Assessing the Vulnerability of the Social-Ecological system of a Smallholder Farmer

(1)

Assessing the Vulnerability of the

Social-Ecological system

of a Smallholder Farmer

Bos, D., Derksen, D., Galen, E. van & Oudshoorn, C.

University of Amsterdam

Course: Interdisciplinary Project

Students: Diederik Bos Human Geography 10791698

Daniel Derksen Earth Sciences 10784373

Ellen van Galen Earth Sciences 10642455

Carolijn Oudshoorn Human Geography 10777962

Expert: Prof. Dr. Marc Davidson

Tutor: Koen van der Gaast

Date: 23 december 2016

(2)

Abstract

In this study a vulnerability assessment of the coupled social-ecological system of smallholder farmers is presented. This is achieved by assigning five main functions and relating proxy-indicators that describe how the system would function in a sustainable way for smallholder farmers and their environment. These functions include a self-regulating ecological system and a buffering capacity, the availability of a market (time, demand, infrastructure), the right crop conditions, sufficient entitlements, and a good social status. The indicators related to the functions provide both quantitative as qualitative data that were partly converted to vulnerability by establishing critical limits. By inserting these into an equation by taking into account the importance of the indicator and the range in which an indicator cannot be considered as vulnerable, information is provided about the vulnerability of a smallholder farmer system.

(3)

1.Introduction

Smallholder farmers make up a significant part of the developing world, with around 50% of the rural population (Morton, 2007). Most smallholder farmers are relatively poor. Therefore they are often considered vulnerable (Alwang, Siegel & Jorgensen, 2001; Luers, Lobell, Sklar, Addams & Matson, 2003). Through aid and other policies governments aim to decrease vulnerability, as well as preventing people from becoming vulnerable in the future (Moret, 2014; Smit & Wandel, 2006). Vulnerability is therefore an important focus of contemporary politics. Despite this importance, there is no clear definition of vulnerability or consensus on vulnerability in contemporary academic literature to date. Contrary, there are multiple definitions of vulnerability in academic literature, depending on different disciplines. Even within a certain academic discipline there are multiple definitions adopted. Little consensus on a single definition makes it a contested research topic. This results in diverse vulnerability assessments.

Our main question of this research is therefore: how can the vulnerability of the socio-agricultural system of smallholder farmers be assessed? But what is vulnerability exactly and when is someone considered vulnerable? By developing a vulnerability assessment of smallholder farmers we expect to find quantified answers to these questions, while taking qualitative aspects of vulnerability into account. Developing such an assessment can be done by integrating multiple definitions of vulnerability into a new holistic definition of vulnerability. Therefore, we start by presenting a number of different definitions of vulnerability that are frequently used in economics, ecology, earth sciences, and human geography. Redefining these different definitions into a single interdisciplinary and workable definition is the starting point of developing a holistic vulnerability assessment for smallholder farmers. A further important aspect of developing this vulnerability assessment is to clearly define the socio-agricultural system of smallholder farmers. This is a difficult task, since no smallholder farmer is the same. Therefore, this definition is highly subjective. However, we try to give a profound and thought-out definition by using a set of functions and indicators that are comprised in the socio-agricultural system of smallholder farmers.

We will continue our paper with a theoretical framework, which includes the different definitions of vulnerability as adopted in the different academic disciplines. In part 3 we will explain our methods as well as the integrative technique that we applied to form a single holistic definition of vulnerability, which will be explained in part 4. In part 5 we will define the smallholder farmer system. In part 6 and 7 we will explain its functions and associated proxy-indicators respectively. In the next part we will introduce the term safe zones, which builds upon the foregoing part of functions and indicators. In part 9 we will present and explain a mathematical equation of the degree of vulnerability, which we developed and is based on different weights of the indicators and

(5)

conclusion in which we answer our main question, followed by a discussion based on our assessment, presenting limitations as well as opportunities for further research.

2. Methods

Finding an interdisciplinary overarching definition of the concept vulnerability can be challenging because of the disciplinary jargon and cultural differences among multiple disciplines. This applies particularly for the social and natural sciences. In addition, the common used definition of vulnerability ‘has emerged from research on risk, hazards and climate impacts’ (Turner et al. 2003), which are external inputs changing the vulnerability of a system. However in social sciences, internal vulnerability also plays an important role. In this case, the internal vulnerability would depend on different socio-economic or cultural characteristics of smallholder farmers in a specific area that occurs within the social system itself (e.g. class, ethnicity). According to Turner et al. (2003), social units could experience and response to exposure differently, however 'vulnerability assessments should follow a common general methodological framework'. In order to follow this framework, we decided to combine these different definitions to achieve a common understanding of the concept that applies for smallholder farmers. This research is preliminary based on literature studies in order to acquire information about the concept vulnerability and its definition that applies differently for each discipline, which is applied to the socio-agricultural system of smallholder farmers.

This system will be defined into more depth as well as the optimal conditions of the system that are necessary in order to maintain a sustainable position. Based on this, five major functions that are necessary to make the system function properly were compiled. These functions are represented by certain measurable factors, called indicators, the indicators can provide both quantitative as qualitative data. Here also applies that not all indicators of a function were included in this study, which indicates that we are limited to approximately three to four indicators for each function. To convert these indicators to vulnerability of the system critical limits should be established. As soon as a critical limit is exceeded, one can speak about a disruption of the corresponding function, which may lead to a more vulnerable system. However, the indicators that we developed can have different types of values, including dichotomous values and percentages. In order to include all indicators so called ‘safe zones’ were compiled. Safe zones help bridge the gap between the indicator values and vulnerability, based on our overarching definition of a vulnerable system. The deviation from the safe zone, valued between 0 and 1, will be inserted into an equation together with the weight of each indicator, since some indicators may be considered as more important because of their influence on the overall system. However, some indicators cannot be easily captured into an equation due to their non-measurable content, which especially applies for

(6)

the social indicators (e.g. emotions and individual concerns). In order to obtain a reliable view of the overall interdisciplinary system they should also be taken into account, which can be obtained from interviews.

To study such an interdisciplinary system, a number of integrating techniques can be used, depending on the aim of your research. It is important to look for common ground per definition and hereby integrating the different subjects into one system (Somerville & Rapport, 2000). For our paper we give a unilateral definition for vulnerability based on the different definitions adopted in the four academic disciplines, which is called redefining (Menken & Keestra, 2016). Redefining is comparing the same term (vulnerability) from multiple disciplines and bringing out a commonality in this term that transcend the disciplines. It hereby becomes applicable to the entire system as well as for the individual systems. Redefining differs from extending, which is broadening a term so that it will incorporate more disciplines, because we minimize or extract the essence of it. This same essence was present in each definition of vulnerability in each discipline.

Our main purpose in this paper is proposing a certain approach of assessing vulnerability in the interdisciplinary smallholder farmer system. This approach is derived from research in earth and soil sciences for standardising research on soil quality. The approach comprises of top to bottom steps from a clear view and definition of the system to determining the functions the system supplies and finding the indicators that represent these functions, a more elaborate explanation can be found above in the vulnerability definition within earth sciences. The integration technique of using this systematic approach used in earth sciences and applying it to the interdisciplinary system is called

organisation (Menken & Keestra, 2016). Organisation can be explained as

identifying commonalities in concepts, in our paper this would be for instance functions and indicators, and organise them in a way to lay causal links between them. We try to find the functions of our system and look for the links of the indicators that touch multiple disciplines. In determining the functions we use the previously mentioned extension technique (Somerville & Rapport, 2000). Redefining smaller and more specific functions to broader and more interdisciplinary functions. In some cases finding the essence of multiple functions show them to be common, this is again a form of redefining.

Eventually, we hope to create a vulnerability assessment that when used in practise will show how the vulnerability in a system is structured, how a certain change can cause more vulnerability in a specific area but perhaps diminish vulnerability of the whole system. This can later help to find pressure points in a system that will make it more resilient although it might mean a little decrease in one part.

(7)

3. Vulnerability according to different disciplines

Before determining an overarching definition of vulnerability for this research, it is important to look at the definitions of vulnerability within the four disciplines of sociology, economics, ecology, and earth sciences. In this chapter we consider the definitions of vulnerability and derive a single definition for each discipline.

Social Vulnerability

In social sciences the term vulnerability is used with different nuances. In this section a definition, which applies to this discipline, is pursued, by examining existing nuances about the concept of vulnerability within the social sciences field.

Firstly, Lange, Rogers & Dodds (2013) argue, in the field of research ethics, that vulnerable is being more prone to harm or exploitation. We will argue that this might also be applicable to social science. Whole social groups could be seen as vulnerable, for instance particular minorities. This is demonstrated by the fact that minorities are mentioned in the equal protection clause of the American Constitution, which purpose is the protection of minorities from discrimination (Galloway Jr. 1989). In this way, minorities are represented as being a victim, dependent or disadvantaged, what Fineman (2008) also argues about the state of being vulnerable. Besides, the manner appears negative and despondent.

Therefore, Fineman (2008) deformed this traditional definition of social vulnerability. She describes it as a concept that is engaged with “privilege and favor [sic] conferred on limited segments of the population by the state and broader society through their institutions” (p.1). Here, focussing on the social framework and equality of people instead, she appoints that being vulnerable is an inescapable social condition that is created by the state and social institutions. So not only are particular social groups more vulnerable, for instance because of their minority position or having less economical potential, but this condition is also constructed by social institutions, society or the state. Goodin (1986) supports this idea, that some vulnerabilities exist naturally and some are shaped by social arrangement (1986, p. xi). So vulnerability within social sciences explicitly encompasses the aspect of focussing on social institutions (Moret, 2014) and their contribution to the meaning of ‘vulnerable’. Furthermore, it is entrenched in power relations, development of technology and knowledge and distribution of resources (Blaikie et al., 1994, Downing et al., 1996, Kelly and Adger, 2000 and Liverman, 2001, as cited in Eakin, 2005).

Adding to the statement by Fineman (2008) that being vulnerable could be an inescapable social condition, the concept of ‘the centrality of the social’ is worth recognizing. This concept entails that the practice of agriculture is always grounded in social relations in and between households, for example gender relations (Morton, 2007). It is therefore deeply influencing farming decisions,

(8)

product marketing and knowledge management (Morton, 2007). If generating negative impact, this could be seen as an aspect of internal social vulnerability of a smallholder farming businesses.

Hence, following the social sciences literature, the definition of social vulnerability that will be used for further analysis, is the proneness to harm, due to the status of the social group, which is caused by by social institutions.

Economic Vulnerability

Economic literature encompasses multiple definitions of vulnerability. A general definition of vulnerability is the risk that a system could be negatively affected by a specific unforeseen perturbation (Alwang et al., 2001; Naudé, Santos-Paulino & McGillivray, 2009). There are two main aspects of vulnerability, which are the exposure to risk, and the susceptibility of this risk (Béné, 2009). Susceptibility is the ability of the system to recover from risk factors (Béné, 2009).

By assessing vulnerability one can look at the risk that is being imposed on the system or on the outcome of this risk. In economic literature vulnerability is mostly assessed by the outcome of a certain risk. By the outcome is meant the inability of the system to react to risks, as well as the underlying conditions of the system (Alwang et al., 2001). Within economic vulnerability, quantifiable data, which are easily put in a single convenient metric and are mostly obtained by using different proxy-indicators, are of focal interest, since the collected data can be added and eventually form a simple function that expresses vulnerability (Alwang et al., 2001; Luers et al., 2003).

In poverty literature, which is strongly related to vulnerability, this single metric is often money, which is almost always easily added up, considering different outcomes of risk. Although poverty is related to vulnerability it is not correlated, since the aspect of risk is intertwined in the definition of vulnerability (Béné, 2009; Chambers, 1989). However, most literature on economic vulnerability focuses on poverty, since it is easily expressed in terms of money, as explained above. By quantifying vulnerability into a single metric, the complexity of vulnerability is nullified, since other important aspects or viewpoints are neglected (Alwang et al., 2001). This is a critique on economic vulnerability assessments. However, more recent literature on economic vulnerability takes this critique into account and work with a more broadly defined definition of vulnerability.

Let us consider two forms of often used vulnerability assessments within the economic literature. The first is more related to poverty literature, and focuses on assets and entitlements. The second form is an example of sustainable livelihoods and is related to well-being. In asset-based assessments, a system (such as a household) is considered less vulnerable since a potential risk is more easily managed through the assets or entitlements that a household

(9)

can command to in order to sustain his/her welfare level (Adger, 2006; Sen, 1997).

In a sustainable livelihood assessments, focus is on well-being of a person or a household and comprises all capabilities, assets and activities that make up well-being (Adger, 2006). According to this view, vulnerability is not just an outcome of poverty but plays an important role to the susceptibility of a household of not being able to sustain a livelihood (Adger, 2006; Morduch, 1994). Although this relates to asset based assessments it considers the system in a more holistic way. In sustainable livelihood assessments well-being depends on a large number of different indicators making up a certain level of well-being, which is being affected by (potential) risks. However, this level of well-being is not easily quantifiable, since there are no clear definitions on a minimumlevel of livelihood (Alwang et al., 2001). Therefore, in the rest of this paper we will work with the former definition of economic vulnerability, based on individual entitlements.

Ecological Vulnerability

Ecological vulnerability can be used for different hierarchical levels within ecosystems (De Lange et al. 2010). Concerning species, the emphasis is on the vulnerability of extinction whereas vulnerability with respect to ecosystems focus on the irreversible damage (Alwang et al. 2001).

According to the World Bank’s Environment Department (WBED) there are two dimensions of ecological vulnerability: exposure to a stressor and its effect, and the recovery potential which can be defined as resilience or adaptive capacity. Resilience refers to the ‘ability to absorb change and disturbance and still maintain the same relationships’ (Holling, 1973; Eaking & Luers, 2006). It is believed that a (ecological) system has a state which is in a stable condition, which is called an equilibrium. The system itself ensures that it remains in this state. However, a system can have multiple equilibria and by increasing disruptions or stressors the state can be in a transition, which is called a tipping point (Gunderson, 2000; Scheffer et al., 2001). During a perturbation, in order to stay in equilibrium the system will adapt, however both the stability and the resilience are changed. This is called the adaptive capacity of the system (Gunderson, 2000). In addition, social and ecological resilience are intertwined due to the dependency on ecosystems by a community. This dependency on ecosystem services (products and services provided by ecosystems that are essential to humans) ‘can be regarded as the link between functioning of ecosystems and their role for society’ (Daily et al. 1997; De Lange et al. 2010). In this way, vulnerability will be defined as the exposure of individuals to stress due to environmental change (Alwang et al. 2001; Ahmed and Lipton, 1999). Nonetheless the definition that we will use will be the susceptibility to stresses and shocks of an ecosystem that can vary depending on the hierarchical levels that are present.

(10)

Soil Vulnerability

Within soil sciences the term vulnerability is used moderately. Tao, Hayashi and Lin (2002) wrote a paper on soil change to acid deposition in China and in their case they defined vulnerability as ‘the extent to which acid deposition may damage or harm a system’. From this definition it becomes clear that it is important to clearly state what the system is before being able to look at the harm by inputs to the system.

Most soil research papers focus on the system namely the soil quality (good state of a system within its context). An article by Nortcliff (2002) described a systematic approach that will help standardise soil research to make the term quality become testable and comparable on different sites. This approach begins with a clear, objective definition a soil quality, the term ‘quality’ in this context means how a soil should ideally function independent of its composition and site. The definition that is used and used by most soil (quality) research is: “the continued capacity of soil to function as a vital living system, within ecosystem and land use boundaries, sustain biological productivity, to promote the quality of air and water environments, and to maintain plant, animal and human health”. The paper continues by giving five basic functions that can be derived from this definition. These functions can be combined and altered to represent the soil functions under specific conditions. To find if these functions are supplied by the soil indicators need to be determined and investigated. These indicators are not bound to a specific function and can sometimes be relevant for multiple functions. This way of dissecting a problem to go from general to specific helps to make research more comprehensive, however, Nortcliff his paper ends with the indicators and leaves a gap to vulnerability, the term is not mentioned.

Research by Arshad and Martin (2002) help bridge the gap to vulnerability by debating the importance of identifying critical limits for the indicators, levels on which the functions they represent will still be supplied by the soil. Including Nortcliff in this would mean the degree of change in the soil functions, reducing the quality. This would happen if the indicator levels (quantifiable) get near the critical limits, limits that would previously be assessed. This would therefore be a cumulative value of all indicators.

To sum up, in soil sciences the term vulnerability is not very common. A way to determine the vulnerability is by defining what a good soil is, determining its functions within its context, finding the right indicators for these functions. Followed by assessing critical limits for the indicators and research to find the change of the value under certain circumstances. If the values get to the critical limits the soil will be more vulnerable to change. The definition that we will work with from the soil sciences is a combination of Tao (2002) and Nortcliff (2002) and is formulated as the extent to which external inputs may change the soil system through its functions.

(11)

4. An overarching definition of vulnerability

The different vulnerability assessments that are developed out of the multiple definitions of vulnerability can be seen as complementary (Eakin & Luers, 2006). However, the strength of interdisciplinary vulnerability assessments is the integration of those different assessments and the collaboration between the different disciplines involved in vulnerability research (Eakin & Luers, 2006).

In the following figure, the four definitions of vulnerability from the different disciplines are combined by means of redefining, explained in section 2 (Methods). Thus, the workable definition of vulnerability, which is used for this research, is stated as; ‘the susceptibility of a system to change state as a consequence of stresses and shocks’.

Figure 1: Diagram of definitions of vulnerability for each discipline and the unilateral definition of vulnerability.

(12)

5. Description of smallholder farmer system

As mentioned before in the methods, the model that is being presented works town down from a clear definition of a system. The system can be any group of elements that are interacting to form a unified whole.

The system for which a vulnerability assessment is created in this research, is the one of a smallholder farmer. The boundaries of the system are given by the fact that it is focussed on the smallholder farmer and limited by the four disciplines. To be able to apply a vulnerability assessment to this system it is crucial to determine the condition we want the system to be in, because that will be the condition that the new data will be compared with when the vulnerability of smallholder farmers is being researched. In this chapter a comprehensive definition of a smallholder farmer is provided.

Smallholder farmers are farmers with an agricultural company that is cultivating about 2500 hectares, whilst a commercial farmer is cultivating around 5200 hectares (Pompi, 1996). On their land they are mainly using family labour and their sales serve as the main form of income (Cornish, 1998 as cited in Morton, 2007). At the same time, the term is also being used in a sense that it refers to rural poor farmers of developing countries in general (Morton, 2007). 50% of the rural population of developing countries is a smallholder farmer (Jazairy, Alamgir & Pannuccio, 1992, as cited in Morton, 2007). In the smallholder farmer system, a lot of issues are occurring. However it is notable that these are not standard in every smallholder farmer situation. For our smallholder farmer system being able to be assessed, we need to set an ideal situation against which change is being measured.

The ideal and, at the same time, the definition of a smallholder farmer system would be; smallholder farmers should have a favorable social status from where they can maintain an ecologically sustainable, self-regulating, economically healthy small farming business, of which they earn adequate money from to sustain family members and have a sufficient well-being.

(13)

6. Functions of the smallholder farmer system

From the above determined definition of the system characteristics or functions can be deducted, we call them functions because this makes the organization from the soil science model by Nortcliff (2002) unequivocal. The functions are provided by the system if this system works properly (deduction). For it to work the other way around, properly working functions to represent a system that works as defined (induction), it is important that the functions that were determined embrace the whole system. This makes them something like the axioms of the system. This may lead to the functions being basic and consequently abstract. This is not problematic it just shows that they are fundamental for the system. The greater the system, the more basic the functions can be because of this, our functions needed to be representative for all smallholder farming systems (independent of context and location). We also narrow the number of the functions to five to make the model more comprehensive. This limitation, however, made the functions even more basic.

The next step is to find a way to assess these ‘working’ functions. If all the functions are supplied the system is in its ideal state, in line with the definition. If they deviate from this ideal one could argue that therefore the system is more vulnerable. The deviation is measured in the values of quantifiable indicators that in their turn significantly represent the given functions. How the deviation is measured from the indicator levels is discussed in the paragraph on critical limits and ‘safe zones’.

For some of the functions of the smallholder farmer system in this paper the integration method of extending was used (Menken & Keestra, 2016). Within the functions contemporary views of the disciplines of economics, human geography, soil sciences and ecology are integrated. The functions made in an earlier stage of our research were adjusted and combined to form more integrated functions to make the vulnerability assessment for the interdisciplinary system more efficient. Note that all functions we consider here represent a subjective stance. As mentioned, they are deduced from the definition which itself was based on literature research. In other research a different definition might surface and thus find different functions. In the following section the functions of the smallholder farmer system are explained.

(14)

Figure 2: The functions of the smallholder farmer system. Enlarged version in Appendix 2.

Ecologically sustainable: self-regulating and buffering

The first function, ecologically sustainable: self-regulating, represents the ecological sustainability within the smallholder farmer system. This function is an example of the extending integration technique as the broadening of the term ecosystem to also incorporate the soil ecosystem. If the system is sustainable, and therefore will not increase in vulnerability over time, it is important that the ecosystem around the farmer, including the soil, conserves a certain quality/standard. This quality will guarantee its functioning to the farmer for short and long term. The ecosystem needs to be able to tolerate negative inputs, such as disease, acidity, climate changes and all the stresses that come as a result of farming practises. The ability to tolerate inputs is the buffering capacity and can be seen as the opposite of high susceptibility, which would be highly reactive responses from the system to inputs (Tao, Hayashi and Lin, 2002). Subsequently, a good quality ecosystem can filter and break down these inputs (for instance toxins), hereby countering possible accumulation (Nortcliff, 2002).

(15)

Availability of market outlet (time, demand, infrastructure)

The second function of smallholder farmers that we will consider is the presence of market outlets, as well as the farmers’ access to markets. This function is of importance, since it represents the smallholder farmers’ ability to sell their products and compete with other selling parties. In literature it is often stressed that the geographical location of a smallholder farmer could impact its business opportunities. This is due to the rural living environment, which could be remote or even isolated (Morton, 2007), causing less opportunities for an outlet. Much like geographical location, a problem that coheres to this is one of sufficient infrastructure, for instance to reach markets. In the function Availability market we tried to include these aspects. For a high quality system all the functions need to be adequate, right, etc. this functions is also read as the sufficient availability of markets.

Right crop conditions

The third function of the system is having the right crop conditions. This function is related to the natural sciences as it entails the right biological and natural conditions that serve an efficient growth of the crop that is being produced by the smallholder farmer. This function was also formed using extension for the same reason as the first function as in comprises both the conditions above the soil, earlier described as ecology, as well as the conditions in the soil. The two

natural functions could themselves be put together into one function (a sustainable and adequate farming ecosystem) but for the sake of overview and workability we chose to keep them separated.

It is an important function as without the word ‘right’ it can be opposed to the first function of a high quality ecosystem that can buffer and filter inputs. When this function is read as the ‘optimal’ crop conditions the farming system might move to damage the ecosystem and reduce its ability to deal with stresses in order to optimize yield. The word ‘right’ is chosen with the purpose of allowing human influence or use of the land without causing a conflict with the first function. There should not be conflict between functions as the system of the smallholder farmer, in a perfect state, will have all the functions working perfectly. This means that there could not be a contradiction between them. The right crop conditions differ for the specific crop or combination of crops that is/are being cultivated, but it includes soil specifications, such as nutrient levels, as well as (above ground) climate conditions.

(16)

Sufficient entitlements

Having sufficiënt entitlements is an important aspect of the smallholder farmer system, since it represents the combination of rights and opportunities that an individual can command to in order to sustain his/her welfare level (Adger, 2006; Sen, 1997). Expanding entitlements are therefore an important characteristic of economic development (Sen, 1997). A poor economical situation is considered one of the greatest constraints smallholder farmers encounter around the developing world. For instance due to the lack of savings or credit history, banks and micro-credit institutions bypass them, which makes scaling up difficult (Salami, Kamara, & Brixiova, 2010). According to a case study in Kenya, the affordable credit issues are seen as the leading factor behind low productivity of smallholder farmers (Salami, Kamara, & Brixiova, 2010). This proved us that having sufficient entitlements is an extremely important function for the success of a smallholder farmer system.

Good social status

The last function that will be considered is the good social status of the smallholder farmer and its family or community. In the section above it has already become clear that a sustainable smallholder farmer system should have sufficient overall entitlements. However, since sufficient entitlements is mostly an economical function, the social status of the smallholder farmer is not incorporated. Though, this social status could impact the entire system if it is allocated as bad. As briefly pointed out in the theoretical framework as ‘the centrality of the social’, social conditions and relations could highly impact farmers’ decision-making and knowledge management (Morton, 2007). Yet the difficulty with a function such as social status is the challenging existence of the concept. It is not to be captured in words, or quantitatively measurable indicators. A good social status is formed by the presence and pressure of different social characteristics, such as different ethnicities, sexes and classes. Just being a minority, female farmer might bring your system into a higher vulnerability than for instance being a white, male farmer. When conducting such an assessment in a specific region, one could collect region-specific data about discrimination, corruption or any type of advantaging of specific groups or sexes and bring more profoundness into the research. Hence, we highly emphasize that qualitative research should be included in the vulnerability assessment of a smallholder farmer, but due to the lack of region-specific data and the inability to operationalize some concepts of the social world into quantifiable ones, this function is therefore left out of the quantitative side of our research.

(17)

7.Indicators

To be able to measure state changes of the functions of the smallholder farmer system, the operationalization of the functions is necessary. In this research we argue that this is achieved most desirably on the basis of multiple associated indicators. The indicators used for operationalizing all functions are derived from academic literature and should best measure these functions in numbers, again according to our academic point of view. The indicators that are presented in the research are the once that were found so far and found to be significant representatives of the functions that were found. Nonetheless, a lot more research is needed to make it complete. Moreover, research is also required on how to determine which indicators are most important in the system. Below, the diagram presenting the functions of the smallholder system with their indicators is shown.

Figure 3: Interdisciplinary functions and indicators of the socio-ecological system of smallholder farmers. Enlarged version in Appendix 3.

Ecological sustainable: self-regulating and buffer capacity

This function includes the sustainability of the environmental system. This includes functioning of nutrient- and water cycle systems and disease

(18)

suppression. According to Bockstaller et al. (1997) several agro-ecological indicators are compiled based on the protection of so called objectives (table 1) in order to optimize the agro-environmental system. Since most of the objectives (soil and water quality) are also covered by the function of the right crop conditions, it has been decided to only focus on the biodiversity objective. 'Biodiversity is an indicator of the capacity of ecosystems to function effectively' (Burke, 2002). Because of specie diversity, species-rich ecosystems have a higher probability to contain species that thrive and survive during environmental perturbation in comparison to species-poor ecosystems (Mooney et al. 1996; Burke, 2002). This indicates that an ecosystem characterized with high biodiversity is more buffered against disturbances and can therefore function better in fluctuating environments in order to achieve and maintain a self-regulating system.

According to the Millennium Ecosystem Assessment (MEA, 2005), biodiversity is considered to be the basis of all ecosystem functions (Garbach, 2014) and is defined as the composition and diversity of species. Agricultural biodiversity can be seen as the foundation for a resilient farm ecosystem (Frison et al. -2011). Due to the complexity and the comprehensiveness of the concept it is assumed that not only one single indicator for biodiversity can be devised (Duelli, 2003). We therefore decided to designate multiple indicators in order to represent biodiversity as a whole while taking ecological and agricultural practices into account. This implies both habitat (ecological) and farming management (agricultural) indicators, which ultimately determine the biodiversity and the associated functions of the socio-agricultural system of smallholder farmers within an ecological perspective. Although many different indicators contribute to the two proposed ecological functions, it has been decided to select two of the main indicators of biodiversity in order to keep a clear overview. Therefore these indicators were included by the term biodiversity (figure 3).

The main indicator of the ecological practices is crop diversity. The way a farm is arranged (i.e. animal, crop or mixed farming) and what kind of agricultural system is applied (polycropping, monocropping) affects the crop diversity and therefore also the agricultural biodiversity. The most pivotal farming management indicator of biodiversity is the application of pesticides (herbicides, insecticides and fungicides). This indicator is a constraining factor for species diversity on farmland (Geiger et al. 2010; Guidebook, A. 2012) due to reduction of arable flowers and invertebrate species and disruptions of food webs (Hole et al. 2005; Guidebook, A. 2012). ‘This indicator can be measured on the amount applied, the rate of degradation, its partitioning to the air, surface water and groundwater and its toxicity to the species in the environmental compartments’ (van der Werf, 1996; Bockstaller et al. 1997). To include the sustainability in a soil microbial diversity and soil organic matter content (SOM) are also indicators for this function (Arshad and Martin, 2002; Schoenholtz et al.,

(19)

Table 1: List of agro-ecological indicators in association with their environmental and agronomic objectives. As presented in Bockstaller et al. (2007).

Availability markets

The second function of smallholder farmers that we will consider is the presence of market outlets, as well as the farmers’ access to markets. This function is of importance, since it represents the smallholder farmers’ ability to sell their products and compete with other selling parties. The indicators that best measure the presence of outlet markets and the access to markets are a combination of: the number of markets, which are accessible within a certain amount of time (number), for which the farming business remains cost effective and the farmer’s well-being is maintained or enhanced; the number of routes that are possible, which lead to these markets; and the total demand of the farmer’s products at these markets in dollars ($) (Hugo, Squalli & Wilson, 2006). When focusing on markets, we only take into account local markets, since most smallholder farmers are not directly involved in the world market. When considering world markets, other indicators such as trade barriers, technical barriers, import tariffs, inflation, etc. should also be accounted for (Hugo et al., 2006; “Measuring market access for agricultural products”, 2009).

(20)

Right crop conditions

As explained in the previous chapter on functions, the right conditions depend on the crop that is cultivated, likewise for the indicators to measure them. Nonetheless, there are some indicators that are always of importance for agriculture. Outside of the soil humidity, temperature and seasonal changes of these factors are most relevant. Within the soil, indicators for nutrient availability are organic matter (concentration), pH and the concentration of extractable nutrients (Nitrogen, Phosphorous and Potassium) (Arshad and Martin, 2002; Schoenholtz et al., 2000). For crop growth, importants indicators are; electrical conductivity (CEC), bulk density and texture.

Figure 4: Key soil indicators for soil quality assesment. As presented in Arshad and Martin (2002).

Sufficient entitlements

The indicators that best measure sufficient entitlements are a combination of: total income per household in dollars ($); household size, which is the total number of people that the smallholder farmer has to provide for; total worth of assets in dollars ($); and the totality of rights and opportunities that a

smallholder farmer can command, such as certain (local) policies, but also opportunities that arise out of social relations for example (Adger, 2006; Alwang et al., 2001; Chambers, 1989; Vincent, 1976). However, these rights and

opportunities are very diverse and are therefore difficult to express in

(21)

8. Safe zones (critical limits)

At this point we are left with a gap between the indicators and vulnerability. We have determined that the indicators that we use need to be quantifiable so we can place them into an equation (This will later be explained). This equation will return a value on a universally applicable scale for the vulnerability of a smallholder farmer system. Within the indicators that we use, research is needed to find critical limits, at what levels do the indicators represent a function that works and at what level does the specific indicator not contribute to the function or maybe even disrupt the function? Multiple studies argue that critical limits are important and link vulnerability to those limits (Arshad and Martin, 2002; Tao, 2002). However, the problem we encountered, concerned the fact that our model (could) include different types of indicators. Among these indicators are those that are dichotomous, as for instance the presence of certain laws, indicators that vary between two values (on a scale), this category includes pH and indicators with percentages, and lastly we have indicators that do not have a specific outer value, indicators that can be sorted in this category are for example the travel time to a local market and the amount of money a farmer owns.

To include all indicators in the same model we decided to use so called safe zones. We define safe zones as the value or range of values of an indicator on or between which the indicator does not contribute to vulnerability and therefore represents an adequately working function. This way we make it possible for indicators that are not dichotomous to also be able to have a value of 0. The deviation of the indicator value from the mean value of the safe zone for the specific indicator is seen as the contribution of the indicator to the vulnerability of the whole system. The deviation will be inserted in the equation as a value between 0 and 1, again this makes the combination possible of the different indicators into one scale, without losing the nuance that can exist for non-dichotomous indicators. This basically turns all deviations into percentages, with 1 being a 100 percent. Determining the safe zones and its limits as well as the 0 to 1 contribution to vulnerability asks for a lot of additional research on groups of indicators or by isolating a single indicator, so called system analysis (Bentley and Whitten, 2007). In the next paragraph more detail will be given on the practise of the research that will make the equation more accurate and including the deviations that arise in a system because of linkage between indicators or of indicator to different function at certain levels. A visualization of the safe zone concept can be found in appendix 4.

Our research does not yet include a sound statistical legitimation of our technique but this simple mathematical model makes an estimation possible.

(22)

9. Equation

The last step now is to insert all the indicators into one equation. As described in the previous paragraph a value between 0 and 1 will be assigned to all the indicators. The indicators are however not of equal importance, some can be more influential as they are important indicators for multiple functions or change in significance to the system as their value changes. As stressed before, we do not have the knowledge to cope with these mathematical and statistical problems yet, but by introducing a weight for every indicator we hope to show our intent to include relative importance, linkages between indicators, and time frames for which the vulnerability is tested. If we set the vulnerability on a scale from 0 to 100 an equal distribution of the weights would be 100 divided by the amount of indicators that are used. If all the indicators would give a value of 1 the vulnerability will consequently be 100. The scale from 0 to 100 is important as this allows the model to be used on different sites and subsequently be compared between multiple studies.

The weights that accompany the indicators need to be estimated. We propose that this could be done by creating a dummy variable of vulnerability, which is drafted from historical data. In order to estimate this dummy variable, indicators such as described above should be collected per farmer. By this, a regression model can be filled in, which eventually would estimate the weight of the indicators in the past, providing a model for the future. Naturally, higher mathematics and statistics need to be applied to these forms of equations, an aspect we could not fulfill from our disciplines.

Suggestion:

Vulnerability: 0 - 100

Vulnerability = (Deviation1 * weight1) + (Deviation2 * weight2) + … +

(deviationn *weightn)

(23)

10. Qualitative research

Both Moret (2014) and Eakin and Luers (2006) stated that the participation of potentially vulnerable groups, in this case the smallholder farmers, is essential when doing, and evaluating on, research about vulnerability. It will only lead to increasing levels of marginalization and disempowerment if the thoughts and concerns of these potentially vulnerable groups are not taken into account when doing research about their vulnerability (Eakin & Luers, 2006).

Therefore, an important aspect within this paper is the difference that is being made between quantitative and qualitative research. The mathematical function explained in the foregoing part is completely developed out of the quantifiable indicators presented in section 6. However, this does not mean that the qualitative side can be neglected within our research. As stipulated in different sections of this paper already, the qualitative part plays an important role throughout our research. This is because qualitative research takes emotions and individual concerns into account. These are social aspects that cannot be measured and for which there are no indicators. Qualitative research is especially important within the social sciences, which is intertwined in this interdisciplinary research. Since qualitative aspects, such as good social status, are not part of the equation presented in section 9, the outcome of the function, although correct as long as the quantified data is sufficient, should be scrutinized. Therefore, and in line with the article of Moret (2014), we want to stress that making use of a mixed methods approach is needed within future research endeavors.

So, in order to conduct a vulnerability assessment, researchers should not only collect and make use of quantitative data, but also examine qualitative data, which may support or detract the outcome of the quantitative findings, resulting in a more thorough and in depth vulnerability analysis/assessment. These qualitative data can be obtained by using multiple research methods, such as: interviews, surveys, focus groups, and ethnography.

(24)

11. Conclusion

Preceding vulnerability assessments are often oriented from natural sciences, which is probably related to the fact that the definition of vulnerability originates from research on risk, hazards and climate impacts. However, to properly describe vulnerability, it is important to take the coupled human-environmental system into account by conducting an interdisciplinary approach.

In our research we emphasized that in creating a vulnerability assessment it is extremely important to build the research from the definition of the system that is being researched. Since the smallholder farmer system is central in our overarching definition of vulnerability, we argue that one should use the system as a starting point when developing a vulnerability assessment. The following step should be deriving a number of functions that the system, described by the definition, should supply if the system is in the state that we want it to be in. Whether these functions are working correctly can be research by looking at indicators that represent the functions.

In order to argue if the vulnerability of smallholder farmers has changed or not, the deviation from the ideal functions of the system should be measured. We stress that this can be achieved by setting critical levels and hereby determining a safe zone per indicator. Then, putting the deviation per indicator in the equation, while taking into account the different weights of indicators, can provide a value between 0 and 100 of vulnerability. A lot still has to be done on the statistical and mathematical side of this research, including the right way to find the weight of each indicator to represent the indicator its relevance in the system.

In addition, we highly emphasize that to provide a full understanding of the vulnerability of the system, qualitative aspects that cannot be captured in the equation should be included in the study by making use of other research methods that are suitable for these indicators, including interviews. The outcome of the quantitative and qualitative values combined offers a far more complete view the degree of vulnerability for the whole system.

Nonetheless, some difficulties emerged while doing the research. Frequently, these could be ascribed to the complexity of interdisciplinary research on vulnerability of smallholder farmers. To fully understand this vulnerability assessment, one should be aware of the fact that this assessment forms a framework for further vulnerability research. This entails that every aspect of the research is context dependent. So the mathematical equation explained in section 9, as well as the number of indicators, and the safe zones, or critical limits do not apply to every single case or location. In further research, these might be narrowed down or set differently. This especially applies to the weights of the different indicators that have been incorporated into the equation presented

(25)

interrelated. Moreover, some critical limits are difficult to determine and comprise a margin, which may in fact be too broad. However, it is follow-up research that needs to deal with these limitations, which will differ per indicator and are, again, context dependent.

(26)

12. References

- Adger, W. N. (2006). Vulnerability. Global environmental change, 16(3), 268-281.

- Altieri, M. A. (1999). The ecological role of biodiversity in agroecosystems.

Agriculture, Ecosystems & Environment, 74(1), 19-31.

- Altieri, M. A. (2000). Agroecology: principles and strategies for designing sustainable farming systems. Agroecology in action.

- Alwang, J., Siegel, P. B., & Jorgensen, S. L. (2001). Vulnerability: a view

from different disciplines (Vol. 115). Social protection discussion paper

series.

- Arshad, M.A., & Martin, S. (2002). Identifying critical limits for soil quality indicators in agro-ecosystems. Agriculture, Ecosystems and Environment, 88, 163-160.

- Barrett, C. B. (1999). On vulnerability, asset poverty and subsidiarity.

Asset Poverty and Subsidiarity (May 1999).

- Béné, C. (2009). Are fishers poor or vulnerable? Assessing economic vulnerability in small-scale fishing communities. The Journal of

Development Studies, 45(6), 911-933.

- Bentley, L.D., & Whitten, J.L. (2007). System Analysis and Design for the Global Enterprise. McGraw-Hill/Irwin.

- Beroya-Eitner, M. A. (2016). Ecological vulnerability indicators. Ecological

Indicators, 60, 329-334.

- Burke, R. A. (2002). Ecological Indicators for the Nation. Journal of Environmental Quality, 31(3), 1045-1045.

- Chambers, R. (1989). “Editorial Introduction: Vulnerability, Coping, and Policy.” IDS Bulletin, 20(2): 1-7.

- Dale, V. H., & Beyeler, S. C. (2001). Challenges in the development and use of ecological indicators. Ecological indicators, 1(1), 3-10.

(27)

- De Lange, H. J., Sala, S., Vighi, M., & Faber, J. H. (2010). Ecological vulnerability in risk assessment—a review and perspectives. Science of

the Total Environment, 408(18), 3871-3879.

- Eakin, H. (2005). Institutional change, climate risk, and rural vulnerability: Cases from Central Mexico. World Development, 11 (33), 1923–1938. - Eakin, H., & Luers, A. L. (2006). Assessing the vulnerability of

social-environmental systems. Annual Review of Environment and Resources,

31(1), 365.

- Fineman, M. A. (2008). The Vulnerable Subject: Anchoring Equality in the Human Condition. Yale Journal of Law and Feminism, 1 (20).

- Frison, E. A., Cherfas, J., & Hodgkin, T. (2011). Agricultural biodiversity is essential for a sustainable improvement in food and nutrition security. Sustainability, 3(1), 238-253.

- Galloway Jr., R. W. (1989). Basic Equal Protection Analysis. Santa Clara Law

Review, 1 (29).

- Gamanou, G., & Morduch, J. (2002). Measuring vulnerability to poverty. United Nations University, World Institute for Development Economics Research.

- Garbach, K., Milder, J. C., Montenegro, M., Karp, D. S., & DeClerck, F. A. J. (2014). Biodiversity and ecosystem services in agroecosystems. Encycl.

Agric. Food Syst.

- Goodin, R. E. (1986). Protecting the vulnerable: A re-analysis of our social

responsibilities. University of Chicago Press.

- Hugo, S., Squalli, J., & Wilson, K. (2006). What Explains Market Access.

Economic & Policy Research Unit Working Paper, (06-08).

- Lange, M. M., Rogers, W., & Dodds, S. (2013). Vulnerability in research ethics: a way forward. Bioethics, 6 (27), 333-340.

- Luers, A. L., Lobell, D. B., Sklar, L. S., Addams, C. L., & Matson, P. A. (2003). A method for quantifying vulnerability, applied to the agricultural system of the Yaqui Valley, Mexico. Global Environmental Change, 13(4), 255-267. - Measuring market access for agricultural products. (2009, August 7).

Retrieved from https://farmingfirst.org/2009/08/measuring-market-access-for-agricultural-products/

(28)

- Menken, S., & Keestra, M. (2016). An Introduction to Interdisciplinary Research, Theory and Practice.

- Morduch, J. (1994). Poverty and vulnerability. The American Economic

Review, 84(2), 221-225.

- Moret, W. (2014). Vulnerability Assessment Methodologies: A Review of

the Literature. United States Agency for International Development

(USAID) Cooperative Agreement .

- Morton, J. F. (2007). The impact of climate change on smallholder and subsistence agriculture. Proceedings of the National Academy of Sciences

of the United States of America, 50 (104), 19680–19685.

- Naudé, W., Santos-Paulino, A. U., & McGillivray, M. (2009). Measuring vulnerability: an overview and introduction. Oxford Development Studies,

37(3), 183-191.

- Nortcliff, S. (2002). Standardisation of soil quality attributes. Agriculture,

Ecosystems and Environment, 88, 161-168.

- Sen, A. (1997). Resources, values, and development. Harvard University Press.

- Schoenholtz, S.H., van Miegroet, H., & Burger, J.A. (2000). A review of chemical land physical properties as indicators of forest soil quality: Challenges and opportunities. Forest Ecology and Management, 138, 335-356.

- Smit, B., & Wandel, J. (2006). Adaptation, adaptive capacity and vulnerability. Global environmental change, 16(3), 282-292.

- Sommerville, M.A., & Rapport, D.J. (2000). Transdisciplinarity: Recreating integrated knowledge -- advances in sustainable development. Oxford, UK

- Tao, F., Hayashi, T., & Lin, E. (2002). Soil vulnerability and sensitivity to acid deposition in China. Water, Air and Soil

pollution, 140, 247-260.

- Thamaga-Chitja, J. M., & Morojele, P. (2014). The context of smallholder farming in South Africa: towards a livelihood

(29)

- Turner, B. L., Kasperson, R. E., Matson, P. A., McCarthy, J. J., Corell, R. W., Christensen, L., Eckley, N., Kasperson, J.X., Luers, A., Martello, M.L., Polsky, C., Pulsipher, A., & Schiller, A. (2003). A framework for vulnerability

analysis in sustainability science. Proceedings of the national academy of

sciences, 100(14), 8074-8079.

- Vincent, D. P. (1976). Economic aspects of farm poverty. Australian Journal

(30)

(31)

(32)

(33)

(34)

Appendix 5

Here we show an example of the vulnerability equation in practise. Since the purpose is show how the equation works we will only use five indicators for which the values will be assumed. If we assume the indicators to be of equal

importance the weight will be 100 divided by 5, the number of indicators, resulting in a weight of 20. In this example we will show how variation in the weights can lead to different values of vulnerability.

The indicators that are used in this example and the functions they represent are listed in the table below, as well as their safezones, ‘measured’ value, deviation number and weight.

Function Indicator Lower

limit (Safezone ) Upper limit (Safezon e) Value Deviation

number Weight (mean is 20), first and second example Sufficient

entitle ments (Monthly) incomeper household, in Dollars 230 dollar - 200 0.4 34 -9 25 Right crop conditions pH 7.5 6.5 7.2 0 10 Right crop

conditions Nitrogen concentration in soil

(concentration in percentages)

1.5 2 1.35 0.1 20

Sustainable

ecosystem Soil organic matter content (SOM), in percentages 1.5 60 45 0 25 +17 42 Market

availability Number of markets reachable in certain amount of time. 1.5 - 1 0.3 11 -8 3

(35)

(36)

With the data we can fill in the equation presented in chapter 9:

Vulnerability: 0 - 100

Vulnerability = (Deviation1 * weight1) + (Deviation2 * weight2) + … +

(deviationn *weightn)

Vulnerability = (0.4 * 34) + (0 * 10) + (0.1 *20) + (0 *25) + (0.3 *11) =

18.9

Eventually additional research may result in a scale with scenario for ranges of values, 0-10, 10-20, etc. This could make decisions for aid and policies less complicated.

As mentioned before we can adjust the weights for different situations or time scales. If for instance, a farmer produces mainly for his or her family the indicators for the function of available markets become less important. Simultaneously we increase the weight of the ecological sustainability to

represent a vulnerability calculation for a longer time scale, we also decrease the weight of the entitlements as these may become less important on a longer time scale.

Vulnerability = (0.4 * 25) + (0 * 10) + (0.1 *20) + (0 *42) + (0.3 *3) = 12.9 In this case it will result in a less vulnerable system. By adjusting the weights in certain situations or for other time scales to specify the model, the right increments for the deviations (0 to 1) need to determined only once.

Assessing the Vulnerability of the Social-Ecological system of a Smallholder Farmer