Agricultural innovations and food security in Malawi: Gender dynamics, institutions and market implications

Technological Forecasting & Social Change 103 (2016) 240–248

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

Agricultural innovations and food security in Malawi: Gender dynamics, institutions and market implications Munyaradzi Mutenje a,⁎, Henry Kankwamba b, Julius Mangisonib b, Menale Kassie a a b

International Maize and Wheat Improvement Center (CIMMYT), Harare, Zimbabwe Bunda College of Agriculture, P.O. Box 219, Lilongwe, Malawi

a r t i c l e

i n f o

Article history: Received 14 March 2014 Received in revised form 19 September 2015 Accepted 3 October 2015 Available online xxxx Keywords: Agricultural innovations Food security Farm households

a b s t r a c t The main objective of this paper was to analyze the driving forces that enhance farm households' decision to adopt agricultural innovations and the implications of these decisions on household food security. Maize variety diversity, soil and water conservation and improved storage or combinations of these accounted for 98% of agricultural innovations followed by the farmers in the study area. Using data from 892 randomly sampled households obtained from six districts of Malawi, the research employed a maximum simulated likelihood estimation of a multinomial endogenous treatment effect model to account for unobservable heterogeneity that influences technology adoption decision and maize productivity. Results revealed considerable heterogeneity in the choice of agricultural innovations practiced by smallholder farmers ranging from none to all practices within their fields. For instance 24% adopted improved maize varieties and storage; 14% improved maize, soil and water conservation, 14% improved maize only and 36% practiced all the technologies while 12% practiced none. In addition, the results showed that spouse's education, marital status, religion and informal networks are important factors in shaping women's participation in agricultural technology choice decisions. Exposure to production shocks such as drought, access to input and output markets, land size and gender of the plot manager of the plots explained the variation in farmers' propensities to adopt agricultural innovations. Respondents from drought prone areas, with small land size had higher incentives to adopt all the agricultural technologies as risk minimizing strategies. Conversely, farmers from high potential regions with bigger land sizes and higher asset and crop diversity indexes were less likely to adopt these agricultural innovations. Overall, adoption of improved maize and storage technologies resulted in significant increase in maize output per unit area though it may be important for researchers and policy makers to understand the social and institutional settings in which the technology is targeted, to benefit both men and women farmers equally. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Technology change has been widely acknowledged as a critical component of agricultural development and economic growth specifically in countries with agro-based economies and large concentration of agricultural households among the poor such as Malawi (World Bank, 2008; Diao et al., 2010). Agricultural technologies can provide a potential means of increasing crop production, improving household food security and subsequently raising incomes of farmers. Agricultural innovation involves the continuous use of new and existing knowledge emanating from diverse sources within and outside research domains to improve food production and household welfare (Spielman, 2005; Hall, 2010). Agricultural innovations which include adoption of improved agricultural practices, crop varieties, inputs and associated products such as crop insurance, have the potential to improve household food security and contribute to economic growth among the poor ⁎ Corresponding author. E-mail address: [email protected] (M. Mutenje).

http://dx.doi.org/10.1016/j.techfore.2015.10.004 0040-1625/© 2015 Elsevier Inc. All rights reserved.

particularly in southern Africa. Agricultural innovations are facilitated by diverse interactions between men and women; shaped by institutions, practices, behaviors and social relations that direct scientific research and technological change and the ultimate socio-economic goal (World Bank, 2008). It is also important to note that economic capacities and incentives are gender differentiated in ways that affect food availability and access, resource allocation, labor productivity and welfare within the household (World Bank, 2005; Quisumbing and McClafferty, 2006). According to FAO (2011) gender inequalities constrain women more than men in competitiveness and entrepreneurship. These gender differences have implications on agricultural research and innovation in terms of flexibility, responsiveness and dynamism. It is recommended that improving access to requisite resources (such as land, seed and fertilizer) for rural women to the same extent as men would increase agricultural production by 20% (Bardisi et al., 2007; DFID, 2007). Thus understanding the dynamic processes of technology change related to gender and agriculture innovation is fundamental in order to enhance faster and sustained agricultural growth, particularly in subSaharan Africa where gender disparities

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

tend to be greatest among the poor (Mason and King, 2001). The social dynamics around different activities and roles that poor communities engage in to address their economic needs through agricultural production systems shows the gender dimension of agricultural innovation. More importantly the interrelationships emanating from social dynamics of a society form a significant component of social capital that drives technological improvements and adoption. Paucity of literature on gender related agricultural innovations and its implications for increased food production is widely acknowledged (Kakooza et al., 2005; Nompumelelo et al., 2009; Meinzen-Dick et al., 2011; World Bank, 2008; Blake and Hanson 2005). Given the inter-causal relationships that exist between men and women in the different activities of the agricultural production cycle, there is need for planners, policy makers, implementers and researchers to focus both on men's and women's roles in agricultural activities simultaneously rather than as separate entities. A policy tool that addresses these diverse challenges, while maximizing on the available innovative opportunities for men and women, will be very useful. Food security is commonly defined as a situation “when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life” (FAO, 1996). Thus, the nutritional dimension is integral to the component of food security (FAO, 2009).It is important to understand the multidimensionality of the food insecurity issue in search of effective, and comprehensive solutions which are vital for improving nutrition. Food security is a fundamental need for individuals to realize both their maximum physical and intellectual potential. It is the basis for the well-being of individuals and households and for human capital formation and, thus it is vital for economic development. Food insecurity has serious, long-lasting economic consequences at the micro and macro levels. Malnutrition and illness reduce household income earning ability; perpetuate poverty, and slow economic growth through direct losses in productivity from poor physical and mental performances and indirect losses from reduced working and cognitive capacity and related deficits in schooling, and losses in resources due to increased health care costs. Even transitory food insecurity can cause irreversible health impairments, particularly in children, limiting the development potential of future generations (World Bank, 2006). Household food security is dependent on agricultural production, food imports and donations, employment opportunities and income earnings, intra-household decision-making and resource allocation, health care utilization and caring practices. Household food security also depends on the characteristics of the decision maker and gender roles, information and education, cultural and social customs (Thomas and Frankenberg, 2002; Tanumihardjo et al., 2007; WHO/FAO, 2003). Hence, there is need to understand how agricultural innovations can be leveraged for improving food security given the complex social and institutional environment under which these innovations occur in Malawi. The objectives of this paper are to identify the pathways and the extent to which agricultural innovation contributes to household food security in a complex social and institutional environment. 2. Methodology 2.1. Conceptual framework Smallholder farmers in Malawi produce and consume a number of maize and legume varieties. Their decision about crop combinations and varieties to grow, agricultural technologies to implement, and the amount of land to allocate to each crop and variety can be explained by household economic theory (Becker, 1965; Sadoulet and de Janvry, 1995). In this theory, due to imperfect input and output markets, a household acts as a unified unit of production and consumption of goods and services with the aim to maximize expected utility. Market imperfections directly influence farm household's investment and

241

production decisions. For example capital market imperfections limit households to their savings and already accumulated capital assets such that low resource endowed smallholder farmers are not able to invest in capital-intensive technologies. Similarly imperfect rural labor market structure, information asymmetry and high transaction cost imply that only larger households are able to invest in labor-intensive technologies (Pender and Kerr, 1998). In many developing countries including Malawi output markets for food grains are highly seasonal and underdeveloped. Empirical evidence show that when output markets are highly imperfect or thin it discourages technology investments. Farmers will opt for technologies that improve food supply and access stability. In such situations, non-separable household models that partially or fully incorporate input and output market imperfections are suitable for modeling household decisions and resource allocation. Our theoretical framework mainly draws from Becker (1965) and Sadoulet and de Janvry (1995). In our model, households are assumed to rely primarily on agriculture. They maximize utility in a specific period t, Ut, which is assumed to be a concave. Utility depends on the consumption of agricultural commodities (ca), manufactured goods (cm), leisure (cl), subject to household characteristics affecting preferences (zh) maxU ðca ; cm ; cl j zh Þ:

ð1Þ

The farm household utility function is subjected to three constraints, a convex, continuous production function, assuming that quantity of maize produced (qa) depends on the selected agricultural technology xj, family labor input available (l), agricultural knowledge(ka) acquired through experience or observing other neighboring farmers using the selected technology, and fixed inputs such land and capital (zf)
 g qa ; x j ; ka ; l j z f ¼ 0:

ð2Þ

It is through the production function that households are differentiated as innovating households (i.e. those that adopt modern technology or not). Adopting modern technology represents picking a production plan that represents a production possibility set that maximizes output associated with qa , x , l ∈ Q. The second constraint relates to household's labor allocation into agriculture, off-farm activities and leisure which cannot exceed the household endowment (l). Finally, the household has a standard budget constraint such that the total household expenditure, measured using market prevailing prices should be less than the net income from agriculture, off farm income generating activities, other income sources (e.g remittances and pensions) and net savings. Becker (1965) laid the foundation for household models while Sadoulet and de Janvry (1995) extended the model to make it a producer, worker and consumer model. This earlier work did not discuss the role that technology could play in altering the outcomes of farm output and eventually household utility. We attempt to indicate that given an isoquant Q = f(x i , … , x j ) = {x ∈ ℝ + n : x ∈ V(q) and x ∉ V(q′) ∀ q ′ N q}, when a new technology becomes readily available, the farm household will attempt to adjust adoption behavior of other technologies in such a way that ∂f ðxÞ ∂f ðxÞ dxi þ dx j ¼ 0: ∂xi ∂x j

ð3Þ

Hence, the household will substitute the technologies in such a way that their technical rate of substitution becomes ∂f ðxÞ dx j ∂xi ¼− : dxi ∂f ðxÞ ∂x j

ð4Þ

242

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

Thus, such a substitution will necessitate farm households choosing a particular technology and foregoing the other. Given factor prices, px in Eq. (4), the rate of substitution between the two technologies could be presented as ∂f ðxÞ pi ∂xi ¼ ; − pj ∂f ðxÞ ∂x j

∀i; j ¼ 1; …; k:

ð5Þ

Considering technology choices this way may present a pathway through which net returns and household utility could be maximized since the objective function is affected by factor prices such as wages, input and output prices, agro-ecological conditions and household resource endowments. Since consumption and production decisions are inseparable, the household's optimal choice depends on farm size(A0), exogenous income(Y0), household characteristics (zh), biophysical (zr) and market characteristics (zm). h



¼



A0 ; Y 0 ; zh ; zr ; zm



ð6Þ

The household micro-economic theory states that each observed choice of consumption and production of the farmer is the optimum and the only feasible way to increase productivity and well-being when an individual is optimizing is through innovation (United Nations Development Programme, 2012). Eq. (6) motivates the econometric procedure discussed in the next section since agricultural innovation choice is a result of optimal household decision making. 2.2. Empirical model Probit regression model was employed to examine the factors that determine female core heads' participation in agricultural technology adoption decisions. The justification for the use of the probit model over the logit model is as a result of its ability to constrain the utility value of the participation decision within 0 and 1, and its ability to resolve the problem of heteroscedasticity (Green (2003). Participation in agricultural technology adoption decisions (Y) was captured as a dummy variable with the value of 1 assigned to female core heads who participated and 0 for otherwise. Following from Green (2003), the binary probit for the two choice models was specified as Y i ¼ β0 þ B1 X þ ε; → 1 ¼ if Y i N Y; 0 otherwise

treatment. Our treatment and outcome equations, thus, take the following structure: Eðyi jyio ; xio ; mi ; lÞ ¼ f ðγyio þ xio β þ γmi þ λli Þ 
 Pr mij ¼ 1jyio ; xi0 ; zi0 ; li ¼ g τyio þ xi0 ζ j þ zi0 α j þ δ j ; li

where mi is a vector of agricultural technology choices and mij is the jth agricultural innovation alternative; li is a vector of latent factors representing unobservable heterogeneity and λ and δj are vectors of factor loadings. Following the specification in Eq. (8) and assuming that y and m are conditionally independent, the joint treatment selection and outcome distribution, conditional on common latent factors is: Prðyi ; mi ¼ 1jyio; xi0 ; zi0 ; li Þ ¼ f ðγyio þ xio β þ γmi þ λli Þ  g τyio þ xi0 ζ j þ zi0 α j þ δ j li

ð9Þ

In order to estimate Eq. (9) we encounter a problem because of lij which is latent (unknown). The agricultural technology preference parameters are unobserved, thus the unconditional probability that the farmer will choose agricultural innovation alternative j is estimated by integrating the conditional probabilities over all values of each random technology preference coefficients weighted by its density function. We hence use maximum likelihood estimation following Deb and Trivedi (2006). We invoke the assumption that lij is independently and identically distributed, also follows standard normal distribution. If we denote the distribution as h then we can integrate lij out of the joint distribution in Eq. (9) as follows: Z Prðyi ; mi jyio ; xi0 ;Þ ¼

½ f ðγyio þ xio β þ γmi þ λli Þ   g τyio þ xi0 ζ j þ zi0 α j þ δ j li  dhðli Þ

ð10Þ

This maximization cannot be solved because the integral of Eq. (10) does not have a closed form solution as its dimension is determined by the number of βn that is random with non-zero variance. This was addressed by using maximum simulated likelihood based on Halton sequences as described in Deb and Trivedi (2006); Train (2003) and Bhat (2001). Following Deb and Trivedi (2006) the maximum simulated log-likelihood estimator is given by lnlðyi ; mi jyio ; xi0 ;Þ  X n o X 1 s n ≈ ln f ðγyio þ xio β þ γmi þ λli Þ  gðτyio þ xi0 ζ j þ zi0 α j þ δ j li g s s ¼ 1 i¼0

ð11Þ

ð7Þ

where Yi represents participation in agricultural technology adoption, X is a set of explanation variables, and ε is the error term. The set of explanatory variables used in the probit model, and their definitions used are presented in Table 3. We attempt to model agricultural innovation choices at household level using an endogenous multinomial treatment approach. The decision to choose a particular agricultural innovation can be considered as a binary variable. In that regard, treatment may be estimated using two-stage least squares method. However, Deb and Seck (2009) indicate that it results in substantial loss of information and efficiency than when you estimate using full information maximum likelihood procedures. We further notice that the decision to adopt agricultural technology procedures may vary given individual/farm characteristics and location amidst other unobservable factors (Doss and Morris 2001; Kristjanson et al. 2012). As such it would be important to model the choice of an agricultural innovation in a multinomial framework. Therefore, we model choice of an agricultural innovation strategy as a multinomial

ð8Þ

where mi is 1 if alternative j is chosen and 0 otherwise, s = 1, 2,…,S are random draws from the h density. We estimate Eq. (11), impact of different agricultural innovations on welfare given a set of socioeconomic variables, institutional factors and other farm level variables using mtreatreg in Stata 12.0. 2.3. Variables and descriptive statistics The outcome variable used in this study is the average maize yield per hectare in the 2009/2010 agricultural season. Average maize yield was chosen, as the use of improved agricultural technology may affect the household resource allocation, and hence affects the overall household food security situation not only food availability. Moreover maize production is the most important indicator of food security in this semi-subsistence agricultural system accounting for 97% of the food consumed in the household (National Statistical Office (NSO), 2008). A number of questions are investigated whether farm households had adopted any technology to meet the 4 key dimensions of food security. A household was described as an adopter, if the technology was

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

applied on at least 50% of the total share of the maize area for at least three years continuously. A dummy variable equal to 1 was created for households that had at least 50% of their total maize area with a given technology and 0 otherwise. Maize variety diversity, soil and water conservation and improved storage or combinations of these were the main agricultural innovations followed by the farmers in the study area. These technologies accounted for 98% of the agricultural innovations followed by the farmers who actually adopted an innovation. Improved maize seeds refer to use of certified seeds of maize either open pollinated varieties or hybrids. The distribution of adopters of agricultural technologies was as follows; 24% adopted improved maize varieties and storage; 14% improved maize, soil and water conservation; 14% improved maize only and 36% practiced all the technologies while 12% practiced none. The share of land area planted to improved maize was calculated as: A p ¼ Xi ; Ai

ð12Þ

where pi is the proportion of the ith crop; Ai is the area under the ith n

crop in hectares and ∑i¼1 Ai is the total cropped area in hectares. To calculate the maize variety diversity index and crop diversification, we employed the Simpson Index of Diversification (SID). According to Minot et al., (2006), it is an ecological measure of diversification. Joshi et al., (2003) defines the Simpson Index as: Xn SID ¼ 1− i¼1 p2i n

ð13Þ

where ∑i¼1 p2i is the HHI1 and 0 ≤ SID ≤ 1. The SID = 0 when there is complete specialization and pi = 1. Similarly, SID = 1 when there is perfect diversification with the number of shares declining as the number of cultivated crops increases. Thus if there are k crops, then the SID falls between zero and 1 − 1 / k(Minot et al., 2006). Empirical evidence has shown that the use of improved maize seeds can increase total factor productivity by 10% and account for as much as 50% of the yield increase (Okoboi 2010, World Bank, 2008 and de Janvry and Sadoulet, 2002). Despite more than two decades of improved maize seed promotion in the smallholder sector of Malawi, the proportion of maize area planted to improved varieties is still significantly low at 30% (Lunduka et al., 2012; World Bank, 2008). Nearly 14% of the sampled households had adopted improved maize only. The commonly practiced soil and water conservation strategies (SWC) included box ridges, soil bounds, and grass strips. Soil fertility decline through nutrient mining and soil erosion has been identified as the major contributing factor to household food insecurity in Malawi (World Bank, 2008). A number of non-governmental organizations, international research and public institutions have been promoting soil and water conservation strategies as a way of improving and stabilizing food production and supply in Malawi. Some studies (for example (FAO, 2001)) noted that SWC increases food production through increased land and labor productivity. Others such as Kassie et al. (2009), discovered that adoption of SWC does not contribute positively to improved food production when all inputs such as labor are factored into the economic analysis. Improved storage was another agricultural technology that was found to be of greater importance in maintaining household food supply stability and access. This refers to storage facilities that reduced postharvest loss such as using polythene bags fumigated with actellic super or storage facilities built with conventional or locally available materials equipped with insecticides and pesticides. Storage facilities are assumed to reduce household food expenditure and increase food access particularly during lean periods. A number of previous studies

1

HHI ¼

n

∑ p2i

i ¼ 1

243

in sub-Saharan Africa have shown that selection of agricultural innovations and food production is influenced by socio-economic factors such as risk reduction, desire to have stable food production, product diversity and resource endowments (Doss and Morris 2001; Sserunkuma 2005; Kristjanson et al. 2012). The descriptive statistics of these variables are given in Table 1.

2.4. Data source This study was part of the sustainable intensification of maizelegume in Eastern and Southern Africa (SIMLESA) project aimed at enhancing agricultural productivity and food security, whilst restoring the degraded environment. In southern Africa, Malawi was chosen as one of the pilot countries because of the importance of maize-legume crops in their farming system and the strong public policy support to smallholder agricultural development with particular emphasis on legumes as cash crops. Integrating these soil fertility management practices with strong policy support is expected to address some of the main challenges threatening future productivity. In Malawi the project focused on the central and southern regions. The baseline survey was done in 2010 covering 892 households using multistage random sampling probabilistic survey.2 Information from the households was gathered through structured questionnaires. The data covered information on socio-economic characteristics such as household composition, consumption expenditure, and income sources, plot and farm management practices, crop varieties and area planted, cost of production, yield data for different crops and varieties, livestock ownership, farm and non-farm assets and exposure to shock in the past ten years, as well as resource and market access. Data was also collected from 719 women that identified themselves as the primary female decision makers in the household such as wives in male head households, mother in laws, daughter in laws, daughters and sisters in female headed households. This included information on their level of participation in key decision making in the household, formal and informal networks within and outside the village.

3. Results and discussion 3.1. Descriptive analysis Table 1 presents descriptive statistics of demographic and farm characteristics of the households that adopted the different agricultural technologies as well as pairwise comparative analysis with the nonusers. The results highlighted that non-users of improved technologies for all the categories had significantly bigger cropped area, more diversified cropping patterns and access to credit compared to the users. The non-users of improved technologies were producing tobacco on contract basis. The result also showed that higher proportion of sampled household (N 80%) had experienced some shocks that affected crop production in the past 3 years. Non-users of improved technologies employed crop diversification as a way of dealing with production risks while adopters of agricultural technologies used various combinations of these technologies ranging from maize variety diversification to a combination of improved maize varieties, improved storage and soil and water conservation. Users of these improved agricultural technologies produced significantly more maize compared to the non-users. These results are in line with previous research in developing countries which asserts that farmers prone to production risks employ various methods to mitigate these risks (Bezemer et al., 2005; Barrett, et al.; Ellis, 2000).

2 The baseline survey covered 6 districts, 64 extension planning areas (EPA's), 89 sections, 235 villages and 892 households.

51.9 51.0 51.6 51.0 50.5 51.9 50.2 53.9 52.3 50.4 51.7

HH received credit (%) HH experienced any shock in the past 3 years (%) Information %

* Significant at the 10% level; ** Significant at the 5% level; *** Significant at the 1% level.

52.8** 40.6

50.7

8.7* 87.3 7.4 86.1 7.6 88.7 8.2 83.6 7.8 87.9 7.9 85.6 8.1* 86.0 7.4 87.7 8.3* 88.4 7.4 84.7 8.7** 88.1 6.5 88.1 8.5** 86.6 2.9 86.4

5.9* 1.8 1.7 0.30 0.119 6.3 84.8 1.1 3.6 0.5** 0.173 5.2 1.5 1.7 0.27 −0.068 6.2 88.2 1.5 3.0 0.42 0.121

Users Non user

5.7* 1.7 1.7 0.28 0.050 6.7 91 1.6 3.3 0.44 0.171 5.2 1.5 1.7 0.28 −0.069 5.7 81.3 1.1 3.0 0.44 0.098

Users Non user

5.8* 1.7 1.7 0.30 0.099 6.2 87 1.8 3.4 0.48** 0.164 5.2 1.6 1.7 0.27 −0.071 6.3 87 1.1 3.0 0.41 0.123

Users Non user

5.7* 1.6 1.7 0.30 0.024 5.4 82 1.5 3.3** 0.48** 0.143 5.1 1.6 1.7 0.24 −0.050 8.1 97 1.2 2.9 0.36 0.134

Users Non user

5.7* 1.8 1.7 0.29 0.069 7 89 1.2 3.4** 0.49* 0.185 5.2 1.5 1.7 0.27 −0.070 5.6 84.8 1.6 2.9 0.40 0.095

Users Non user

5.7* 1.6 1.7 0.31 0.035 5.4 80.5 1.1 3.4** 0.49** 0.150 5.1 1.7 1.7 0.24 −0.053 7.6 96.8 1.9 2.8 0.37 0.125

Users Non user

5.5* 1.6 1.7 0.28 0.012 6.4 86.0 1.1** 3.3*** 0.44* 0.150 5.0 2.1 1.6 0.22 −0.092 4.9 93.8 3.7 2.3 0.42 0.068

Variable definition

Years of formal education completed Proportion of defacto female head households Aggregated using adult equivalence scale Proportion of income from different sources Asset index Distance to input market (walking minutes) Distance to output market (walking minutes) Tropical livestock units Total cropped area (ha) Crop diversification index 1 if individual member of an agricultural association, 0 otherwise 1 if individual had access to credit, 0 otherwise 1 if the household experienced shock that affected production, 0 otherwise Household has access to information yes = 1

Variable name

Users

3.2. Social dynamics in agricultural technology innovation and adoption

Head's years of schooling Defacto female headed HH (%) Labor availability Income diversity index Asset index Distance to input market Distance to output market Tropical livestock units Total cropped area (ha) Crop diversity index Membership in association

Soil & water conservation only Improve storage and soil & water conservation Improved storage only Improved maize and soil & water conservation Improved maize varieties & storage Improved maize varieties (variety diversity) Household characteristics

Table 1 Household and farm characteristics by agricultural technology type adopted.

Non user

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

Improved maize varieties, improved storage & soil and water conservation

244

The empirical results in Table 2 revealed that both men and women are involved in decision making regarding resource allocation and agricultural technology innovation choice. It further revealed that both men and women have distinctive responsibilities and they are key decision makers in the spheres of production cycle that they control. For example, the decision of cash crop varieties such as tobacco, soybeans and cotton and fertilizer allocation was mainly the responsibility of men in all the districts except in Salima where men derived their income from fishing. While legume crop variety choice particularly groundnuts and cowpeas and storage technology choice was regarded as women's tasks. It is worth noting that Malawi presents a special scenario, more than half of the communities are matriarchal mainly found in the central and southern regions of the country and patriarchal societies are mainly found in the north and some parts of central region. Our result revealed that decision-making process in the farm household is influenced by the culture of the community to which the household belongs. In patriarchal societies such as Kasungu, decision on land improvement such as soil and water conservation technology is taken by men while in matriarchal societies men are formal decision makers but in reality women in the household greatly influence the decision on land improvements. Though the decision making regarding soil water conservation technologies was revealed to be jointly determined, focus group discussion with both men and women in the matriarchal societies confirmed that women core heads of households had significant influence on land improvement decision. Our results also showed that farm labor contribution was also influenced by inter- and intrahousehold dynamics. In households where farming was the main source of livelihood 60% of men indicated that they contributed fully to farming, whilst in households and societies such as Ntcheu and Salima where men were engaged in off–farm activities as sources of livelihood their contribution to farm labor was limited to those that were ascribe by society to be men activities. Overall our results revealed that though household heads could be considered as the main decision makers in this farming system, some decisions were made by men or women only, others were reached at after consultations with core household heads and other member(s) of the farm household. We further analyzed the factors that influenced women's participation in decision making of agricultural resource allocation and technology choices (Table 3). Probit regression empirical results confirmed the importance of intra-household dynamics. Women with higher intrahousehold decision-making power index and skills training influenced greatly the selection of agricultural innovations. Spouse's education level positively and significantly influenced women's participation in decision making of agricultural resource allocation and technology choice. Each additional year of spouse's education increases the probability of the woman's participation in key decision making by 18.3%. The result further revealed that marital status is an important factor in shaping participation in decision making regarding agricultural technology choices and resource allocation. They are able to indirectly influence their husbands' agricultural innovation choices and take an active role in household decision making. The results suggest that being married can be important for ensuring women's participation and building their self-confidence because they are better trusted and respected in these farming systems. However, belonging to the Muslim and traditional religion and informal social networks negatively and significantly influences women's participation in agricultural innovation decisions. Belonging to the Muslims or traditional religion decreased the probability of active women participation in agricultural resource allocation and technology choice decision by 6.9% and 7.2% respectively. The results imply that informal networks influence greatly the attitudes, perceptions, preferences and use of technologies and thus determine their choices. In line with Schuller et al. (2000) and Mayoux (2001) finding our results confirms that traditional informal networks stifle agricultural innovation. In societies where agricultural production is the mainstay

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

245

Table 2 Decision making on resource allocation and agricultural technology innovation choice. Districts

High agricultural potential areas Lilongwe

Resource/technology Land allocation Cash crop varieties Legume crop varieties Food crop varieties Credit Soil & water conservation technologies Fertilizer allocation Storage technology Food grain storage quantity Use of cash income Marketing Agricultural extension training

HHa 29 87 0 18 23 36 52 24 12 13 44 62

FCH 11 9 94 13 3 14 17 59 56 2 17 8

Own farm labor contribution % 100 75 50

60 6.8 33.2

95 5 0

Low agricultural potential areas

Kasungu Jointly 60 4 6 69 74 50 33 17 32 85 39 30

HH 16 73 2 15 38 52 57 33 31 49 56 73

FCH 6 10 79 44 9 8 10 42 16 21 12 11

61 6.7 31.3

98 2 0

Mchinji Jointly 78 17 19 41 55 40 33 25 53 30 34 16

HH 36 42 23 22 35 26 37 33 16 12 33 31

FCH 28 30 51 25 21 15 24 53 73 10 28 26

62 17 21

92 4 4

Ntcheu Jointly 36 28 26 53 44 59 39 14 11 78 39 43

HH 23 28 6 12 14 18 58 9 14 14 13 17

FCH 37 41 87 67 8 39 23 77 64 24 61 64

57 14 29

96 3 1

Balaka Jointly 40 31 7 21 78 43 19 14 22 62 26 19

Salima

HH 39 61 9 23 67 26 87 48 59 82 89 77

FCH 5 2 78 46 11 18 3 26 24 8 7 9

58 8 33

97 3 0

Jointly 56 37 13 31 22 56 10 26 17 10 4 14

HH 33 13 4 9 10 41 26 12 9 48 17 43

FCH 57 76 88 78 56 29 43 66 87 16 63 44

37 12 51

93 6 1

Jointly 10 11 8 13 34 30 31 22 4 36 20 13

a HH refers to household head; FCH primary female decision-maker in the household, such wife of the male household head, mother in law, daughter in law, sister, and daughter in female headed households.

of economic production, informal insurance networks, and conformity with social norms have tangible impacts on farmers' livelihoods (Läpple and Kelley 2013). Overall, our empirical results suggested that gender dynamics, socio-economic and cultural settings influence agricultural innovation decisions. These results further suggest that it is important for researchers and policymakers to understand the social and institutional context in which the technology is targeted, to ensure that the technology will benefit both men and women farmers equally. 3.3. Impact of agricultural technology innovations on household food security The endogenous multinomial treatment-effects regression was fitted on 891 observations. One observation was dropped since it had missing values in all variables. Estimation was done after some rigorous exploratory data analysis procedures to remove outliers, leverage points

and influential observations. Results of the estimated model are shown in Table 4. The model was significant at 1% given a Wald chi-square statistic of 21,153.76 with 50° of freedom. This implies that at least one of the independent variables considered was significant either in the selection or the outcome equations. 3.3.1. Factors affecting choice of agricultural innovation strategy The mixed multinomial logit component of the model was used to determine factors affecting choice of an agricultural innovation strategy. The multinomial treatment variable was transformed by combining the “improved storage only (IS)” (n = 49) and “soil and water conservation only (SWC)” (n = 18) with the base category of “non-adopters (NA)” (n = 16) since there were too few observations in these categories such that treating them separately would make the model unstable due to negative degrees of freedom. These categories combined form our base category while the “improved maize & soil and water conservation

Table 3 Factors influencing women's participation in agricultural technology innovation choice. Variable

Description

Depend variable participation in Individual characteristics Educational level Marital status Wife's decision-making power index Innovativeness Skill training Jointly make income use decision Spouse's education level

Number of years in school Marital status (1 = married) Women's intra-household decision-making power indexa If the women have adopted a technology outside agriculture (1 = yes) Adult skills training ≥3 months Jointly make income use decision with spouse Number of years in school of spouse

Religion Muslim Protestant Traditional

Muslim faith (1 = yes) Non-traditional churches (1 = yes) Traditional faith (1 = yes)

−0.084 (0.051) 0.011 (0.003) −0.471 (0.195)

−0.069⁎⁎ 0.086 −0.072⁎⁎

Number of formal connections both in the village & outside Number of friends and relatives within the village and outside the village they share with agricultural information Constant

0.078 (0.009) −0.137 (0.154)

0.061 −0.096⁎⁎⁎

Social capital Formal network Informal network Constant 2

Coefficients

0.142 (0.006) 0.209 (0.163) 0.821 (0.217) 0.157 (0.078) 0.056 (0.044) 0.242 (0.073) 0.209 (0.099)

Marginal effects

0.119 0.174+⁎ 0.251⁎⁎⁎ 0.144 0.052⁎⁎ 0.178 0.183⁎⁎⁎

−0.634 (0.722)

Pseudo-R : 0.42 sample size: 793. Robust standard errors in parentheses. a Wife's decision-making power index is constructed based on the 5 key agricultural technology decisions, agricultural produce marketing, and household expenditures. A value of 5 is given if a wife has fully autonomous decision-making; 3 participates jointly (50:50); 1 only consulted for an opinion and 0 no influence at all. ⁎ Significant at the 10% level. ⁎⁎ Significant at the 5% level. ⁎⁎⁎ Significant at the 1% level.

246

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

Table 4 Regression results of the mixed multinomial-logit model, determinants of choice of a CA strategy in Malawi. Variables

Improved maize varieties (variety diversity)

Improved maize varieties & storage

Improved maize and soil & water conservation

All technologies

Head's age Head's years of schooling HH received credit HH experienced any production shocks in the past 3 years Gender of plot manager Farm size Crop diversity index Asset index Distance to input market Distance to output market Membership in association Access to information Local agro-ecology-low potential Constant Wald-test (χ2)

0.001 (0.02) −0.008 (0.04) 1.213 (0.925)⁎⁎⁎ −0.045 (0.167)⁎⁎ −0.037 (0.021)⁎⁎⁎ −0.041 (0.116) 0.050 (0.029) 0.083 (0.024) −0.059 (0.012) −0.201 (0.005)⁎⁎ 0.013 (0.529) 0.433 (0.270)⁎ 0.049 (0.025)⁎

0.04 (0.009) 0.025 (0.037) 1.081 (0.772)⁎⁎⁎ 0.031 (0.054)⁎ −0.055 (0.061)⁎ −0.531 (0.049) −0.017 (0.009) 0.032 (0.006) −0.036 (0.020) −0.033 (0.017)⁎ 0.450 (0.456) 0.422 (0.251)⁎ 0.011 (0.034) 2.431 (3.988) 14.2⁎

0.15 (0.023) 0.005 (0.039) 1.121 (1.233)⁎⁎ 0.091 (0.006)⁎⁎ −0.020 (0.007)⁎⁎ −0.028 (0.032)⁎⁎

−0.02 (0.05) 0.055 (0.035)⁎ 1.392 (0.987)⁎⁎⁎ 0.141 (0.072)⁎⁎ −0.003 (0.015)⁎ −0.334 (0.011)⁎ −0.187 (0.207)⁎⁎⁎ 0.091 (0.023)⁎

−2.526 (4.165) 20.26⁎⁎

0.026 (0.018) −0.056 (0.042) −0.181 (0.021) −0.096 (0.053)⁎⁎ 1.253 (0.451)⁎⁎⁎ 0.648 (0.279)⁎ 0.839 (0.367)⁎⁎⁎ −4.082 (3.450)⁎ 17.68⁎⁎

−0.003 (0.014) 0.116 (0.019)⁎⁎⁎ 0.903 (0.422)⁎⁎ 0.406 (0.237)⁎ 0.078 (0.031) −3.955 (3.381)⁎ 22.77⁎⁎⁎

Note: The base category is farm households that did not adopt any agricultural innovation. Pseudo-R2: 0.31 sample size: 891. Robust standard errors in parentheses. ⁎ Significant at the 10% level. ⁎⁎ Significant at the 5% level. ⁎⁎⁎ Significant at the 1% level.

(IMSWC)”, “Improved storage & improved maize (SIM)”, “improved maize only (IM)”, and “all technologies” form the treatment categories. Table 4 results show that membership in association and access to information increased the adoption propensity of all the different agricultural innovations. Membership in an association seems to be panacea to access to information, inputs and innovations. The results highlight the importance of social capital in agricultural technology diffusion and adoption of improved agricultural innovations (Claridge 2007; Woolcock and Sweetser 2007). Social capital may be the only resource available for poor communities that are often characterized with low incomes, poor education, and few physical and financial assets (Woolcock and Sweetser 2007). Distance to output market (proxy for output market access) had negative and significant impact on the adoption of all the different agricultural innovations. These results confirm findings in Zambia asserting that poor functioning output markets may serve as a disincentive for market oriented production for rural households in subSaharan Africa (Rusike et al. 1997; Mwape 2004). Seshamani et al. (2002) identified poor transport infrastructure and weak regional market integration as major constraints to agricultural productivity in northern Zambia. Weak market integration implies that smallholder farmers cannot respond timely to price signals to make optimal production decision and investments. Markets are centers of exchanging information with supplier of inputs, buyers of commodities and other farmers (Maddison 2006). Agro-ecological setting also explained the heterogeneity in farmers' choice of agricultural innovations. For example, farming in the low potential drier agro-ecological regions such as Balaka significantly increased the propensity to adopt improved maize variety only, and improved maize variety with soil and water conservation. 3.4. Impact of CA strategies on farm output Choice of a particular agricultural technology has direct effect on maize productivity. Table 5 presents the impact of adopting the different agricultural technologies on maize yield per hectare. Overall, we found that adoption of the different agricultural innovations significantly increases maize productivity per hectare holding other things constant. Results show that adoption of improved maize only resulted in 14% increase in maize productivity per hectare. Farmers that had adopted improved storage and improved maize had the highest (29%) maize productivity per hectare. Implementation of all the technologies (improved maize, soil & water conservation and improved storage) leads to 10% increase in maize productivity. These results suggest that

adoption of more than three technologies does not lead to increased maize productivity per hectare. These results may also indicate the important trade-offs and cost of adopting a number of strategies in this agro-ecological setting of Malawi. Other exogenous factors that affected maize output per unit area were crop diversification, household labor supply, farm size, and whether the farm was managed by a male or female. Farmers that were further from the market also had significantly low farm output.

4. Conclusions Access to output markets and information on innovations play a significant role in determining agricultural innovation strategy to implement. This has considerable implications on how extension messages concerning innovations should be tailored so that every smallholder farmer should have access to information. The statistical significance of access to information across all technologies also indicates that farmers require basic knowledge of how the technologies work. Farmer groups make access to productive inputs easy for members and that has significant implications on farm output. Using participatory approaches, scientists need to work with farmers to develop and target agricultural technologies that meet farmers' multiple objectives. The positive and significant influence of spouse's education and women's intra-household decision-making power index on women's participation in resource allocation and agricultural innovation choice further suggest that it is important for researchers and policymakers to understand the social and institutional context in which the technology is targeted, to ensure that the technology will benefit both men and women farmers equally. The negative effect of informal networks, Muslim and traditional faith on women's participation in agricultural technology choice decision n imply that food security policies should go beyond food production measures and address the constraints to women's participation in technology development and dissemination. Overall, adoption of improved maize and storage technologies resulted in significant increase in maize output per unit area. Adopting improved maize only had positive impacts as well. Adopting all technologies was not necessarily the best option since it did not lead to the highest proportional increase in maize output per hectare. These findings have important implication to policy makers, researchers and development practitioners involved in the fight against hunger in developing countries. Policies can play an important role in helping farm households identify and adopt the best bet technology portfolio that can improve their agricultural productivity sustainably within their agro-ecological settings.

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

247

Table 5 The impact of agricultural innovations on maize yields. IMSWC Variables

Coef.

SIM Std. Err.

Coef.

IM Std. Err.

Coef.

Std. Err.

All technologies

ln (Yield)

Coef.

Coef.

Std. Err.

0.164 0.232 0.146 0.103

0.087 0.106 0.090 0.141

⁎⁎⁎ ⁎⁎⁎ ⁎ ⁎

−0.011 0.284 0.523 0.139 −0.275

0.103 0.093 0.169 0.025 0.074

⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎

−0.002 −0.020

0.001 0.116

⁎⁎⁎

5.760

0.184

⁎⁎⁎

Std. Err.

Outcome model Treatment effects IMSWC SIM IM All tech. Household/farm resource Income index Adult equiv. units Crop diversity index Farm size (A0) Female manager Institutional factors (zm) Dist. to the market Access to credit Selection model Household characteristics (zh) Education of head Household shocks Asset index Sex of head Information access Club membership (zr) Regression constants Constant lnsigma Lambda Observations (n)

−0.060 0.047 −0.547 −0.040

0.039 0.377 0.393 0.329

−1.045

0.450

⁎⁎

4.393 −0.159 0.312 125

3.915 0.071 0.068

⁎⁎ ⁎⁎⁎

−0.058 0.418 −0.243 0.239

0.032 0.393 0.339 0.329

0.421

⁎

−0.081 −0.022 −0.021 0.532

0.036 0.345 0.054 0.352

⁎⁎

0.293

−0.921

0.400

⁎⁎

5.840

3.854

−0.923

0.675

−0.180 125

0.046

0.117 323

0.052

−0.043 −0.440 −0.633 0.599

0.029 0.282 0.390 0.290

0.303

−0.369

0.839

3.384

−0.048 216

0.028

⁎

⁎⁎

⁎⁎⁎

⁎⁎ 891

Robust standard errors in parentheses. ⁎ p b 0.10. ⁎⁎ p b 0.05 ⁎⁎⁎ p b 0.01.

Acknowledgments We acknowledge the Australian Center for International Agricultural Research (ACIAR) for funding project activities under the CIMMYT-led Sustainable Intensification of Maize-Legume Cropping Systems in Eastern and Southern Africa project. We also would like to thank Malawian farmers, enumerators and technical staff at Chitedze research station for their supportive collaboration.

References Bardisi, E., Backden, C.M., Guzman, J.c., 2007. Gender, Entrepreneurship and competitiveness in Africa. World Economic Forum, the Africa Competitiveness. Becker, G.S., 1965. A theory of the allocation of time. In: Becker, G.S. (Ed.), The Economic Approach to Human Behavior, pp. 89–114. Bezemer, D., Balcombe, K., Davis, J., Fraser, I., 2005. Livelihoods and farm efficiency in rural Georgia. Applied Econ. 37 (15), 1737–1745. Bhat, C.R., 2001. Quasi-random maximum simulated likelihood estimation of the mixed multinomial logit model. Transp. Res. B 35 (7), 677–693. Blake, M., Hanson, S., 2005. Rethinking innovation: context and gender. Environ. Plan. 35, 681–701. Claridge, T. 2007. Social capital. URL: www.gnudung.com Accessed 20/11/2013 Deb, P., Seck, P.A., 2009. Internal migration, selection bias and human development: evidence from Indonesia and Mexico. Human Development Report 2009: Overcoming Barriers: Human Mobility and Development, United Nations Development Programme. Deb P and P. K. Trivedi. 2006. Maximum simulated likelihood estimation of a negative binomial regression model with multinomial endogenous treatment. Stata J. 6 (2) : pp. 246–255. de Janvry, A., Sadoulet, E., 2002. World Poverty and the Role of Agricultural Technology: Direct and Indirect Effects. J. Dev. Studies 38 (4), 1–26. DFID, February 2007. Gender Equality Action Plan 2007-2009 Making faster progress to gender equality Briefing. A DFID Practice Paper. DFID, London. Diao, X., Hazell, P., Thurlow, J., 2010. The role of agriculture in African development. World Dev. 38 (10), 1375–1383. Doss, C.R., Morris, M.L., 2001. How does gender affect the adoption of agricultural innovations? The case of improved maize technology in Ghana. Agric. Econ. 25, 21–39.

Ellis, F., 2000. Rural Livelihoods and Diversity in Developing Countries. Oxford University Press, Oxford. FAO, 1996. The State of Food and Agriculture Rome Italy. FAO, 2001. The economics of soil productivity in Africa. Soils Bulletin Rome, Itally. FAO, 2009. Food and Agricultural Organization of the United Nations. World Summit on Food Security, Rome, Italy November 16-18. FAO, 2011. Food and Agricultural Organization of the United Nations. The State of Food and Agriculture: Women in Agriculture- Closing the Gender Gap for Development. FAO, Rome. Green, H.W., 2003. Econometric Analysis. 5th edition. Pearson Education, Inc., Upper Saddle River, New Jersey, USA. Hall, A., 2010. Who champions putting research into use? RIU Emerging lessons Paper. Research Into Use (RIU) Publication (April 2010). Joshi, P.K., Gulati, A.A., Birtha, P.S., Twari, L., 2003. Agriculture diversification in South Asia: Pattern, determinants and policy implications. Discussion Paper No. 57. Market structure studies division. International Food Policy Research Institute, Washington, DC. Kakooza, J., Kabasimba, E., Ssemakula, B., Musisi, A., 2005. Gender variation in agricultural technology: a comparative analysis of two ecological zones of Uganda. A Report for ASARECA and ECAPAPA (August 2005). Kassie, M., Zikhali, P., Manjur, K., Edward, E., 2009. Adoption of sustainable agriculture practices: evidence from a semi-arid region of Ethiopia. Nat. Res. Forum 39, 189–198. Kristjanson, P., Neufeldt, H., Gassner, A., Mango, K., Kyazze, F.B., Desta, S., Sayula, G., Thiede, B., Forch, W., Thornton, P.K., Coe, R., 2012. Are food insecure smallholder households making changes in their farming practices? Evidence from east Africa. Food Secur. 4 (3), 381–397. Läpple, D., Kelley, H., 2013. Understanding the uptake of organic farming: accounting for heterogeneities among Irish farmers. Ecol. Econ. 88, 11–19. Lunduka, R., Fisher, M., Snapp, S., 2012. Could farmer interest in diversity of seed attributes explain adoption plateaus for modern varieties in Malawi? Food Policy 37, 504–510. Maddison, D., 2006. The perception of and adaptation to climate change in Africa. CEEPA Discussion Paper No. Centre for Environmental Economics and Policy in Africa.Unversity of South Africa, Pretoria, South Africa Mason, A.D., King, E.M., 2001. Engendering development through gender equality in rights, resources, and voice. A World Bank policy research report. World Bank, Washington DC http://documents.worldbank.org/curated/en/2001/01/891686/ engendering-development-through-gender-equality-rights-resources-voice [online] URL: (Accessed 12/11/2013). Mayoux, l., 2001. Tackling the down side: social capital, women's empowerment and micro-finance in Cameroon. Dev. Chang. 32 (3), 435–464.

248

M. Mutenje et al. / Technological Forecasting & Social Change 103 (2016) 240–248

Meinzen-Dick, R., Behrman, J., Menon, P., Quisumbing, A., 2011. Gender: a key dimension linking agricultural programs to improved nutrition and health. 2020 Conference Brief 9. International Food Policy Research Institute, Washington, DC. Minot, N., Epprecht, M., Anh, T.T.T., Trung, L.Q., 2006. Income diversification and poverty in the Northern Upland of Vietnam. Research Report No. 145. International Food Policy Research Institute, Washington, DC. Mwape, F., 2004. Institutional analysis of fertilizer in Zambia. Background report for Zambia PSIA. Prepared for Social Development Department. The World Bank, Washington, D.C (mimeo). Nompumelelo, P., Gumede, U., Oketch, R., 2009. Women and technology in South Africa: a case of four communities in KwaZulu-Natal". Gend. Technol. Dev. 13 (1), 103–125. NSO, 2008. National Statistical Office of Malawi. Population and Housing Census Report. NSO, Malawi. Okoboi, G., 2010. Improved inputs use, productivity and commercialisation in Uganda maize production. Contributed Paper Presented at the Joint 3rd African Association of Agricultural Economists (AAAE) and 48th Agricultural Economists Association of South Africa (AEASA) Conference, Cape Town, South Africa, September 19–23, 2010. Pender, J., Kerr, J., 1998. Determinants of Farmers' Indigenous Soil and Water Conservation Investments in Semi-Arid India. Agricl Econ. 19, 113–125. Quisumbing, A., McClafferty, B., 2006. Food Security in Practice. Using Gender Research in Development. IFPRI, Washington, D.C. Rusike, J., Reardon, T., Howard, J., Kelly, V., 1997. Developing cereal-based demand for fertilizer among smallholders in southern Africa: lessons learned and implications for other Africa regions. Policy Synthesis #30 Prepared for USAID Bureau for Africa Office of Sustainable Development. Michigan State University, East Lansing. Sadoulet, E., de Janvry, A., 1995. Quantitative Development Policy Analysis. The John Hopkins University Press, London. Schuller, T., Baron, S., Field, J., 2000. Social capital: a review and critique. In: Baron, S., Field, J., Schuller, T. (Eds.), Social Capital: Critical Perspectives. Oxford University Press, Oxford, pp. 1–38. Seshamani, V., H. White, N. Chisupa, and J. Leavy (2002) Northern Province Field Report. Draft Report Prepared for DFID. Seewww.ids.UK. mimeo Spielman, D.J., 2005. Innovation systems perspectives on developing-country agriculture: a critical review. ISNAR Discussion Paper 2. International Food Policy Research institute (IFPR), Washington, D.C. Sserunkuma, D., 2005. The adoption and impact of improved maize and land management technologies in Uganda. J. Agric. Dev. Econ. 2, 67–84. Tanumihardjo, S.A., Anderson, C., Kauter-Horwitz, M., Bode, L., Emenaker, N.J., Hagg, A.M., Satia, J.A., Silver, H.J., Stadler, D.D., 2007. Poverty, obesity, and malnutrition: an international perpective recognising the paradox. J. Am. Diet. Assoc. 107 (11), 1966–1972. Thomas, D., Frankenberg, E., 2002. Health, nutrition and prosperity: a microeconomic perspective. Bull. World Health Organ. 80 (2), 106–113. Train, K.E., 2003. Discrete Choice Methods with Simulation. Cambridge University Press, Cambridge. United Nations Development Programme, 2012. Annual Report 2011/2012: The Sustainable Future We Want. United Nations, New York. WHO/FAO, 2003. Diet nutrition and the prevention of chronic diseases. Report of a joint WHO/FAO expert consumption. WHO Technical Report 916. World Health Organisation and Food and Agriculture Organisation of United Nations, Geneva and Rome. Woolcock, M., Sweetser, A.T., 2007. Social capital: the bonds that connect. Asian Development Bank (ADB), Manila, Philippines. www.adb.org [online] URL: (Accessed 18/11/ 2013). World Bank, 2005. World Development Report: A Better Investment Climate for Everyone. Worldbank, Washington, D.C. World Bank, 2006. World development report: Equity and Development. World Bank, Washington, D.C.

World Bank, 2008. World Development Report: Agriculture for Development. Worldbank, Washington, D.C. Munyaradzi Mutenje is an agriculture economist, working with CIMMYT based in Harare, Zimbabwe. She received her Ph.D from the University of Kwazulu-Natal South Africa in 2011 and joined CIMMYT thereafter. Her professional and research interest focuses on food security, poverty and livelihood analyses, impact assessments and sustainable development. She is involved in four projects on sustainable intensification in southern Africa. She possesses vast experience as an extension officer, monitoring and evaluation specialist, lecturer and researcher. She has authored and co-authored 10 peer-reviewed publications. Henry Kankwamba is a lecturer at Bunda College of Agriculture. He has experience spanning 5 years in agricultural development, especially with smallholder communities. His research and consultancy work have included policy and impact analysis of agricultural projects, livelihood studies and micro-projects, and socio-economic and demographic analyses of selected communities in Malawi. Julius Mangisoni was awarded a Ph.D in Agricultural and Applied Economics, specializing in natural resources and environmental economics, production economics and prices and marketing by the University of Minnesota, USA in 1999. He is currently a professor of the Agricultural and Applied Economics in the Department of Agricultural & Applied Economics at Bunda College of Agriculture. He was a guest/visiting lecturer at the University of Pretoria in South Africa, where he taught environmental economics to MSc and Ph.D students in 2002 and in 2003 and agricultural price analysis from 2006 to 2012. From 1990 to 1993, he worked as a project preparation/marketing economist in the Planning Division of the Ministry of Agriculture in Lilongwe. Professor Mangisoni has been involved in a number of professional assignments, notable among them are: (i) measuring the poverty and food security impacts of improved maize in Africa: a combined econometric and microeconomy wide modeling approach (CIMMYT); (ii) evaluation of community based land redistribution program(World Bank); (iii) evaluation of USAID/OFDA small scale irrigation programs on rural poverty and food security in Zimbabwe and Zambia; (iii) impact assessment of the treadle pump irrigation technology on rural poverty and food security; (iv) water poverty package of the Limpopo river basin focal project; (v) evaluation of irrigation technologies in Chingale ADP; (vi) input voucher study in Malawi, Mozambique and Zambia; (vii) beneficiary assessment of the World Bank-supported community subproject of MASAF (vii) integration of environmental concerns into sustainable forest management; and (viii) review of the Canadian Physicians for Aid and Relief (CPAR)-supported drought mitigation project in Malawi. Menale Kassie is a development economist working at the International Maize and Wheat Improvement Center (CIMMYT). Since he joined CIMMYT in 2010, he has been coordinating the socioeconomic components of the program “Sustainable intensification of maizelegume cropping systems for food security in eastern and southern Africa (SIMLESA)” supported by the Australian government through the Australian Center for International Agricultural Research (ACIAR). He is also currently a project leader of the Adoption Pathways Project funded by the Australian government through the Australian International Food Security Research Center (AIFSRC). This project focuses on establishing panel data in five African countries in order to understand the drivers of adoption of technologies and their impacts from a dynamic perspective. Menale's research focuses on adoption and impact of crop and natural resource management technologies on rural household welfare, using advanced cross-section-panel econometrics and mathematical programming models. He has analyzed the contribution of sustainable land management technologies on agricultural productivity and the production risks of technologies on crops such as maize, wheat, groundnuts and pigeon pea and their effects on poverty and food security.