Factors influencing crop diversification strategies among smallholder farmers in cotton production zone in Mali

Advances in Agricultural Science 06 (2018), 03: 01-16

Factors influencing crop diversification strategies among smallholder farmers in cotton production zone in Mali

Bandiougou Dembele 1,2*, Hillary K. Bett 1, Isaac Maina Kariuki 1, Marjorie Le Bars 3, Kevin O. Ouko 1

Department of Agricultural Economics and Agribusiness Management, Faculty of Agriculture, Egerton University, Box 536-20115, Egerton, Kenya.
Institut Economie Rurale (IER),  Sikasso,  Mali.
3 Institut de recherche pour le développement (IRD), UMR GRED, 34394 Montpellier cedex 5- France.


Smallholder farmers in Sub-Saharan Africa (SSA) seek to improve their livelihood through diversification of crops, livestock production systems and non-farm activities. This study was conducted in the cotton growing areas in Mali, which constitutes the major agricultural production zone in Mali. This study examines crops diversification strategies and identifies the main factors that influence diversification strategies. Cross-sectional data were obtained from 134 selected randomly smallholder farmers in three villages drawn from different agro-ecological zones in Southern Mali.  A multinomial logistics (MNL) regression model was used to estimate the main factors that determine diversification strategies of smallholder farmers. Results from MNL model revealed that the age of family head, education level, family size, oxen ownership, farm income per capita and crop pest significantly influenced families’ participation in the four diversification strategies. Policy interventions should be employed to encourage and promote profit-oriented activities through diversification strategies. Institutional interventions should be implemented to increase smallholder farmers access to inputs and both agricultural and non-agricultural credit.

Keywords: Diversification strategies, Multinomial logistics regression, Southern Mali

How to Cite: Dembele, B., Bett, H. K., Maina Kariuki, I., Le Bars, M., & Ouko, K. O. (2018). Factors influencing crop diversification strategies among smallholder farmers in cotton production zone in Mali. Advances in Agricultural Science6(3), 01-16.  


conomic development, food security and poverty alleviation in developing countries is directly linked to the agricultural sector (Pretty et al., 2011; Kassie et al., 2012; Mugendi, 2013 Fedoroff, 2015). However, agriculture in the developing countries is for subsistence and mainly undertaken by smallholder farmers who constitute over two-thirds of the global poor, food insecure and most vulnerable population (FIDA, 2011; FAO, 2014). Smallholder farmers in Sub-Saharan Africa (SSA) endured with low income from agricultural production and continue to struggle with food insecurity, poverty and climatic risks. Most of the governments in SSA are faced with the dilemma of achieving food security, while reducing poverty in the face of increasing population, climate change and the associated environmental consequences (Teklewold et al., 2013; Vanlauwe et al., 2014; Kuivanen et al., 2016; Binswanger-mkhize & Savastano, 2017). Although smallholder farmers in developing countries depend on rain-fed agriculture, they continue to contribute to improvement of rural and urban livelihoods. For instance, in SSA,  agriculture employs over 50% of labour force and contributes to an average of about 15% of the total gross domestic product (GDP) (OECD-FAO, 2016).

Smallholder farmers engage in multiple farm and non-farm activities in order to generate income, enhance food security and reduce poverty by utilizing their farms and selling surplus products (Meert et al., 2005; Rahut & Scharf, 2012; Senadza, 2012; Tittonell, 2013; Rahut et al., 2014; Wan et al., 2016). A majority of smallholder farmers undertake more than one activity and generate income from more than one source such as crop  diversification, which refers to a mix of farming systems rather than the shift from one given enterprise to another (Babatunde et al., 2008; Babulo et al., 2014; Abdullah et al., 2017). Participation in a mix of activities contributes to increased level of smallholder farmers’ incomes and maximizes their income (Khatun & Roy, 2012; Sultana et al., 2015). Farmers producing cash crops in the developing world diversify their agricultural production systems to increase their incomes, improve and maintain food security and reduce vulnerability to poverty (Kanyua et al., 2013; Schroth & Ruf, 2013; Mulwa  et al., 2017).

Agricultural diversification is one of the strategies for income generation, poverty and food insecurity reduction and improvement of nutritional status of rural population (Ellis, 2000; Barrett, 2001; Reardon, 2001; Fabusoro et al., 2010; Makate et al., 2016; Barrett, 2001; Ellis, 2000; Fabusoro et al., 2010; Reardon, 2001). Diversification involves growing more than one crop and, at the same time, practicing livestock production in order to increase income and enhance livelihoods. However, food crop production, the primary income generating enterprise in rural areas in SSA, is inadequate to enhance the well-being of smallholder farmers. In addition, its contribution to rural livelihoods is hampered by high cost of production (Woldenhanna & Oskam, 2001; Abimbola & Oluwakemi, 2013; FAO, 2014; Gautam & Andersen, 2016). This is attributed to low input use, low mechanization and poor soil fertility which lead to low agricultural output (Baquedano et al., 2010; Mariano et al., 2012; Sheahan & Barrett, 2017; Asfaw et al., 2018). Although agricultural diversification reduces production-related risks and increases farm earnings, few farmers diversify their agricultural activities in SSA. The lack of access to agricultural inputs, equipment and other factors of production as well as institutional constraints are important obstacles to diversification (Kasem & Thapa, 2011; Nguyen, 2017). Low or lack of diversification causes a decline in production of important commodities such as cash crops (cotton), food crops (maize, millet and sorghum) and livestock products. This results in low income, food insecurity and increased poverty levels among smallholder farmers (Ellis & Freeman, 2004; Okoboi & Barungi, 2012; Makate et al., 2016). Hence, agricultural diversification is important for the improvement of smallholder farmers’ livelihoods because of its potential of providing a reliable source of food and income in rural areas.

Several studies have attempted to describe the factors that may influence smallholder farmers in developing countries to diversify. Results indicate that education level, farmers resource endowment, agro-ecological and institutional factors constitute major constraints to farm and non-farm income diversification (Jansen et al., 2006; Mariano et al., 2012; Rahut & Scharf, 2012; Piya & Lall, 2013). The understanding of smallholder farmers’ decisions to participate in a particular strategy from among the available choices should put into consideration the enabling factors or constraints. Smallholder farmers’ choices for agricultural diversification are determined not only by agricultural production systems but also by low soil fertility, climate conditions and income among others. Likewise, it is rare for farmers in the rural areas of developing countries to sustain their livelihoods from one source of income. Most farmers in rural areas depend on a diverse portfolio of activities. Moreover, smallholder farmers may engage in crop and livestock production in order to overcome food insecurity and poverty (Ellis, 2000; Ellis & Freeman, 2004; Mariano et al., 2012).

Smallholder agriculture dominates the agricultural sector in Mali. About 80% of Malian farms are smallholder farms (Staatz et al., 2011). Southern Mali is the major agricultural production region in Mali. The region covers 12% of Malian geographical territory and is occupied by about 7.4 million people General Census of Housing and Population (RGPH,2009). This is around 51% of a total population of 14.5 million people (RGPH, 2009). The importance of Southern Mali in Malian agricultural production is due to its climatic conditions. Southern Mali zone supplies the rest of country with agricultural products. Cotton is the major cash crop grown in the region and constitutes the main source of income. The region is the Malian food basket and provides the majority of livestock products in the country (Ba et al., 2011). However, agricultural productivity in the region is hampered by limited resources and dependency on rainfall. Furthermore, despite the importance of cotton to the national economy, smallholder farmers in the region face several livelihood and agricultural production challenges. These include low access to agricultural inputs, climate change, food insecurity, poverty and malnutrition. Many of these challenges have been depicted in the literature as major factors that hinder the socioeconomic progress of smallholder farmers in SSA (Woldenhanna & Oskam, 2001; Teklewold et al., 2013; Meraner et al., 2015).

Despite the critical role played by Southern Mali in the Malian agricultural growth, little attention has been drawn to understanding factors that drive smallholder farmer diversification of income sources. A few studies have attempted to establish the socioeconomics determinants of production (Abdulai & Crolerees, 2001). Also, several studies have been conducted, focusing on agriculture and climate condition in cotton belt of Mali (Sanogo et al., 2016; Traore et al., 2013; Traore et al., 2015). Therefore, this paper analyses the factors affecting crop diversification strategies amongst smallholder farmers in cotton growing zone of Mali using Multinomial logistic regression. The study contributes to the understanding of crop diversification in cotton growing of Mali. It is underlined the following hypotheses; socioeconomic factors have no significant influence on the level of diversification in cotton growing zone of Mali and institutional factors have no significant effect on the level of diversification in cotton growing zone of Mali.


Methodology of research

Study area

The study was conducted in Southern Mali. The selection of the study area was justified based on a number of reasons. Southern Mali has a well-developed and diversified agricultural sector compared to other regions in Mali. Secondly, the region receives enormous public and private investment in agriculture. For instance, the Malian Company of Textile Development (CMDT) which is in charge of cotton production has over the years encouraged agricultural production through promotion of smallholder farmer access and use of farm inputs. The access and use of technologies such as improved seeds, manure, crop residues, composting among others have resulted in significant improvement in smallholder agriculture compared to other regions. Lastly, Malian agriculture is rain-fed and Southern Mali has favourable agro-ecological conditions that support diverse agricultural productions systems.

Three villages that were randomly selected from three districts located in different agro-ecological zones in cotton growing areas in Mali (Figure1). Beguene (old basin) is a village in Bla district in the northern part of Southern Mali. The main agricultural enterprise in Beguene is based on cotton, which is the grown as cash crop, maize, sorghum, millet and livestock (cattle, sheep and



Figure1. Location of villages in the study area
Source: (Traore et al., 2016)



Table 1. Description of explanatory variables and expectation of sign

Explanatory variables Description Expected sign
Age Age of family headed +/-
Family size Number of family members +
Education Level of formal education +
Farm income Income from all farming activities +
Non-farm income Income from all off-farm activities +
Oxen (livestock) Oxen ownership +
Land size Land size (ha) +
Land ownership Land ownership +
Lagriinput Low agricultural inputs used
priceinputs Agricultural input prices
Crop pest Crop pest infestation +
Access to credit Access to agricultural  inputs credit +
Extension services Access to agricultural extension services +


goat) production. Beguene is characterized by high population pressure on natural resources and low soil fertility. The area receives an average annual rainfall of around 850 mm, with high inter-annual variability. The second village, Ziguena (intermediate or centre), is located in Sikasso district and is characterized by low population pressure on natural resources. Ziguena receives an annual rainfall of about 1000mm. Cotton, maize, sorghum, millet and livestock production are the dominant agricultural activities in Ziguena. The last village is Nafegue which is located in Kadiolo district in Southern part of the cotton growing zone. Nafegue receives an average rainfall of more than 1200mm per year. Agricultural activities in Nafegue include cotton, maize, millet, sorghum and livestock production. Besides the main agricultural enterprises, there is also rice, groundnut and cowpea production in cotton growing zones in Mali.


Sampling procedures and data collection

This study is based on cross-sectional survey data that were collected from farming families across three villages located in different agro-ecological areas in the cotton growing zones in Mali. The study used multistage sampling technique. It involved a combination of purposive, stratified and simple random sampling procedure. The first stage involved purposive selection of three districts in the cotton growing area. In the second stage, one commune from each district was randomly selected, thus obtaining a total of three communes. In the third stage, a source list was obtained from the cotton producer cooperatives. Farming families were then randomly sampled using proportionate to size sampling procedure. The total population of farming was 66, 67, and 69 for Beguene, Ziguena, and Nafegue respectively.  A total sample of 134 farming families was selected from the three villages using the formula by Yamane (1967).

where n is the sample size, N is the total number of families in the three villages, e is the level of significance at 5% confidence level and 1 is the constant value.  The unit of study was based on farming families where only men are involved and registered in cotton cropping in Mali and possess factors of production. This is attributed to the criteria of admission to the cooperatives. For instance, it is mandatory for the member to allocate at least 0.5 ha of land under cotton. In the study area, women do not own land.

Data were then collected using structured questionnaires through one on one interview. Data were collected on the socio-economic factors (gender, age, education level, livestock ownership, land ownership and productive physical assets), institutional factors (access to agricultural inputs, credit and extension services) and bio-physical factor (field location). The analytical techniques used included descriptive statistics such as mean and standard deviation and multinomial logistics regression.


Econometric model specification and estimation of multinomial logistic regression

Econometric models such multivariate probit or logit, multinomial probit, nested logit, conditional fixed effects logit, among others are useful for analysis categorical outcomes. In the study, multinomial logit (MNL) is appropriate for analysis of categorical dependent variables when farmers have to choose only one outcome from among the set of crop diversification strategies. As opposed to MNL, the above econometric models allow the possibility of simultaneous choice of dependent outcome (Greene, 2002; Wooldridge, 2002). The choice of the MNL model was also backed by previous related studies that applied the same model in estimating the effects of socioeconomic and institutional factors on crop diversifications ( Jansen et al., 2006; Rahut et al., 2014;Belaye et al., 2017).

Multinomial logistics (MNL) regression is an analytical method that is commonly used to analyse smallholder farmers’ choices of agricultural strategies. It allows analysis of decisions across more than two categories (Greene, 2002; Wooldridge, 2002). The MNL model is also used in assessing the choice of alternative combinations of strategies in smallholder crop and livestock production systems (Babulo et al., 2014; Deressa et al., 2009). The model is used to analyse the factors that influence smallholder farmers’ decisions to diversify crop enterprises in cotton growing zone of Mali.

The Simpson Index of Diversity (SID) was used to characterize and measure the degree of diversification at smallholder farmer level (Joshi et al., 2003; Fabusoro et al., 2010; Ahmed et al., 2015). The measure of diversification is based on the area (hectare) of land under the main crops (cotton, maize, millet and sorghum). Livestock owned by smallholder farmers is expressed in terms of Tropical Livestock Units (TLU). The SID ranges between zero (0) and one (1) where 0 denotes specialization and 1 means extremity of diversification. The SID general formula is given as:

where SID denotes Simpson’s Index of the Diversity, is the proportion of enterprises coming from ,  is number of enterprises Following the random utility model (RUM), we assume that smallholder farmers aim to maximize their income, , by comparing the income generated by alternative strategies. The expected income,  that the smallholder farmer derives from engaging in strategy is a latent variable determined by the observed farming family characteristics  and unobserved . Therefore,

Where  is the parameter associated with that remains constant across alternatives and is a random disturbance term that capture intrinsically random choice behaviour, measurement or specification error and unobserved attributes of the alternatives.

To describe the MNL model, let  (j=1,2,3,4) denote the probability associated with cropping activity choices of a smallholder farmer with:

if the smallholder farmer combines cotton plus maize,  if the smallholder farmer combines cotton, maize and millet,  if the smallholder farmer combines cotton, maize, millet and sorghum  and  if the smallholder farmer only practices food crop production.

Following Greene (2002), the MNL model is given as:

Given a convenient normalization that solves the indeterminacy problem inherent in equation 3 is , then MNL model can be rewritten as:

Where  is a vector of coefficients on each of the independent variables . This can be estimated using maximum likelihood. For this study, the diversification strategies or their probabilities are described in Table 2. Crop diversification strategies. Unbiased and consistent parameter estimates of the MNL in Eq. (4) require the assumption of independence of irrelevant alternative (IIA) to hold. Specifically, the IIA assumption requires that the probability of using certain diversification strategies by the smallholder farmer needs to be independent of choosing other strategies. The premise of the IIA assumption is the independent and homoscedastic disturbance terms of the basic model.


Description of explanatory variables     

The study grouped the explanatory variables into three categories Table 1. First, family characteristics include the age, education level, family size, non-farm income and farm income. The second category of predictor variables is factor endowments which include ownership of oxen and land. Last, institutional variables include access to credit and agricultural extension services. The description of explanatory variables used in the MNL model is given below. The description of the predictor variables and the hypothesized direction of relationship with diversification are as follows.

Ages of family headed: Captures the age of the principal decision maker in years. Age of the household head plays an important role in diversification into several enterprises since it can be used to indicate farmer’s experience in different farming systems. It is expected that the age of the family head increases the probability of engaging in multiple agricultural enterprises.

Family size: Refers to the number of people working together and sharing a common pool of resources. Family size might positively influence diversification strategies in agriculture. The study hypothesizes that large family size increases the probability of smallholder farmers diversifying agricultural production systems. Large family size implies availability of labour which allows participation in multiple farm activities.

Education level of family head: Refers to the literacy level of the family head. It is considered as an important factor of agricultural diversification strategies. A family headed by a highly educated member is more likely to diversify into different agricultural enterprises owing to the available knowledge on different crop and livestock enterprises and the importance of diversification to family income and food security.

Farm and non-farm income: Indicates the financial position of smallholder farmers. We hypothesize that farm and non-farm incomes positively influence diversification because it enables farmers to have increase access to farm inputs. Increased access to inputs then enhances the level of diversification.

Oxen ownership: Livestock is an important physical asset because it represents an important source of capital for smallholder farmers. Oxen constitute an important part of agricultural systems in Southern Mali because it is the primary source of animal power. It is expected that ownership of oxen positively influences diversification into different crop production systems.

Farm size and land ownership: Land is an important factor in agricultural production and can be considered as a proxy for family wealth in Southern Mali. Possessing large sizes of arable land increases the probability of diversification into different cropping systems.

Access to agricultural inputs use: Access to farm inputs induces smallholder farmers to diversify their activities. The quantity of chemical fertilizer used per hectare is still low in Southern Mali due to the high prices. Smallholder farmers have to be members of cotton cooperatives so as to access important inputs.

Input prices: it is an important factor for crop diversification. It is hypothesized that high inputs prices negatively influence diversification due to liquidity constraints.

Crop pest: Crop pest is an important constraint in agricultural production. It is hypothesized that crop pests influence farmer decision to diversify. A farmer may diversify into different crops to cushion against output declines of another crop as a result of pest infestation.

Extension services: Extension services provide technical information and advice to farmers. It contributes to the dissemination of agricultural information, knowledge and skills, thereby enhancing the uptake of different agricultural technologies. It is expected that increased access to extension information and advice encourages diversification.


Results and discussions

Description of dependent variables

The dependent variable in the empirical estimation is the combinations of crop diversification strategies. The combinations that smallholder farmers were engaged in are provided in Table 2. The sampled smallholder farmers were grouped into four categories based on their diversification strategies. Accordingly, those who diversified into cotton and maize represented 24.63% of the sampled farmers. Cotton, maize and millet combination was practiced by 29.10% of the sampled farmers. A combination of cotton, maize, millet and sorghum was practiced by 39.55% of the sampled farmers. About 6.72% of the sampled smallholder farmers were only engaged in food crop production.


Field location characteristics

The average level of diversification was 0.6 implying that agricultural production systems were well diversified. Table 3 presents soil type in the study area based on farmer perceptions. Sampled farmers reported four types of soils: clay, sand, gravels and silt. There was also a combination of soil types which was dominated by sand and silt and silt and clay. About 9% of farmers’ fields are located on clay soils. This type of soil has a good potential for food and cash production. The gravel type of soil was reported by about 12% of the farmers and was ranked second after sand plus silt. The combination of two types of soils, sand plus silt, was the most dominant in the study area. About 74% of farmers’ fields were reported to be composed of the combination. The rest of fields were located on sand and silt plus clay at 4% and 1% respectively. The diversity of land allocated to crops across the different types of soils indicates that field location plays an important role in the diversity of cropping systems. On average, farmers had three fields of land. This represents about 72% of total cultivated land area of the sampled farming families.


Summary statistics

Table 4 provides the summary statistics of family characteristics, endowment and institutional variables used in the analysis. On average, the family head was 56 years old. The average farming family size was 23 people. The average number of oxen owned was 4 oxen per family. On average, the total land size under the main crops (cotton, maize, millet and sorghum) was averagely 11.51 ha per farming family. Smallholder farmers practiced various non-farm activities such as informal trade, traditional gold mining, casual work, and handcraft among others. The average annual earnings from non-farm activities was about USD 665.36 compared to annual farm income of about USD 60. Turning to education, about 51% of farmers had at least primary education while 49% reported that they did not attend formal school.

The survey results showed that about 46% of farmers thought that they used sufficient quantities of agricultural inputs against 54% who thought otherwise. About 77% of sampled farmers reported that the price of agricultural inputs was high. This could be attributed to the tendency of the cotton company to charge interest on the inputs supplied on credit to its contracted farmers. About 53% of sampled farmers indicated that cotton and maize outputs were severely affected by pests despite interventions by the cotton company.  Most of the cultivated land in the study area, about 85%, is owned by the family. About 15% of the cultivated land area is freely used (no rental fee) by immigrants and others related people. Turning to institutional

Table 2. Crop diversification strategies

Diversification strategies Percentage of farmers Share of cotton Share of food crop
Cotton +Maize 24.63 44 56
Cotton +2Food crop 29.10 50 50
Cotton +3Food crop 39.55 48 52
Food crop only 6.72 0 100



Table 3. Fields location per soil types in the study area (farmer perceptions)

Soil types Field1(ha) Field2 (ha) Field3 (ha) Percent  %
Clay 127 49 0 9
Gravels 202 42 1 12
Sand 71 16 0 4
Sand+ silt 1045 357 92 74
Silt + clay 17 11 2 1
Total 1462 475 94 100



Table 4. Summary and descriptive statistics of explanatory variables

Explanatory variable Mean S.D Description
Age of family headed 56 15.022 Continuous
Family size 23 19.112 Continuous
Oxen 4 2.851 Continuous
Nonfarm income 371269 744503 Continuous
Income per capita 33252 30576 Continuous
Farm size (ha) 11.51 8.31 Continuous
Qualitative descriptor Frequency Percent
Education Yes 66 51 Dummy
No 68 49
Low inputs used Yes 62 46 Dummy
No 72 54
Agricultural  price inputs Yes 103 77 Dummy
No 31 23
Crop pest Yes 71 53 Dummy
No 63 47
Land ownership Yes 114 85 Dummy
No 20 15
Access to credit Yes 107 80 Dummy
No 27 20
Access to extension services Yes 120 90 Dummy
No 14 10

Note: the exchange rate at the time of survey was 558 FCFA (Franc of the African Financial Community) for USD1


factors, about 87% and 90 % of the sampled smallholder farmers expressed that they had no problem accessing extension and credit services respectively. The farmers indicated that they were trained on good agricultural practices and crop production by field agents from the cotton milling company.


Econometrics analysis

Multinomial logistic regression was used to determine factors influencing choice of alternative diversification strategies. Food crop production was selected as the base category with full results from the MNL presented in Table 5. The model was highly significant as indicated by likelihood ratio test (LR (30)= 119.72, p = 0.000), suggesting strong explanatory power of the model. The variables were tested for multicollinearity using the variance inflation factor (VIF). The variance inflation factors for all variables were less than 10, indicating absence of multicollinearity. Out of the ten explanatory variables included in the model, four variables influenced farmers’ decision to diversify into a combination of cotton and maize. Five variables influenced diversification into a combination of cotton and two food crop (maize and millet) and cotton and three food crop (maize millet sorghum).


Age of family head

This variable was positively and significantly associated with the probability of diversification at 5% and 10% significance levels. Elderly farmers were more likely to either engage in cotton and maize, cotton and two food crops or cotton and three food crop production relative to only engaging in food crop production. This implies that the likelihood of diversification into several crop enterprise increases with age of the farmer. This probably implies that older farmers put much emphasis on family food security and increased income. Besides producing food crops for consumption purposes, older farmers engage in cash crop production in order to earn additional income. On the other hand, these three groups of smallholder farmers are large families and get inputs through their participation in cotton production. This allows farmers to apply a significant proportion of the inputs on maize, the main cereal crop. This also implies that the older heads of families are interested in maintaining cash crop production and also seek to feed the family while younger heads of families are mainly interested in ensuring food self-sufficiency. This finding is consistent with results of studies by Deressa et al. (2009), Teklewold et al.( 2013), Meraner et al. (2015) and Gautam & (Andersen, 2016) who found that age positively affected diversification decisions. On the other hand, Hassan & Nhemachena (2008) and Aneani et al.( 2011) indicated that the age of smallholder farmers increased the probability of diversification in order to improve livelihoods.


Family size

This variable had a positive and significant effect on the probability of diversification at 5% significance level. This implies that large families are more likely to grow a mixture of cash and food crops. Large families are able to engage in multiple cropping systems as compared to smaller families. In cotton growing zone of Mali, access to factors of production such as land and labour contribute to smallholder farmers’ decisions to diversify. Therefore, family size has a significant association with these diversification strategies in agricultural production systems. This finding is in line with results reported by Kassie et al. (2012), Piya & Lall (2013) and Babulo et al. (2014) who found that families with high labour availability are more likely to  diversify into several agricultural enterprises.



Table 5. Multinomial logistics regression model on crop diversification strategies

Explanatory variables Cotton+ Maize Cotton+2FC Cotton+3FC
Coef. P>z Coef. P>z Coef. P>z
Age 0,308 0.042** 0,297 0.049** 0,286 0.058*
Family size 1,171 0.033** 1,135 0.038** 1,167 0.033**
Education 4,574 0.218 7,612 0.038** 8,591 0.019**
Oxen 2,898 0.072* 3,331 0.039** 3,258 0.044**
Ln nonfarm 0,322 0.191 0,398 0.111 0,368 0.141
Lagriinput -3,237 0.239 -2,699 0.330 -2,360 0.396
hcostagriinput -4,711 0.404 -6,650 0.238 -6,879 0.222
Ln income per capita 0,500 0.067* 0,394 0.107 0,503 0.051**
Crops pest -4,852 0.121 -5,423 0.085* -4,736 0.133
Extension -6,514 0.995 -6,148 0.995 -5,294 0.996
Constant -21,167 0.983 -20,098 0.984 -22,159 0.982
Base category Food crop
Number of observation                                      134
LR Chi2(30) 119.72
Prob > Chi2 0.000
 Pseudo R2 0.357
Log likelihood -107.987

Note: **, * significant at 5% and 10 % of probability levels respectively.


Table 6. Hypotheses test statistics of model

Hypotheses Chi2 p- Value Decision
H1: 32.11 0.006*** H1  rejected
H2: 2.04 0.564 Fail to reject H2

Note: H1 = null hypothesis and = alternative hypothesis for socioeconomic factors and significant at 1% level and H2 = null hypothesis and = alternative hypothesis for access to agricultural extension services



Education level

Education is an important factor influencing diversification of livelihood strategies. The education level of the family head positively and significantly influenced diversification of crop enterprises. Well educated heads of family were more likely to engage in cotton and two food crop (maize and millet) and cotton and three food crop production compared to food crop growers. Similarly Onya et al. (2016), Zereyesus et al. (2016) and Asfaw et al. (2018) found that highly educated smallholder farmers were more likely to engage in several enterprises in order to improve their livelihoods. In addition, the finding is consistent with Jansen et al. ( 2006), Rahut & Scharf ( 2012) and Rahut et al. (2014)  who found a positive relationship between education level and income diversification strategies.


Oxen ownership

Livestock (oxen) had a positive and significant influence on the probability of diversification at 5% and 10% significance levels. An increase in the number of oxen owned by smallholder farmers increased the likelihood of diversifying into cotton and maize, cotton and two food crops and cotton and three food crops as opposed to engaging only in food crop production. Oxen constitute the main source of animal power for agricultural work. All agricultural work such as ploughing, seeding and transportation of farm input and output utilizes animal power. Oxen ownership permits cultivation of larger areas of arable land and provides manure, an important farm input. This explains the low use of tractor and others machineries in Southern Mali. This finding is consistent with Cunguara and Darnhofer ( 2011), Ghimire et al. (2014) and  Khonje et al. (2015) who suggested oxen provide animal power for ploughing in rural areas in SSA due to lack of tractors and small size of cultivated land area.


Farm income per capita

Farm income had a positive and significant relationship with diversification into cotton and maize and cotton and three food crops at 10% and 5% significance levels respectively. Higher incomes allow farmers to have access to critical productive resources such farm assets, inputs and land which increase the likelihood of crop diversification. The extra income earned by farmers from one crop is also important in providing financial resources that are used for diversification into other crops. This finding underlines results by (Basantaray & Nancharaiah, 2017) who indicated that crop diversity is strongly associated with significantly higher farm income.


 Crop pest

Threats to production (pests) were negatively and significantly associated with diversification into cotton and two food crops at 10% level.  The possible explanation for this negative direction in the relationship between crop pest and crop diversification is that pests cause crop damage which discourage diversification into cotton and two food crops. The food crop may be prone to the same pests, dis-incentivizing farmers from diversifying. This finding is inconsistent with Murrell (2017) who  reported that diversification of crops has a potential of suppressing and breaking down pest lifecycles.



Table 6 provides test results of the two hypotheses. The test statistics for hypothesis 1 (Chi2 = 32.11; p value = 0.006) suggest that socioeconomic factors considered in this study such as age of family, family size, oxen education, non-farm and farm income jointly influence crop diversification. Therefore, we reject the null hypothesis and conclude that socioeconomic factors significantly influence crop diversification. Further, the test statistic for Hypothesis 2 (Chi2 =2.04; p = 0.566) suggest that the coefficient of extension services is not statistically different from zero.  Therefore, we fail to reject the null hypothesis and conclude that extension services have no significant influence on crop diversification.


Conclusion and policy implications 

The study used cross-sectional data to analyse the factors that influence diversification strategies among smallholder farmers in Southern Mali. Agriculture in Southern Mali is characterized by mixed farming where cotton constitutes the main source of income. Findings show that farmers in Southern Mali engage in four diversification strategies such as cotton and maize; cotton maize and millet; cotton maize, millet and sorghum and food crop production. The share of cotton in the total cultivated land was 44%; 50%; 48% and 0% across the four systems respectively. The results also show that 74% of fields were located on sand and silt soils while the rest were spread across silt and clay, sand, clay and gravel soils at 1%; 4%; 9% and 12% respectively. Most of the results are reasonably consistent and in line with the previous studies. The MNL regression model revealed that the likelihood of diversification strategies is positively influenced by farmer and family characteristics and factors endowment. The estimates show that the ages of family head, education level, family size, oxen ownership, income per capita and crop pests significantly influenced smallholder farmers’ participation in the four diversification strategies. The results also indicate that smallholder farmers with larger family size were more likely to diversify into three diversification strategies compared to farmers only engaging in food crop production. Similarly, farmers owning oxen were more likely to diversify into cotton and maize. cotton and two food crops and cotton and three food crops.

The policy implication of these findings is to encourage smallholder farmers to diversify their agricultural production system in order to achieve food self-sufficiency and enhance family income. In addition, having a cash crop and engaging in food crop production and livestock rearing contribute to the reduction of extreme poverty, malnutrition and food insecurity. Policy interventions should encourage and promote better access to agricultural inputs and improve options for diversification. Future research has to consider implementation of adequate and sustainable agricultural technologies in order to provide pathways for diversification of crop enterprises.



“This material is based upon work supported by the United States Agency for International Development as part of the Feed the Future initiative under the CGIAR Fund award number BFS-G-11-00002 and the predecessor fund the Food Security and Crisis Mitigation II grant award number EEM-G-00-04-00013.” Special thanks to three anonymous reviewers for their helpful comments and suggestions on the first version of the paper. The assistance of technicians in data collection and farmers in three villages is well appreciated.



Abdulai, A., & Crolerees, A. (2001). Determinants of Income Diversification among Rural Households in Southern Mali Determinants of income diversification amongst rural households in Southern Mali. Food Policy, 26, 437–452. http://doi.org/10.1016/S0306-9192(01)00013-6

Abdullah, Rabbi, F., Ahamad, R., Ali, S., Ali, A., Ahmad, W., & Ilyas, A. (2017). Determinants of commercialization and its impact on the welfare of smallholder rice farmers by using Heckman ’ s two-stage approach. Journal of the Saudi Society of Agricultural Sciences, 10. http://doi.org/10.1016/j.jssas.2017.06.001

Abimbola, O. A., & Oluwakemi, A. O. (2013). Livelihood diversification and welfare of rural households in Ondo State , Nigeria. Journal of Development and Agrciultural Economics, 5(12), 482–489. http://doi.org/10.5897/JDAE2013.0497

Ahmed, T., Bhandari, H., Gordoncillo, P. U., Quicoy, C. B., & Carnaje, G. P. (2015). Diversification of rural livelihoods in Bangladesh. Journal of Agricultural Economics and Rural Development, 2(2), 32–38.

Aneani, F., Anchirinah, V. ., Owusu-Ansah, F., & Asamoah, M. (2011). An analysis of the extent and determinants of crop diversification by cocoa ( Theobroma cacao ) farmers in Ghana. African Journal of Agricultural Research, 6, 4277–4287. http://doi.org/10.5897/AJAR10.1083

Asfaw, S., Pallante, G., & Palma, A. (2018). Diversification Strategies and Adaptation Deficit : Evidence from Rural Communities in Niger, 101, 219–234. http://doi.org/10.1016/j.worlddev.2017.09.004

Ba, A., Lesnoff, M., Poccard, R. C., & Moulin, C.-H. (2011). Demographic dynamics and off-take of cattle herds in southern Mali. Tropical Animal Health and Production, 43(6), 1101–1109. http://doi.org/10.1007/s11250-011-9808-2

Babatunde, R. O., Omotesho, O. A., & Olorunsanya, E. O. (2008). Determinants of Vulnerability to Food Insecurity : A Gender- based Analysis of Farming Households in Nigeria. Indian Journal of Agricultural Economics, 63(1).

Babulo, B., Muys, B., Nega, F., Tollens, E. F., Nyssen, J., Deckers, J., & Mathijs, E. (2008). Household livelihood strategies and forest dependence in the Highlands of Tigray. Agricultural Systems, 98, 147–155. http://doi.org/10.1016/j.agsy.2008.06.001

Baquedano, F. G., Sanders, J. H., & Vitale, J. (2010). Increasing incomes of Malian cotton farmers: Is elimination of US subsidies the only solution? Agricultural Systems, 103(7), 418–432. http://doi.org/10.1016/j.agsy.2010.03.013

Barret, C. B., Reardon, T., & Webb, P. (2001). Nonfarm income diversification and household livelihood strategies in rural Africa: Concepts, dynamics and policy implications. Food Policy.

Barrett, C. B. (2001). Nonfarm income diversification and household livelihood strategies in rural Africa : Concepts , Dynamics and and policy implications. http://doi.org/10.1016/S0306-9192(01)00014-8

Basantaray, A. K., & Nancharaiah, G. (2017). Relationship between Crop Diversification and Farm. Agricultural Economics Research Review, 30, 45–58. http://doi.org/10.5958/0974-0279.2017.00021.0

Belay, A., Recha, J. W., Woldeamanuel, T., & Morton, J. F. (2017). Smallholder farmers ’ adaptation to climate change and determinants of their adaptation decisions in the Central Rift Valley of Ethiopia. Agriculture & Food Security, 1–13. http://doi.org/10.1186/s40066-017-0100-1

Binswanger-mkhize, H. P., & Savastano, S. (2017). Agricultural intensification : The status in six African countries. Food Policy, 67, 26–40. http://doi.org/10.1016/j.foodpol.2016.09.021

Cunguara, B., & Darnhofer, I. (2011). Assessing the impact of improved agricultural technologies on household income in rural Mozambique. Food Policy, 36, 378–390. http://doi.org/10.1016/j.foodpol.2011.03.002

Deressa, Tadesse, T., Hassan, R. M., Ringler, C., Alemu, T., & Yusuf, M. (2009). Determinants of farmers ’ choice of adaptation methods to climate change in the Nile Basin of Ethiopia. Global Environmental Change. http://doi.org/10.1016/j.gloenvcha.2009.01.002

Ellis, F. (2000). The determinants of rural livelihood diversification in developing countries. Journal of Agricultural Economics, 51(2), 289–302.

Ellis, F., & Freeman, H. A. (2004). Rural livelihoods and poverty reduction in four African countries strategies. Journal of Development Studies, 40, 1–30. http://doi.org/10.1080/00220380410001673175

Fabusoro, E., Omotayo, A. M., Apantaku, S. O., & Okuneye, P. A. (2010). Forms and Determinants of Rural Livelihoods Diversification in Ogun State , Nigeria. Journal of Sustainable Agriculture, 34, 417–438. http://doi.org/10.1080/10440041003680296

FAO. (2014). The State of Food and Agriculture.

Fedoroff, N. V. (2015). Food in a future of 10 billion. Agriculture & Food Security, 4(1), 11. http://doi.org/10.1186/s40066-015-0031-7

FIDA. (2011). New realities , new challenges : new opportunities for tomorrow ’ s generation.

Gautam, Y., & Andersen, P. (2016). Rural livelihood diversification and household well-being : Insights from Humla , Nepal. Journal of Rural Studies, 44, 239–249. http://doi.org/10.1016/j.jrurstud.2016.02.001

Ghimire, R., Wen-chi, H. U., & Hrestha, R. B. S. (2014). Factors Affecting Adoption of Improved Rice Varieties among Rural Farm Households in Central Nepal. Rice Science, 22(1), 35–43. http://doi.org/10.1016/S1672-6308(14)60278-X

Greene, W. H. (2002). Econometric analysis Seventh Edition.

Hassan, R., & Nhemachena, C. (2008). Determinants of African farmers’ strategies for adapting to climate change: Multinomial choice analysis, 2(1), 83–104.

IFPRI. (2004). Ending hunder in Africa: Prospects for the smallfFarmer. Policy Brief, International Food Policy Research Institute, WashingtonDC.

Jansen, H. G. P., Rodriguez, A., Damon, A., Pender, J., Chenier, J., & Schipper, R. (2006). Determinants of income-earning strategies and adoption of conservation practices in hillside communities in rural Honduras. Agriculutral Systems, 88, 92–110. http://doi.org/10.1016/j.agsy.2005.06.005

Joshi, P. ., Gulati, A., Birthal, P. S., & Tewari, L. (2003). Agriculture diversification in South Asia : patterns , determinants and policy implications No. 57. Washington, D.C.IFPRI MSSD. http://doi.org/10.2307/4415148

Kanyua, M. J., Ithinji, G. K., Muluvi, A. S., Gido, O. E., & Waluse, S. K. (2013). Factors influencing diversification and intensification of horticultural production by smallholder tea farmers in Gatanga District , Kenya. Current Research Journal of Social Sciences, 5(4), 103–111.

Kasem, S., & Thapa, G. B. (2011). Land Use Policy Crop diversification in Thailand : Status , determinants , and effects on income and use of inputs. Land Use Policy, 28(3), 618–628. http://doi.org/10.1016/j.landusepol.2010.12.001

Kassie, M., Jaleta, M., Shiferaw, B., Mmbando, F., & Mekuria, M. (2012). Technological Forecasting & Social Change Adoption of interrelated sustainable agricultural practices in smallholder systems : Evidence from rural Tanzania. Technological Forecasting & Social Change, 80(3), 525–540. http://doi.org/10.1016/j.techfore.2012.08.007

Khatun, D., & Roy, B. C. (2012). Rural Livelihood Diversification in West Bengal : Determinants. Agricultural Economics Research Review, 25(1), 115–124.

Khonje, M., Manda, J., Alene, A. D., & Kassie, M. (2015). Analysis of Adoption and Impacts of Improved Maize Varieties in Eastern Zambia. World Development MulT, 66, 695–706. http://doi.org/10.1016/j.worlddev.2014.09.008

Kuivanen, K. S., Alvarez, S., Michalscheck, M., Descheemaeker, K., & Groot, J. C. J. (2016). Characterising the diversity of smallholder farming systems and their constraints and opportunities for innovation: A case study from the Northern Region, Ghana. NJAS – Wageningen Journal of Life Sciences. http://doi.org/10.1016/j.njas.2016.04.003

Makate, C., Wang, R., Makate, M., & Mango, N. (2016). Crop diversification and livelihoods of smallholder farmers in Zimbabwe : adaptive management for environmental change. SpringerPlus. http://doi.org/10.1186/s40064-016-2802-4

Mariano, J. M., Villano, R., & Fleming, E. (2012). Factors influencing farmers ’ adoption of modern rice technologies and good management practices in the Philippines. Agricultural Systems, 110, 41–53. http://doi.org/10.1016/j.agsy.2012.03.010

Meert, H., Huylenbroeck, G. Van, Vernimmen, T., Bourgeois, M., & Hecke, E. Van. (2005). Farm household survival strategies and diversification on marginal farms. Journal of Rural Studies, 21, 81–97. http://doi.org/10.1016/j.jrurstud.2004.08.007

Meraner, M., Heijman, W., Kuhlman, T., & Finger, R. (2015). Determinants of farm diversification in the Netherlands. Land Use Policy, 42, 767–780. http://doi.org/10.1016/j.landusepol.2014.10.013

Mugendi, E. (2013). Crop diversification : A potential strategy to mitigate food insecurity by smallholders in sub-Saharan Africa Institute of Life Sciences , Sant ’ Anna School of Advanced Studies , and Kenyatta University. Journal of Agriculture, Food Systems and Comminity Development, 3(4), 63–69.

Mulwa, C., Marenya, P., Bahadur, D., & Kassie, M. (2017). Climate Risk Management Response to climate risks among smallholder farmers in Malawi : A multivariate probit assessment of the role of information , household demographics , and farm characteristics. Climate Risk Management, 16, 208–221. http://doi.org/10.1016/j.crm.2017.01.002

Murrell, E. G. (2017). Can agricultural practices that mitigate or improve crop resilience to climate change also manage crop pests? Current Opinion in Insect Science. http://doi.org/10.1016/j.cois.2017.07.008

Nguyen, H. Q. (2017). Analyzing the economies of crop diversi fication in rural Vietnam using an input distance function. Agricultural Systems, 153, 148–156. http://doi.org/10.1016/j.agsy.2017.01.024

Okoboi, G., & Barungi, M. (2012). Constraints to Fertilizer Use in Uganda : Insights from Uganda Census of Agriculture 2008 / 9. Journal of Sustainable Development, 5(10), 99–113. http://doi.org/10.5539/jsd.v5n10p99

Onya, S. C., Okezie, C. A., & Ejiba, I. V. (2016). Determinants of Participation in Non-farm Employment among Rural Farmers in Ebonyi State , Nigeria, 11(6), 1–11. http://doi.org/10.9734/JSRR/2016/28047

Piya, L., & Lall, K. (2013). Determinants of adaptation practices to climate change by Chepang households in the rural Mid-Hills of Nepal. Regional Environmental Change, 437–447. http://doi.org/10.1007/s10113-012-0359-5

Pretty, J., Toulmin, C., & Williams, S. (2011). Sustainable intensification in African agriculture Sustainable intensification in African agriculture. International Journal of Agricultural Sustainability, (February). http://doi.org/10.3763/ijas.2010.0583

Rahut, D. B., Ali, A., Kassie, M., & Marenya, P. P. (2014). Rural Livelihood Diversification Strategies in Nepal. Poverty & Public Policy, 6(3).

Rahut, D. B., & Scharf, M. M. (2012). Livelihood diversification strategies in the Himalayas *. Journal of the Australian Agricultural and Resource Economics, 56, 558–582. http://doi.org/10.1111/j.1467-8489.2012.00596.x

Sanogo, K., Binam, J., Villamor, G. B., & Dodiomon, S. (2016). Farmers’ perceptions of climate change impacts on ecosystem services delivery of parklands in southern Mali. Agroforestry Systems, 2–19. http://doi.org/10.1007/s10457-016-9933-z

Schroth, G., & Ruf, F. (2013). Farmer strategies for tree crop diversification in the humid tropics. Agronomy for Sustainable Development, 4. http://doi.org/10.1007/s13593-013-0175-4

Senadza, B. (2012). Non-farm income diversification in rural Ghana : Patterns and determinants. African Development Review, 24(3), 233–244.

Sheahan, M., & Barrett, C. B. (2017). Ten striking facts about agricultural input use in Sub-Saharan Africa. Food Policy, 67, 12–25. http://doi.org/10.1016/j.foodpol.2016.09.010

Staatz, J., Kelly, V., Boughton, D., Sohlberg, M., Skidmore, M., Diarrah, C. O., … Sako, M. (2011). Évaluation du secteur agricole du Mali.

Sultana, N., Hossain, E., & Islam, K. (2015). Income diversification and household well-being : A Case study in rural areas of Bangladesh. International Journal of Business and Economics Research, 3, 172–179. http://doi.org/10.11648/j.ijber.20150403.20

Teklewold, H., Kassie, M., Shiferaw, B., & Köhlin, G. (2013). Cropping system diversification , conservation tillage and modern seed adoption in Ethiopia : Impacts on household income , agrochemical use and demand for labor. Ecological Economics, 93, 85–93. http://doi.org/10.1016/j.ecolecon.2013.05.002

Tittonell, P. (2013). Livelihood strategies , resilience and transformability in African. Agricultural Systems. http://doi.org/10.1016/j.agsy.2013.10.010

Traore, B. B., Wijk, M. T. V. A. N., Descheemaeker, K., Corbeels, M., Rufino, M. C., & Giller, K. E. N. E. (2015). Mali : Learning from farmer perceptions and on farm trials, 51, 615–634. http://doi.org/10.1017/S0014479714000507

Traore, B., Marc, C., Wijk, M. Van, Rufina, C. M., & Giller, K. E. (2013). Effects of climate variability and climate change on crop production in southern Mali. European Journal of Agronomy, 115–125. http://doi.org/10.1016/j.eja.2013.04.004

Vanlauwe, B., Coyne, D., Gockowski, J., Hauser, S., Huising, J., Masso, C., … Asten, P. Van. (2014). Sustainable intensification and the African smallholder farmer. Current Opinion in Environmental Sustainability, 8, 15–22. http://doi.org/10.1016/j.cosust.2014.06.001

Wan, J., Li, R., Wang, W., Liu, Z., & Chen, B. (2016). Income diversification : A strategy for rural region risk management. Sustainability MDPI, 1–12. http://doi.org/10.3390/su8101064

Woldenhanna, T., & Oskam, A. (2001). Income diversification and entry barriers : evidence from the Tigray region of northern Ethiopia. Food Policy, 26, 351–365.

Wooldridge, J. M. (2002). Econometric Analysis of Cross Section and Panel Data.

World, B. (2008). Agriculture Development: World development report 2008, Washington,DC.

Yamane, T. (1967). Statistics: An Introductory Analysis, 2nd Edition., New York: Harper and Row.

Zereyesus, Y. A., Shanoyan, A., Ross, K. L., & Amanor-, V. (2016). The determinants of household-level well-being in Northern Ghana. Development Studies Research, 3(1), 1–10. http://doi.org/10.1080/21665095.2016.1229128