Skip to main content

Farmers versus nature: managing disaster risks at farm level

Abstract

Farmers have number of options available in managing disaster risks at farm and many of them utilize these risk management tools simultaneously. However, previous studies have ignored the correlation among the risk management adoption decisions. This study, therefore, aimed at investigating factors affecting the adoption of off-farm diversification and agricultural credit to manage catastrophic risks while taking into account the contemporaneous correlation in the decisions to adopt the two risk management tools simultaneously. Bivariate and multinomial probit approaches are applied to data collected from four districts of Khyber Pakhtunkhwa Province in Pakistan. Findings revealed a strong correlation between the decisions to adopt the two risk management tools and conclude that adopting one risk management tool may induce farmers to adopt the other tool at the same time. Moreover, socioeconomic characteristics, losses due to natural disasters, farmers’ risk perceptions and their risk attitude play significant role in shaping their decisions of adopting risk management tools.

Introduction

Exposure to and losses resulting from natural disasters in general and hydrological disasters in particular are increasing worldwide, affecting peoples’ livelihoods and food security (UN 2014). Their frequency and subsequent economic losses have steadily increased over the past few decades, stretching the response capacities of governments and humanitarian organizations (Hellmuth et al. 2011). Agriculture is one of the sectors most affected by natural hazards and disasters, which enhance vulnerabilities of resource-poor farmers in particular and often threaten their livelihood security (UN 2014). Natural hazards determine production in ways that are outside the control of the farmer (Antón 2009). Managing uncertainties and risks associated with extreme weather conditions in agriculture is crucial as it affects other sectors of the economy (Kammar and Bhagat 2009; Ullah et al. 2015a) and is generally considered as a key matter in farmers’ decision-making and to the policies that affect these decisions (Robinson and Barry 1987).

The agricultural sector of Pakistan has been devastated by two successive years of massive floods. The monsoon floods in 2010 caused unprecedented damages to agriculture crops, livestock, fisheries and forestry and primary infrastructure such as tube wells, water channels household storages, houses, animal sheds, personal seed stocks, fertilizers and agricultural machinery. One-fifth (17 million acres) of the country’s total land area remained inundated. 1.9 million houses were destroyed, and over 2000 people have died. The economic impact of the tragedy has been estimated at over US$43 billion (RISE 2010). The floods struck just before the harvesting of main crops, such as cotton, rice, maize, vegetables and sugarcane. Overall production loss of sugar cane, paddy and cotton was estimated at 13.3 million mt, over 2 million ha of standing crops was lost or damaged, and over 1.2 million head of livestock (excluding poultry) died due to the flood (WFP 2010). In 2011, another massive flood struck Pakistan leaving a significant impact on people’s lives, especially related to the loss of livelihoods, primarily those related to agricultural activities. The flood destroyed standing crops of cotton, rice, sugar cane, sorghum, vegetables and pulses on about 0.84 million hectares of land. Livestock also suffered heavy losses, approximately 115,500 livestock have perished and about 5 million surviving livestock were directly affected (GoP 2012). After the devastating floods in 2010 and 2011, farmers are adopting sophisticated risk management strategies to reduce the impacts of such kind of risks in future. The socioeconomic factors, farmers’ risk perception, their attitude toward risk and previous incidence of losses due to floods play an important role in these decisions, and thus, these factors need to be studied and should be key elements while drafting risk management strategies for agriculture sector.

As agriculture is the only source of livelihood for most of the people in rural areas, it is important for the farmers to protect and safeguard their farms from risks and uncertainties to continue earning their livelihood from agriculture. Among the most essential and complex decisions that an agricultural producer has to make is the choice of a combination of risk management instruments to provide the best income safety net for their particular situation (Coble et al. 2000; Ke and Wang 2002). Farmers have number of options available to manage farm risk and many of them use these risk management tools simultaneously (Velandia et al. 2009; Ullah et al. 2015a). However, most of the previous studies ignored the potential for simultaneous adoption of the risk management tools and considered the risk management decisions independent of each other. Examples of studies that analyze adoption of a single risk management tool are as follows: Shapiro and Brorsen (1988) and Makus et al. (1990) for hedging with futures and options; Goodwin and Schroeder (1994) and Davis et al. (2005) for forward contracting/pricing; and Sherrick et al. (2004) and Makki and Somwaru (2001) for crop insurance. Velandia et al. (2009), Ullah et al. (2015a) and Ullah and Shivakoti (2014) are few exceptions where the authors have considered the possibility of simultaneous adoption of multiple risk management tools. However, the potential of off-farm diversification and agricultural credit to cope with the growing instabilities in farm incomes and factors influencing the adoption of these risk coping tools are rarely investigated particularly in Pakistan. This study is therefore aimed at investigating factors responsible for the adoption of off-farm diversification (ex ante risk coping tool) and agricultural credit (ex post risk absorbing tool) keeping in view the correlation between these decisions. The specific objectives of the study are (1) to figure out correlation between farmers’ decisions of adopting the two risk management tools and (2) to assess the impact of farm and farm characteristics, losses due to natural disasters, farmers’ perceptions of the catastrophic risk sources and their attitude toward catastrophic risks on their decisions of adopting the two risk management tools. The findings of the study will be of great importance for government line agencies, extension educators and other researchers in a number of ways. Policy-makers can use the findings to identify which type of farmers will use government supported risk coping tools (i.e., Crop Loan Insurance Scheme) in the presence of traditional risk management strategies. The present study will also help to understand how farmers’ perception and attitude toward risk can affect their risk management decisions.

Methodology

Study area and sampling

The required sample households for data collection were selected using multistage sampling technique. In the first stage, Khyber Pakhtunkhwa (KP) Province was purposively selected for the study. In the second stage, four districts in KP Province viz Peshawar, Charsadda, Swat and Shangla were selected purposively. The main reason behind the selection of these four districts is that two districts, Peshawar and Charsadda, are located in Peshawar valley and the farmers have relatively higher access to input–output market, extension and other publicly provided services, credit, information, etc, while Swat and Shangla are relatively less developed districts and the farmers have limited access to markets, information, agricultural extension services and other government provided services and are less adoptive of modern technologies (Ahmad et al. 2007; Shahbaz et al. 2010). In the third stage, 8 villages in the selected districts are randomly selected for data collection. In the fourth stage, a total of 330 respondents are randomly selected using the below equation as suggested by Yamane (1967). The required data were collected using a structured questionnaire during February to April 2013.

$$n \, = N/(1 + Ne^{2} )$$
(1)

where n = sample size in each village; N = total number of farming households in a village; e = precision which is set at 15 % (0.15).

Estimation procedure

A bivariate probit model is used to analyze the effect of farm and farm household characteristic losses due to natural disasters, farmers’ perceptions of the catastrophic risk sources and their risk attitude on their decisions to adopt diversification and agricultural credit to manage disaster risk at farm level. The bivariate probit model, considering the possibility of correlation between the risk management adoption decisions, is given as follows:

$$Y_{ij} = x_{ij }^{\prime} \beta_{j} + \varepsilon_{ij}$$
(2)

where Y ij (j = 1, …, m) represent the risk management options (in this case m = 2) faced by ith farmer (i = 1, …, n), \(x_{ij }^{\prime}\) is a 1 × k vector of observed variables that affect the risk management adoption decision, \(\beta_{j}\) is a k × 1 vector of unknown parameters (to be estimated), and \(\varepsilon_{ij}\) is the unobserved error term. In this specification, each Y j is a binary variable and, thus, Eq. (2) is actually a system of m equations (m = 2 in this case) to be estimated (Ullah et al. 2015a):

$$\begin{aligned} Y_{1}^{*} & = \alpha_{1} + X\beta_{1} + \varepsilon_{1} \\ Y_{2}^{*} & = \alpha_{2} + X\beta_{2} + \varepsilon_{2} \\ \end{aligned}$$
(3)

where \(Y_{1}^{*}\) and \(Y_{2}^{*}\) are two latent variables underlying each of the risk strategy adoption decision such that y j  = 1 if \(y_{j}^{*}\) > 0; 0 otherwise.

A multinomial probit approach is also used to analyze the impact of independent variables on the adoption of a combination of risk management tools. In the multinomial approach, the choice set is made of all possible combinations of the risk management tools instead of just the risk management tools by themselves (Velandia et al. 2009; Ullah and Shivakoti 2014). The two risk management alternatives make four possible combinations a farmer can choose from the following: (1) no risk management strategy is used, (2) only diversification is used, (3) only agricultural credit is used, and (4) both diversification and agricultural credit are used. Given the choice set, the multinomial probit model can be specified as follows:

$$Y_{i} = \, x_{i}^{\prime } \beta \, + \, \varepsilon_{i}$$
(4)

Y i in this case represents the risk management tool combination (Y i  = 1, …, m) that the ith producer (i = 1, …, n) chooses, \(X_{i }^{\prime}\) is a 1 × K vector of observed variables that affect the risk management combination chosen, β is a k × 1 vector of unknown parameters (to be estimated), and ε i is the unobserved error term. A maximum-likelihood (ML) procedure is used to estimate the unknown parameters in Eq. (4) using computer software Stata (version 12).

Variables used in the analysis

Dependent variables

Off-farm diversification Diversification is one of the basic and obvious approaches used since mankind began to engage in agriculture (Tangermann 2011). Households smooth income by diversifying income sources and thus minimizing the effect of a negative shock to any one of them. For the current study, diversification includes any off-farm strategy adopted by agricultural producers to cope with the variability in farm production and/or income. The variable associated with off-farm diversification is a binary variable (1 if farming household adopts off-farm diversification for risk management and 0 otherwise).

Agricultural credit Credit plays an important role in the process of modernization of agriculture and commercialization of the rural economy (Abedullah et al. 2009), and the credit reserves are one way farmers manage risk (Skees 1999). Holding a credit reserve can be an efficient way to provide liquidity to guide a business through hard times (Anderson 2001). Credit is used as a dependent variable in the model. As discussed above, there are various uses of agricultural credit in farm operations/activities; however, in this study, its value depends on the farmers’ decision to utilize credit only for the purpose of farm risk management. The variable associated with farm credit is a binary variable (1 if farming household uses agricultural credit as an ex post risk absorbing tool and 0 otherwise).

Independent variables

  1. 1.

    Farm and farm household characteristics.

Farm and farm household characteristics including age, education and income significantly affect the risk attitudes of the agricultural producers (Ullah et al. 2015b) and are potential contributors to farmers’ risk management decisions (Sherrick et al. 2004). Some studies also found that risk preferences differ significantly based on education (e.g., Harrison et al. 2007), age (e.g., Tanaka et al. 2010) and/or income (Cohen and Einav 2007). Farming experience also plays a significant role in their decisions to adopt sophisticated risk management tools. Larger proportion of own land is expected to discourage the adoption of the risk management tools as it reflects greater wealth, greater stability of land control and a lesser need for risk management instruments (Velandia et al. 2009).

  1. 2.

    Losses due to natural disasters.

These are the actual losses faced by the farming household due to natural disasters during the last 5 years. It includes:

  1. (a)

    Land affected measured as a ratio of total land operated to land affected by the disasters (mainly floods and heavy rains).

  2. (b)

    Animals affected measured as number of animals affected by the natural disasters excluding poultry.

  3. (c)

    Farming tools lost/damaged measured as number of farming tools lost/damaged due to natural disasters.

  1. 3.

    Farmers’ risk perceptions.

Three catastrophic risk sources are considered for the study including (1) risk of floods, (2) risk of heavy rains and (3) risk of drought. Farmers are asked to score the incidence and severity of each source of risk, on a Likert scale, from 1 (very low) to 5 (very high) to express how significant they consider each source of risk in terms of its potential impact on their farm enterprise (Ullah et al. 2015b). The scores are then combined in a risk matrix (Cooper et al. 2005) and categorized into two groups (as depicted in Fig. 1) as low if the score lies between 2 and 5 and high if the score is from 6 to 10 (Ullah and Shivakoti 2014). The results of risk matrix are then used for further analysis as 1 if the risk matrix score is high and 0 otherwise.

  1. 4.

    Farmers’ risk attitude.

Equally Likely Certainty Equivalent (ELCE) model is used to elicit farmer’s attitude toward risk. Certainty equivalents (CE) are derived for a sequence of risky outcomes and match them with utility values (Binici et al. 2003). For instance, the respondent was asked to specify the monetary value of a sure outcome that makes him/her indifferent between the two risky outcomes of total annual household income, say PKRFootnote 1 50,000 and PKR 0 with equal probabilities. Suppose the response is PKR 26,000, the respondent is asked again to specify the monetary value of a sure outcome that makes him/her indifferent between the two risky outcomes of PKR 26,000 and PKR 0 with equal probability. In this way, several certainty equivalents are derived and matched with utility values.

Fig. 1
figure1

Risk matrix

For the other half of the income distribution, the farmer is asked to specify the monetary value of a sure outcome that makes him/her indifferent between PKR 26,000 and PKR 50,000 each with equal probabilities. In this way, several CE points are obtained and matched with their respective utility values. Utility value attached with the lower outcome (PKR 0) is 0 and with the higher outcome (PKR 50,000) is 1. The farmer response of PKR 26,000 is his CE for uncertain payouts of PKR 50,000 and PKR 0 with equal probabilities (0.5 each) and utility value for this CE is calculated as follows:

$$U\left( {26,000} \right) = 0.5u\left( 0 \right) + 0.5u\left( {50{,}000} \right) = 0.5\left( 0 \right) + 0.5\left( 1 \right) = .50$$
(5)

Similarly utility values for all the CE points are calculated and presented in Table 1.

Table 1 Example of elicitation of certainty equivalents and computation of utility values

The most commonly used utility functions when assessing risk preferences are the negative exponential, power, expo-power and cubic functions (Binici et al. 2003). Cubic utility function is used to elicit risk attitude of the farmers as it is consistent with risk aversion, risk preferring and risk indifferent attitudes. The cubic utility function is given as follows:

$$u\left( w \right) = \alpha_{1} + \alpha_{2} w + \alpha_{3} w^{2} + \alpha_{4} w^{3}$$
(6)

where u (w) is the utility of wealth. Following Olarinde et al. (2007), income is substituted for wealth. The second derivative is given by 2α 3 + 6α 4 w, the sign of which depends on the sign and magnitude of the parameters α 3 and α 4, and the level of wealth W. Thus, increasing and decreasing marginal utilities are both possible (Torkamani and Rahimi 2001). If the second derivative of the utility function is positive (U″ > 0), it implies a risk seeking attitude; a negative second derivative of utility function (U″ < 0) implies risk aversion, and if the second derivative is zero (U″ = 0), it reflects the risk neutral attitude of the individual (Olarinde et al. 2007). Utility is generally measured on an ordinal scale; however, the shape of the utility function on an ordinal scale can be transformed into a quantitative measure of risk aversion called absolute risk aversion (Pratt 1964; Arrow 1964). The absolute risk aversion is mathematically defined as follows:

$$r_{a} \left( W \right) = - \frac{{U^{\prime } \left( W \right)}}{{U^{\prime \prime } \left( W \right)}}$$
(7)

where r a (W) is coefficient of absolute risk aversion, U′ and U″ are first- and second-order derivatives of wealth (W), respectively. The coefficient of absolute risk aversion is positive if individual is risk averse, negative if individual prefers risk and zero if individual is indifferent to risk (Binici et al. 2003). Following Kouame (2010), the variable of risk aversion enters into the model as a categorical variable having value of 1 if the farmer is risk averse and 0 otherwise.

Limitations and future studies

This study is limited only to the decisions making process of the farmers and the factors affecting these decisions. The potential outcomes of these decisions on their livelihoods, income, food security and poverty are beyond the scope of the research. Future studies should investigate the important role of these decisions in farming household’s overall welfare.

Results and discussion

Descriptive statistics of the variables used in the analysis are presented in Table 2. Off-farm diversification as a risk management tool is adopted by 52 % of the farmers, while agricultural credit is adopted by 46 % of the sampled respondents in the study area. Farmers faced significant losses due to natural disasters in the form of losses in agricultural land under crops, livestock and farming tools and equipment. Farmers consider risk of flood, risk of heavy rains and risk of pest and diseases as major catastrophic risk sources that may lead to potential losses in yields of crops and subsequently alter their incomes from farm sector.

Table 2 Descriptive statistics of variables used in the analysis

As indicated in the table, only 26 % of the sampled respondents considered risk of drought as a high risk source that may alter their farm earnings. A possible explanation for this is the fact that all districts in the province, except Karak, have easy access to the water of a river passing nearby or flowing through for irrigation (Khan 2012). Majority of the farmers (80 %) in the area are risk averse in nature and will avoid risk when facing a risky situation.

Parameter estimates of the bivariate probit model

The parameter estimates of the bivariate probit model are presented in Table 3. The positive and significant value of likelihood ratio (LR) test of ρ kj justifies the estimation of the bivariate probit and the hypothesis H0 of conjoint nullity of ρ kj can be rejected. The Wald chi-square test (128.00) also allows us to reject the H0 hypothesis of conjoint nullity of variable coefficients included in the estimation.

Table 3 Parameter estimates from bivariate probit model and its marginal effects

Factors significantly affecting the adoption of off-farm diversification as risk management tool are age, education, farming experience, monthly off-farm income, proportion of own land, farmers’ risk perception of drought and risk averse nature of farmers. Our result for age is in accordance with Rehima et al. (2013) and Deressa et al. (2010) who also found a positive impact of age on the adoption of diversification. More educated farmers are likely to adopt diversification as they have more ability to assess the merit of diversification as a strategy to cope with the negative shocks resulting from unfavorable weather conditions. Our result for education is in line with Tavernier and Onyango (2008), Kouame (2010) and Ashfaq et al. (2008) who observed a positive relationship of education and the adoption of enterprise diversification as risk management strategy. However, Mesfin et al. (2011) and Rehima et al. (2013) found the higher education levels discourage farmers to adopt diversification to cope with farm income variability.

As indicated by the results, more experienced farmers tend to avoid the use of diversification. One possible explanation for this may be the fact that more experienced farmers generally stick to their profession and are reluctant to invest in enterprise other than agriculture. Mesfin et al. (2011) also observed a negative impact of farming experience on farmers’ decisions of adopting diversification to manage farm risk. However, Ashfaq et al. (2008) reported a positive impact of farming experience on farmers’ decisions of adopting diversification. Higher off-farm incomes attract and motivate farmers to diversify their income sources and smooth their consumption. This translates into a positive and significant relationship of off-farm income and the adoption of off-farm diversification for farm risk management. Our result is in line with Rehima et al. (2013) and Deressa et al. (2010) who also reported a positive relation of non-farm income with the use of diversification. However, Ashfaq et al. (2008) found a negative relation of off-farm income with the adoption of diversification to manage farm risks.

Ullah and Shivakoti (2014) found that higher off-farm monthly income significantly encourages off-farm diversification while it strongly discourages on-farm diversification. Results revealed that larger proportion of own land discourages the use of diversification to manage farm risk. Larger proportion of owned land is related to greater wealth, greater stability of land control and a larger asset base (Velandia et al. 2009). Risk perception of farmers regarding drought significantly influences their decisions to adopt off-farm diversification to offset negative shocks to their farm enterprise arising from adverse weather conditions. Droughts may lead to significant yield losses of the major crops resulting in a decline of net returns from crop production. In order to continue earnings their livelihoods, farmers have to diversify their income sources. The risk averse nature of the farmers also forces them to adopt diversification to minimize risk at farm. Kouame (2010) also found a significant positive affect of high risk aversion with the adoption decisions of diversification.

Factors significantly affecting the adoption of agricultural credit to manage farm risk are farming experience, farm size, proportion of own land, farmers’ perceptions of the risk sources and their attitude toward risk. More experienced farmers are more likely to use credit to deal with climatic risks at farm compared to their less experienced counterpart. Larger farms may require more capital to rehabilitate its potential after a negative shock strike the farm; therefore, the use of credit as an ex post risk management tool is higher for larger farms compared to small farms. Higher proportion of own land reflects greater wealth and greater stability of land control (Velandia et al. 2009) and therefore lesser need of the risk management tools. Loss of livestock due to natural disasters also induces farmers to adopt agricultural credit as an ex post risk absorbing tool to smooth their consumptions. Farmers’ perception of risk sources also necessitates their demand for the use of credit to overcome the negative shocks of adverse weather conditions and hence induce the farmers to adopt credit for farm risk management. Farmers’ risk averse nature also induces them to adopt sophisticated risk management tools including agricultural credit to overcome the negative shocks of the adverse climatic conditions.

The bivariate probit approach allows us to calculate the conditional marginal effects (i.e., marginal effects conditional on the adoption of the other risk management tool) while the individual probit approach does not allow such calculation (Velandia et al. 2009). A general observation of Table 3 indicates that significant variables in the bivariate probit model also have significant marginal effects which guarantee the robustness of the results.

Parameter estimates of multinomial probit model

The multinomial probit results provide information/inference that is different from the bivariate probit model because it focuses on factors affecting the combination of risk management tools that a producer chooses (Ullah et al. 2015a). Table 4 provides proportion of respondents using different combinations of the risk management tools considered in this study.

Table 4 Proportion of respondents adopting different combinations of risk management tools

Parameter estimates from the multinomial probit model are presented in Table 5. As discussed above, these results are different from the bivariate probit approach because in multinomial probit model emphasis is on the combinations of the risk management tools. For example, age of the household head and risk perception of drought significantly encourage the use of off-farm diversification in the bivariate probit approach; however, the effects of these factors are insignificant in the multinomial probit approach. By contrast, family size, land affected due to natural disasters, risk perception of floods and pest and diseases have insignificant effect on the adoption of off-farm diversification in the bivariate probit estimation; however, these factors significantly affect the adoption of off-farm diversification in the multinomial approach. Similarly in the adoption equation of agricultural credit, the effect of farming experience is significant in the bivariate probit model, however, it is insignificant in the multinomial approach. Parameters associated with family size and land affect due to natural disasters have insignificant effect in the bivariate probit approach; however, these factors significantly affect the adoption of agricultural credit in the multinomial probit approach.

Table 5 Parameter estimates from multinomial probit model

The results from the analysis of factors affecting the use of combination of 4 merits further the discussion here since this is the combination most frequently adopted by the farmers in our sample (i.e., aside from the “use no risk management tool considered in this study” combination). Age, level of education, farming experience, off-farm monthly income, farm size and proportion of own land are the socioeconomic factors that significantly affect (encourage or discourage) the adoption of off-farm diversification and agricultural credit simultaneously. Previous studies also reported a mix effect of the socioeconomic characteristics on farmers’ decisions of adopting risk coping tools (Ullah and Shivakoti 2014; Ullah et al. 2015a; Velandia et al. 2009; Kouame 2010 and Sherrick et al. 2004). Losses due to natural disasters (land and livestock affected) also induce farmers to adopt the two risk management tools simultaneously to cope with instability of farm income due to natural disasters. Farmers’ risk perceptions and their attitude toward risk are also important factors that shape farmers’ decision of adopting off-farm diversification and agricultural credit at the same time. Higher risk perceptions and risk averse nature of the farmers force them to adopt sophisticated risk coping tools to mitigate the risk of extreme weather conditions at farm.

Conclusion

The use of multiple risk management tools at the same time is a common practice in farming community. The findings of our study revealed that these decisions are correlated, i.e., farmers’ decision of adopting one risk management tool may induce them to adopt the other risk management tool at the same time. Moreover, these decisions are influenced by variety of factors including the socioeconomic characteristics of the farmers, the losses farmers faced due to natural disasters, farmers’ risk perceptions and their attitude toward risk. The analysis of risk management choices using both bivariate and multinomial probit approaches provides richer interpretations, better inferences and more information that may further improve understanding of the risk management decisions of farmers. The information on farmers’ decision-making under risk, gleaned through bivariate and multinomial probit approaches, may serve as basis for government line agencies and policy-makers to draft sound policies for farm risk management particularly in favor of small and/or landless farmers.

Notes

  1. 1.

    PKR is abbreviation for Pakistani Rupee, 1 PKR is approximately equal to 0.01 USD.

References

  1. Abedullah MN, Khalid M, Kouser S (2009) Role of agriculture credit in the growth of livestock sector: a case study of Faisalabad, Pakistan. Vet J 29(2):81–84

    Google Scholar 

  2. Ahmad S, Jamal M, Ikramullah A, Himayatullah (2007) Role of extension services on the farm productivity of district Swat: a case study of two villages. Sarhad J Agric 23(4):1265–1272

    Google Scholar 

  3. Anderson JR (2001) Risk management in rural development: a review. Rural Development Strategy Background Paper 7, Rural Development Department, the World Bank, Washington, DC

  4. Antón J (2009) Risk management in agriculture—a holistic conceptual framework. Organization for Economic Co-operation and Development. Trade and Agriculture Directorate Committee for Agriculture. TAD/CA/APM/WP(2008)22/FINAL

  5. Arrow KJ (1964) The role of securities in the optimal allocation of risk bearing. Rev Econ Stud 31:91–96

    Article  Google Scholar 

  6. Ashfaq M, Hassan S, Naseer MZ, Baig IA, Asma J (2008) Factors affecting farm diversification in rice–wheat. Pak J Agric Sci 45(3):91–94

    Google Scholar 

  7. Binici T, Koc AA, Zulauf CR, Bayaner A (2003) Risk attitude of farmers in terms of risk aversion: a case study of lower Seyhan plain farmers in Adana-province, Turkey. Turk J Agric For 27:305–312

    Google Scholar 

  8. Coble KH, Heifner RG, Zuniga M (2000) Implications of crop yield and revenue insurance for producer hedging. J Agric Resour Econ 25:432–452

    Google Scholar 

  9. Cohen A, Einav L (2007) Estimating risk preferences from deductible choice. Am Econ Rev 97(3):745–788

    Article  Google Scholar 

  10. Cooper DF, Grey S, Raymond G, Walker P (2005) Project risk management guidelines. Wiley, Chichester

    Google Scholar 

  11. Davis TD, Patrick GF, Coble KH, Knight TO, Baquet AE (2005) Forward pricing behavior of corn and soybean production. J Agric Appl Econ 37:145–160

    Google Scholar 

  12. Deressa TT, Ringler C, Hassan RM (2010) Factors affecting the choices of coping strategies for climate extremes. The case of farmers in the Nile Basin of Ethiopia IFPRI Discussion Paper 01032

  13. Goodwin BK, Schroeder TC (1994) Human capital, producer education programs, and the adoption of forward-pricing methods. Am J Agric Econ 76:936–947

    Article  Google Scholar 

  14. Government of Pakistan (GoP) (2012) Economics survey of Pakistan (2010–11), Economics Advisory Wing. Finance Division, Islamabad

    Google Scholar 

  15. Harrison G, Lau M, Rutström E (2007) Estimating risk attitudes in Denmark: a field experiment. Scand J Econ 109(2):341–368

    Article  Google Scholar 

  16. Hellmuth ME, Mason SJ, Vaughan C, van Aalst MK, Choularton R (eds) (2011) A better climate for disaster risk management. International Research Institute for Climate and Society (IRI), Columbia University, New York

    Google Scholar 

  17. Kammar SK, Bhagat R (2009) Constraints experienced by farmers in adopting risk and uncertainty management strategies in rainfed agriculture. Pusa AgriSci 32(2009):70–74

    Google Scholar 

  18. Ke B, Wang HH (2002) An assessment of risk management strategies for grain growers in the Pacific Northwest. Agric Fin Rev 62(2):117–133

    Article  Google Scholar 

  19. Khan MA (2012) Agricultural development in Khyber Pakhtun Khwa: prospects, challenges and policy options. Pak A J Pak Stud 4(1):49–68

    Google Scholar 

  20. Kouame EBH (2010) Risk, risk aversion and choice of risk management Strategies by cocoa farmers in western Cote D’ivoire. University of Cocody-AERC Collaborative PHD Program. http://www.csae.ox.ac.uk/conferences/2010-EDiA/papers/267-Kouame.pdf. Retrieved 28th September, 2012

  21. Makki SS, Somwaru A (2001) asymmetric information in the market for yield and revenue insurance products. Washington, DC: USDA Economic Research Service, Technical Bulletin No. 1892, April 2001

  22. Makus LD, Lin BH, Carlson J, Kreibill-Prather R (1990) Factors influencing farm- level use of futures and options in commodity marketing. Agribus Int J 6:621–631

    Article  Google Scholar 

  23. Mesfin W, Fufa B, Haji J (2011) Pattern, trend and determinants of crop diversification: empirical evidence from smallholders in eastern Ethiopia. J Econ Sust Dev 2(8):78–89

    Google Scholar 

  24. Olarinde LO, Manyong VM, Akintola JO (2007) Attitudes towards risk among maize farmers in the dry savanna zone of Nigeria: some prospective policies for improving food production. Afr J Agric Res 2(8):399–408

    Google Scholar 

  25. Pratt J (1964) Risk Aversion in the small and in the large. Econometrica 32(1):122–136

    Article  Google Scholar 

  26. Rehima M, Belay K, Dawit A, Rashid S (2013) Factors affecting farmers’ crops diversification: evidence from SNNPR, Ethiopia. Int J Agric Sci 3(6):558–565

    Google Scholar 

  27. RISE (2010) Road to innovative social entrepreneurship (Pakistan). 2010 floods in Pakistan. The aftermath, the lives affected. http://rise-pak.com/flood-facts/. Retrieved 23rd August, 2012

  28. Robinson LJ, Barry PJ (1987) the competitive firm’s response to risk. MacMillan Publishing Company, New York

    Google Scholar 

  29. Shahbaz B, Ali T, Khan IA, Ahmad M (2010) An analysis of the problems faced by farmers in the mountains of northwest Pakistan: challenges for agricultural extension. Pak J Agric Sci 47(4):417–420

    Google Scholar 

  30. Shapiro BI, Brorsen BW (1988) Factors affecting farmers’ hedging decisions. North Cent J Agric Econ 10:145–153

    Article  Google Scholar 

  31. Sherrick BJ, Barry PJ, Ellinger PN, Schnitkey GD (2004) Factors influencing farmers’ crop insurance decisions. Am J Agric Econ 86(1):103–114

    Article  Google Scholar 

  32. Skees JR (1999) Agricultural risk management or income enhancement? Regul CATO Rev Bus Gov 22(1):35–43

    Google Scholar 

  33. Tanaka T, Camerer CF, Nguyen Q (2010) Risk and time preferences: linking experimental and household survey data from Vietnam. Am Econ Rev 100(1):557–571

    Article  Google Scholar 

  34. Tangermann S (2011) Risk management in agriculture and the future of the EU’s Common Agricultural Policy; ICTSD Programme on Agricultural Trade and Sustainable Development; Issue Paper No. 34; ICTSD International Centre for Trade and Sustainable Development, Geneva, Switzerland

  35. Tavernier EM, Onyango BM (2008) Utilization of farm management risk strategies at rural/urban fringe. Afr J Agric Res 3(8):554–565

    Google Scholar 

  36. Torkamani J, Rahimi MH (2001) Evaluation of farmer’s risk attitudes using alternative utility functional forms. J Agric Sci Technol 3(2001):243–248

    Google Scholar 

  37. Ullah R, Shivakoti GP (2014) Adoption of on-farm and off-farm diversification to manage agricultural risks: are these decisions correlated? Outlook Agric 43(4):265–271. doi:10.5367/oa.2014.0188

    Article  Google Scholar 

  38. Ullah R, Jourdain D, Shivakoti GP, Dhakal S (2015a) Managing catastrophic risks in agriculture: simultaneous adoption of diversification and precautionary savings. Int J Disaster Risk Reduct 12(2015):268–277. doi:10.1016/j.ijdrr.2015.02.001

    Article  Google Scholar 

  39. Ullah R, Shivakoti GP, Ali G (2015b) Factors effecting farmers’ risk attitude and risk perceptions: the case of Khyber Pakhtunkhwa, Pakistan. Int J Disaster Risk Reduct 13(2015):151–157. doi:10.1016/j.ijdrr.2015.05.005

    Article  Google Scholar 

  40. United Nations (2014) Agriculture and disaster risk. http://www.wcdrr.org/documents/wcdrr/prepcom1/UN/ATTFONWO.pdf. Retrieved 9th February 2015

  41. Velandia M, Rejesus RM, Knight TO, Sherrick J (2009) Factors affecting farmers’ utilization of agricultural risk management tools: the case of crop insurance, forward contracting and spreading sales. J Agric Appl Econ 41(1):107–123

    Google Scholar 

  42. WFP (2010) World Food Programme. September. Pakistan Flood Impact Assessment. http://vam.wfp.org.pk/Publication/WFP%20Pakistan%20Flood%20Impact%20Assessment_Sept%202010.pdf. Retrieved 28th July, 2012

  43. Yamane T (1967) Statistics, an introductory analysis, 2nd edn. Harper and Row, New York

    Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Raza Ullah.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ullah, R., Shivakoti, G.P., Kamran, A. et al. Farmers versus nature: managing disaster risks at farm level. Nat Hazards 82, 1931–1945 (2016). https://doi.org/10.1007/s11069-016-2278-0

Download citation

Keywords

  • Disaster risk
  • Risk management
  • Simultaneous adoption
  • Bivariate probit
  • Multinomial probit