatova, P. Geurts, and A. van der Veen. 2015. “Empirical Analysis of Farmers’ Drought
Risk Perception: Objective Factors, Personal Circumstances, and Social Influence.” Risk Analysis 35 (4):
741–55.
Everitt, B., and T. Hothorn. 2011. “Principal Components Analysis.” In An Introduction to Applied Multi-
variate Analysis with R, 61–103. New York: Springer.
FAO (Food and Agriculture Organization). 2018. The State of World Fisheries and Aquaculture 2018: Meeting
the Sustainable Development Goals. Rome: FAO.
Flaten, O., G. Lien, M. Koesling, P. S. Valle, and M. Ebbesvik. 2005. “Comparing Risk Perceptions and Risk
Management in Organic and Conventional Dairy Farming: Empirical Results from Norway.” Livestock
Production Science 95:11–25.
Fox, J., and W. Weisberg. 2013. An R Companion to Applied Regression. 2nd ed. Thousand Oaks, CA: Sage.
Guttormsen, A. G. 1999. “Forecasting Weekly Salmon Prices: Risk Management in Fish Farming.” Aquacul-
ture Economics and Management 3:159–66.
Hair, J., J. F. Anderson, R. E. Tatham, and W. C. Black. 2006. Multivariate Data Analysis. Upper Saddle River,
NJ: Prentice-Hall.
Hayran, S. 2019. “Perceptions of Wheat Producers towards Risk and Risk Management Strategies: A Case
Study from Turkey.” Ciência Rural 49 (12). https://doi.org/10.1590/0103-8478cr20190396.
Hernandez, R., B. Belton, T. Reardon, C. Hu, X. Zhang, and A. Ahmed. 2018. “The ‘Quiet Revolution’ in the
Aquaculture Value Chain in Bangladesh.” Aquaculture 493:456–68.
Islam, I., M. Nielsen, B. S. Ehlers, B. Zaman, and I. Theilade. 2020. “Are Trade Credits a Gain or a Drain?
Power in the Sale of Feed to Pangasius and Tilapia Farmers in Bangladesh.” Aquaculture Economics and
Management 24 (3): 338–54.
Jolliffe, I. T., and J. Cadima. 2016. “Principal Component Analysis: A Review and Recent Developments.”
Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
374 (2065): 20150202.
Kabir, M. J., M. Alauddin, and S. Crimp. 2017. “Farm-Level Adaptation to Climate Change in Western Ban-
gladesh: An Analysis of Adaptation Dynamics, Profitability and Risks.” Land Use Policy 64:212–24.
Kabir, M. J., R. Cramb, M. Alauddin, C. Roth, and S. Crimp. 2017. “Farmers’ Perceptions of and Responses
to Environmental Change in Southwest Coastal Bangladesh.” Asia Pacific Viewpoint 58 (3): 362–78.
Khan, A., A. Guttormsen, and K. H. Roll. 2018. “Production Risk of Pangas (Pangas hypophthalmus) Fish
Farming.” Aquaculture Economics and Management 22 (2): 192–208.
Kobayashi, M., S. Msangi, M. Batka, S. Vannuccini, M. M. Dey, and J. L. Anderson. 2015. “Fish to 2030: The
Role and Opportunity for Aquaculture.” Aquaculture Economics and Management 19 (3): 282–300.
Krämer, N., and M. Sugiyama. 2011. “The Degrees of Freedom of Partial Least Squares Regression.” Journal
of the American Statistical Association 106 (494): 697–705.
Le, T. C., and F. Cheong. 2010. “Perceptions of Risk and Risk Management in Vietnamese Catfish Farming:
An Empirical Study.” Aquaculture Economics and Management 14 (4): 282–314. https://doi.org/10.1080
/13657305.2010.526019.
Lebel, L., P. Lebel, and C. J. Chuah. 2018. “Governance of Aquaculture Water Use.” International Journal of
Water Resources Development. https://doi.org/10.1080/07900627.2018.1457513.
Lebel, L., P. Lebel, and B. Lebel. 2016. “Impacts, Perceptions and Management of Climate-Related Risks
to Cage Aquaculture in the Reservoirs of Northern Thailand.” Environmental Management 58 (6): 931–45.
Levy, P. S., and S. Lemeshow. 1991. Sampling of Populations: Methods and Application. Hoboken, NJ: Wiley.
Lien, G., O. Flaten, A. M. Jervell, M. Ebbesvik, and M. Koesling, and P. S. Valle. 2006. “Management and
Risk Characteristics of Part-Time and Full-Time Farmers in Norway.” Review of Agriculture Economics
28 (1):111–31.
May, D., S. Arancibia, K. Behrendt, and J. Adams. 2019. “Preventing Young Farmers from Leaving the Farm:
Investigating the Effectiveness of the Young Farmer Payment Using a Behavioural Approach.” Land Use
Policy 82:317–27.
Meuwissen, M. P. M., R. B. M. Hurine, and J. B. Hardker. 2001. “Risk and Risk Management: An Empirical
Analysis of Dutch Livestock Farmers.” Livestock Production Science 69:43–53.
Mitra, S., M. A. Khan, and R. Nielsen. 2019. “Credit Constraints and Aquaculture Productivity.” Aquacul-
ture Economics and Management 23 (4): 410–27.
Mitra, S., M. A. Khan, R. Nielsen, N. Islam. 2019. “Total Factor Productivity and Technical Efficiency Dif-
ferences of Aquaculture Farmers in Bangladesh: Do Environmental Characteristics Matter?” Journal of
the World Aquaculture Society. https://doi.org/10.1111/jwas.12666.
Neira, I., C. R. Engle, and C. Ngugi. 2009. “Economic and Risk Analysis of Tilapia Production in Kenya.”
Journal of Applied Aquaculture 21 (2): 73–95.
Patrick, G. F., and W. N. Musser. 1997. “Sources of and Responses to Risk: Factor Analyses of Large-Scale US
Cornbelt Farmers.” In Risk Management Strategies in Agriculture: State of the Art and Future Perspec-
tives—Mansholt Studies, edited by R. Huirne, J. B. Hardaker, and A. A. Dijkhuizen, 7:45–53. Wageningen,
Netherlands: Wageningen Agricultural University.
Pennings, J., and P. Garcia. 2001. “Measuring Producers’ Risk Preferences: A Global Risk Attitude Con-
struct.” American Journal of Agricultural Economics 83:993–1009.
Quagraine, K. K., T. H. Kuethe, and C. R. Engle. 2007. “Arkansas Catfish Farmers and Marketing Contracts:
Some Insights into Motivations.” Aquaculture Economics and Management 11:39–51.
Rahman M. T., R. Nielsen, and M. A. Khan. 2018. “Agglomeration Externalities and Technical Efficiency:
An Empirical Application to the Pond Aquaculture of Pangas and Tilapia in Bangladesh.” Journal of
Aquaculture Economics and Management 23 (2): 158–87.
———. 2019. “Efficiency and Production Environmental Heterogeneity in Aquaculture: A Meta-frontier
DEA Approach.” Aquaculture 509 (2019): 140–48.
Rahman, M. T., R. Nielsen, M. A. Khan, and I. Ankamah-Yeboah. 2020. “Impact of Management Practices
and Managerial Ability on the Financial Performance of Aquaculture Farms in Bangladesh.” Aquacul-
ture Economics and Management 24 (1): 79–101.
Rowe, W. D. 1977. An Anatomy of Risk. New York: John Wiley and Sons.
Sarker, A. A., M. F. Alam, Z. A. Mohamed, and M. A. Khan. 2016. “Stochastic Modeling of Production Risk
and Technical Efficiency of Thai koi (Anabas testudineus) Farming in Northern Bangladesh.” Journal
of Aquaculture Economics and Management 20 (2): 165–84.
Sawatsky, M. L., M. Clyde, and F. Meek. 2015. “Partial Least Squares Regression in the Social Sciences.”
Quantitative Methods for Psychology 11 (2): 52–62.
Shaw, P. J. A. 2003. Multivariate Statistics for the Environmental Sciences. New York: Hodder Arnold
[Google Scholar].
Sitkin, S. B., and A. L. Pablo. 1992. “Reconceptualizing the Determinants of Risk Behavior.” Academy of Man-
agement Review 17 (1): 9–38.
Sjoberg, L. 1998. “World Views, Political Attitudes and Risk Perception.” Risk: Health, Safety and Environ-
ment 9:137–52.
Sjoberg, L., B. Moen, and T. Rundmo. 2004. Explaining Risk Perception: An Evaluation of the Psychometric
Paradigm in Risk Perception Research. Trondheim, Norway: Rotunde.
Slovic, P. 1987. “Perception of Risk.” Science 236 (80): 280–85. https://doi.org/10.1126/science.3563507.
Tisdell, C., N. Hishamunda, R. van Anrooy, T. Pongthanapanich, and M. A. Upare. 2012. “Investment, In-
surance and Risk Management for Aquaculture Development.” In Farming the Waters for People and
Food, edited by R. P. Subasinghe, J. R. Arthur, D. M. Bartley, S. S. De Silva, M. Halwart, N. Hishamunda,
C. V. Mohan, and P. Sorgeloos, 303–33. Proceedings of the Global Conference on Aquaculture 2010,
Phuket, Thailand, September 22–25, 2010. Rome: FAO and Bangkok: NACA.
Tveteras, R. 1999. “Production Risk and Productivity Growth: Some Findings for Norwegian Salmon Aqua-
culture.” Journal of Productivity Analysis 12:161–79.
Valderrama, D., and C. R. Engle. 2001. “Risk Analysis of Shrimp Farming in Honduras.” Aquaculture Eco-
nomics and Management 5 (1–2): 49–68. https://doi.org/10.1080/13657300109380277.
Van Winsen, F., Y. de Mey, L. Lauwers, S. Van Passel, M. Vancauteren, and E. Wauters. 2016. “Determinants
of Risk Behaviour: Effects of Perceived Risks and Risk Attitude on Farmer’s Adoption of Risk Management
Strategies.” Journal of Risk Research 19 (1): 56–78. https://doi.org/10.1080/13669877.2014.940597.
Watson, J. R., F. Armerin, D. H. Klinger, and B. Belton. 2018. “Resilience through Risk Management: Coop-
erative Insurance in Small-Holder Aquaculture Systems.” Heliyon 4 (9): e00799.
Weber, E. U., A.‐R. Blais, and N. E. Betz. 2002. “A Domain‐Specific Risk‐Attitude Scale: Measuring Risk
Perceptions and Risk Behaviors.” Journal of Behavioral Decision Making 15 (4): 263–90.
Weber, E. U. E., and R. A. R. Milliman. 1997. “Perceived Risk Attitudes: Relating Risk Perception to Risky
Choice.” Management Science 43:123–44.
Willock, J., I. Deary, and M. McGregor. 1999. “Farmers’ Attitudes, Objectives, Behaviors, and Personality
Traits: The Edinburgh Study of Decision Making on Farms.” Journal of Vocational Behavior 54:5–36.
Wolff, K., S. Larsen, and T. Øgaard. 2019. “How to Define and Measure Risk Perceptions.” Annals of Tourism
Research 79:102759.

Rahman 2020

Uploaded by

Copyright:

Available Formats

Rahman 2020

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Rahman 2020

Uploaded by

Copyright:

Available Formats

Perceived Risk and Risk Management

Strategies in Pond Aquaculture

Key words: Aquaculture, management strategies, risk perception.

MOTI VATI ON OF T HE ST UDY

B AN GL ADE SH POND A QUAC U L TU R E IN DU STR Y

Figure 4. Study Areas. A color version of this ﬁgure is available online.

Table 1. Characteristics of Farms and Farmers in Bangladesh

Variable Name Farms in Survey

Personally owned farms (%) OwnerF 90

RF i p b0 1 b1 OwnerF 1 b2 OwnerL 1 b3 LandCon 1 b4 Age 1 b5 OccM

1 b6 Edu 1 b7 Exp 1 b8 Train 1 b9 DummyL 1 b10 DummyM (1)

1 b11 FamilyS 1 b12 LnIncome 1 bDi 1 ε,

RMSj p b0 1 b1 OwnerF 1 b2 OwnerL 1 b3 LandCon 1 b4 Age 1 b5 OccM

1 b6 Edu 1 b7 Exp 1 b8 Train 1 b9 DummyL 1 b10 DummyM (2)

1 b11 FamilyS 1 b12 LnIncome 1 bRF i 1 bDi 1 ε,

where the following applies:

RFi: Standardized principal components scores deﬁned as categories of risk sources (i p

Di3: Regional dummies (i p Mymensingh [reference region], Bogura, Comilla, Khulna,

F ARM ERS ’ P ER C EPT I ON S O F R IS K S OUR C E S

Table 2. Means, Standard Deviations, and Ranks of Risk Consequences

Risk Sources Rank Mean Std. Dev.

Fish price variability 1 3.763 1.065

F ARM ERS ’ PERCEPTI ONS OF RISK MA NAG EMENT ST RAT EGI ES

Table 3. Varimax Rotated Factor Loadings for Risk Sources

Principal Components / Factorsa

Production Financial Market- Institutional

Percentage of total variance explained 44.021 6.152 5.528 3.822

KMO test 0.961

Risk Management Strategies Rank Mean Std. Dev.

Choose a good brand of feed 1 3.957 0.935

RELATI ONSHIP S A MONG PERCEPTI ONS OF RISKS A ND FARM

Principal Components / Factorsa (Types of Risk Management Strategies)

Financial Farm Quality Marketing Farmers’ Extension and Disease Input

Principal Components / Factorsa (Types of Risk Management Strategies)

Financial Farm Quality Marketing Farmers’ Extension and Disease Input

Involve production contract with

KMO test 0.801

Table 6. Results of Multiple Regression for Categories of Risk Sources

Intercept 1.790*** 1.283*** 0.820*** 0.824*

Bogura 0.121 –0.004 –0.400 –0.103

Adjusted R2 0.043 0.040 0.046 0.055

Financial Farm Quality Marketing Farmers’ Extension and Disease

Ahmed, N. 2007. “Economics of Aquaculture Feeding Practices: Bangladesh.” In Economics of Aquaculture

You might also like

Intercept 1.790* 1.283* 0.820*** 0.824*