Livelihood Diversification in Rural Laos

World Development Vol. xx, pp. xxx–xxx, 2016 0305-750X/Ó 2016 Elsevier Ltd. All rights reserved. www.elsevier.com/locate/worlddev

http://dx.doi.org/10.1016/j.worlddev.2016.01.018

Livelihood Diversification in Rural Laos SARAH M. MARTIN a and KAI LORENZEN b,* a Indian Ocean Tuna Commission, Victoria, Seychelles b University of Florida, Gainesville, USA Summary. — Livelihoods in rural southern Laos are highly diverse, comprising a wide range of different productive activities. In this paper motivations for diversification within livelihood strategies in rural communities in southern Laos are investigated through theories of distress-induced risk spreading in response to crises and progressive success and wealth. Results indicate that livelihood diversification at the household level is associated with higher wealth status and ownership of a range of assets as part of a progressive, accumulation livelihood strategy for those with fewer constraints. Diversification strategies across all socioeconomic groups commonly include a combination of agricultural and non-agricultural activities as well as migrant remittances, however, the types of activities undertaken are dependent on wealth. The increased asset status of households extends to enabling higher income-generating migration opportunities and may facilitate transition into both non-agricultural employment, as well as into a wider variety of agricultural employment activities. This further strengthens livelihoods through the mutually reinforcing complementarities across livelihood activities and reducing the risks associated with each. But this has the effect of leaving the poor, with lower levels of diversification, at most risk to natural or economic disasters or other shocks. Results have implications for development policy for rural southern Laos, highlighting the importance of recognizing the positive aspects of livelihood diversification for rural poverty reduction. Broadened policy mechanisms which support and encourage diversification and mobility at the household level are needed. Likewise, pro-poor development initiatives that focus on increasing the diversity of assets (rather than the quantity of any one single asset) of the poor are more likely to be successful in supporting livelihood diversification and reducing vulnerability. Ó 2016 Elsevier Ltd. All rights reserved. Key words — agricultural livelihoods, diversification, asset wealth, Laos, resilience, risk mitigation

1. INTRODUCTION

be broadly distinguished as reflecting either: (i) livelihood distress or (ii) progressive success.

(a) Livelihood diversification (i) Distress diversification Distress diversification is where diversification is seen as a strategy of spreading risk to reduce vulnerability to unpredictable crises such as floods, droughts, and illness as well as the seasonal fluctuations of natural resources (Bruge`re, Holvoet, & Allison, 2008; Ellis, 2000; Freeman & Ellis, 2005; Lohmann & Liefner, 2009; Smith et al., 2005). The extent to which risk affects different households and their behavior depends on their risk aversion which varies inversely with wealth and liquid assets. It is thought that in general, poorer households with few liquid assets have a higher risk incentive to diversify than richer households with more assets (Haggblade, Hazell, & Brown, 1989). It is often stated that this type of ‘‘distress-pushed” diversification forces people into a variety of low-return activities, leading to more stable but lower household income (Lohmann & Liefner, 2009; Reardon et al., 2000). This is due to the trade-off between diversification as a risk avoidance strategy and specialization as a more efficient, higher income, but risk-prone strategy (Wilen, Lockwood, & Botsford, 2002). Risk-averse households are willing to pay the implicit insurance premium in the form of foregone gains from specializing as a method of insuring against income shocks (Anderson & Deshingkar, 2005; Iiyama, 2006; Reardon et al., 2000; Wilen et al., 2002). In this light, diversification is seen as an involuntary reversion of the

Livelihood strategies are the combinations of activities and assets that generate the means of household survival. Despite increasing attention in the literature, rural livelihood strategies are still not well understood and limit the understanding of resource-user behavior (Pauly, 2006; Salas & Gaertner, 2004). Although agriculture predominates in many rural communities, livelihoods are complex and rural households are often pluriactive, maintaining a diverse portfolio of activities among which crop and livestock production feature alongside many other contributions to family well-being (Dorward, 2002; Smith, Nguyen Khoa, & Lorenzen, 2005). The process by which rural households construct a diverse portfolio of activities and assets in order to survive has been defined as ‘‘livelihood diversification” by Ellis (2000). Livelihood diversification is widespread and found in all locations as well as across farm sizes and wealth groups, and is not only a distinguishing feature of rural survival strategies in contemporary poor countries, but is also present in urban areas of developing and rural areas of developed countries (Ellis, 2000). The many different combinations of livelihood activities that are practised imply that the goals and motivations for diversification are themselves varied, so allow for multiple interpretations of what these comprise (Perz, 2005b). In rural southern Laos, occupational multiplicity is becoming more common and more pronounced (Rigg, 2006b), however little is known about why, how, and what implications this has for the rural poor. There is extensive debate about the role of livelihood diversification in the literature (e.g., Allison & Ellis, 2001; Anderson & Deshingkar, 2005; Ellis, 1998, 2000; Marschke & Berkes, 2006; Reardon, Taylor, Stamoulis, Lanjouw, & Balisacan, 2000) which can

* This research was funded by the Natural Environment Research Council and the Economic and Social Research Council. The authors would like to thank the Department of Livestock and Fisheries in Savannakhet and the Wetlands Alliance for providing field support with particular thanks to Lamgneun Phengsikeo. Final revision accepted: January 11, 2016. 1

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

2

WORLD DEVELOPMENT

process of specialization, brought on by crises such that the multiplication of activities is an adaptation necessary to ensure survival (Cinner, Mcclanahan, & Wamukota, 2010). (ii) Progressive diversification Although risk theory predicts that households will diversify less as their wealth increases and risk aversion motives decline, liquidity and credit constraints on diversification for the poor may lead to the opposite outcome: more diversification among the non-poor than among the poor. Progressive success and wealth, which in turn lead to increased access to resources, may lead to increased livelihood diversification as although they may have lower risk incentives than the poor, the nonpoor may be more capable of financing this diversification if it is costly, has high entry barriers, and is initially risky (Reardon et al., 2000). From this point of view, diversification can be seen as a deliberate strategy adopted by pro-active households with greater opportunities (Cinner et al., 2010). Although this seems contrary to the theory of specialization, diversification at the household level may not necessarily reduce efficiency, as it still allows for individuals to specialize and develop skills within a household (Ellis, 2000). This progressive diversification can be further investigated in terms of household assets. The asset categories: natural, physical, human, financial, and social capital have been defined as analytically useful components of the assets that underpin individual and household livelihood strategies by determining the livelihood options available to households as part of the Sustainable Rural Livelihoods framework (Ellis, 2000; Iiyama, Kariuki, Kristjanson, Kaitibie, & Maitima, 2008). Variation in assets may be due to differences in both the amount and diversity of those assets, and both may affect livelihood diversification. It has been hypothesized that possessing a diverse array of assets may be important for maintaining a variety of livelihood activities (Anderson & Deshingkar, 2005; Bebbington, 1999; Perz, 2005b). This theory of diversification suggests the amount of diversity in a household’s portfolio reflects the amount of diversity in the assets (or factors of production) it owns or has access to, rather than the quantity of those assets. In contrast, Perz (2005b) suggested that inequality in terms of a given asset may be important in enabling a household to add a specific livelihood income to a household, as the level of capital asset endowment affects the capacity to invest (Iiyama, 2006), and a study in Vietnam proposed that this ability to mobilize capital is essential to enabling diversification (Thulstrup, 2015). (b) Types of diversification strategy There are many risks associated with agriculture, and many rural farm households may be unable to meet basic needs, so household members often search for alternative means of livelihoods to cope (Chianu, Ajani, & Chianu, 2008). As a result, most rural households depend on some combination of agricultural and non-agricultural activities to make a living. Increasing livelihood diversification in Laos is being accompanied by a trend toward deagrarianization as the role of non-farm activities grows in importance (Bouahom, Douangsavanh, & Rigg, 2004). There are a number of different motivations for allocating labor to the non-farm sector including: (i) better relative returns, (ii) inadequate farm output, (iii) a need for non-farm cash sources to pay for farm inputs, and (iv) risky returns to farming (Reardon et al., 2000). Non-farm activities have the potential to play a crucial role in reducing vulnerability to poverty by providing households with a form of insurance against the risks of farming

and reducing reliance on natural resources (Lohmann & Liefner, 2009; Rigg, 2006a). Non-farm activities also enable households to generate capital to adopt new production methods and raise output (Evans & Ngau, 1991). However, the reported effects of off-farm and non-farm employment on rural income inequality are mixed. Bouahom et al. (2004) described a situation in northern Laos where due to the little hope of intensified agricultural production lifting households into food security, households were driven to become dependent on off- or non-farm work by necessity, not choice, in an example of distress diversification. In the Amazon region Perz (2005b) found that few households diversified into non-agricultural income sources, and those that did also had incomes comparable to households primarily reliant on agriculture. Illustrating further differences in results, Iiyama et al. (2008) concluded from a study of African farmers that the wealthiest households were those with non-farm income, and those pursuing higher return agricultural activities. This positive association between non-farm income shares and total income or wealth levels in Africa was also observed by Reardon et al. (2000) who suggested a rough pattern: a positive relationship between non-farm income share (and level) and total household income and/or landholding in much of Africa, a negative relationship in much of Latin America, and a very mixed set of results in Asia. In summary, the evidence is very varied as to the effect of non-farm employment on rural income. The increase in non-farm activities in rural southern Laos has been accompanied by evidence of heightened levels of mobility and delocalization with remittances playing a growing role in household income (Rigg, 2005, 2006b). However, in Laos, there are significant gaps in knowledge regarding migration, including basic information such as the incidence, type, and geographical patterns of movement (Rigg, 2007). Migration may be propelled by poverty, reflecting resource scarcities at the local level (Rigg, 2006a) or encouraged by wealth as an outcome of prosperity. (c) Objectives The reasons for livelihood diversification are varied, ranging from an attractive choice for accumulation purposes, enabled by asset wealth and the diversity of those assets, to a distressinduced insurance strategy brought on by crises. Identifying the type of diversification that takes place is important for forming appropriate policy and development initiatives to support the rural poor. If livelihood diversification is a sign of progress by which households are lifting themselves out of poverty then this should be supported by relevant policy mechanisms and initiatives, however, diversification may alternatively be providing a signal that a particularly vulnerable household is in distress and in need of support through a different approach. Determining why occupational multiplicity is becoming more common in rural southern Laos and how households are diversifying their livelihoods therefore has important implications for pro-poor policy and development initiatives. This study investigates the motivations for livelihood diversification in rural southern Laos by exploring the two hypotheses of progress-pulled and distress-pushed diversification. This was investigated through analyzing whether the level of occupational diversification of a household was positively or negatively correlated with its wealth status. As the measurement of wealth was derived from a combination of the amount and variation of assets owned, the effect of these were also analyzed explicitly by exploring the relationship between the quantity of assets owned

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS

by a household and its level of diversification as well as the diversity of assets held and diversification. The type of diversification strategies undertaken were then explored further to investigate differences in the kind of activities undertaken among different socio-economic groupings, including diversification into non-agricultural livelihood activities and spatial diversification through the role of remittances. This analysis focussed on whether there were any potential barriers to these types of diversification for the rural poor such as land-poverty, low education, or lack of available household labor. 2. METHODS (a) The study area The field study was carried out in Savannakhet Province, southern Laos (Figure 1). Laos has a tropical climate with an average daily maximum temperature of 31 °C and an average annual precipitation of 1,500 mm, about 75% of which occurs in the monsoon season (May–October). Laos is one of the poorest countries in the world and has a primarily agricultural economy; the sector produces over 50% of GDP (Bouahom et al., 2004). Results from the most recent Laos agricultural census indicated that 77% of the total population is engaged in farming, with 97% of farmers owning their land (APCAS, 2012). In Savannakhet Province, the majority of farmers are subsistence farmers, producing rice mainly for their own consumption (Chinvanno, Souvannalath, Lersupavithnapa, Kerdsuk, & Thuan, 2008). They have farms of moderate but sufficient size for producing rice to support the annual consumption of the farm house-

3

hold, and use of mechanized and advanced farm technology is limited. The field study was carried out in the district of Champhone, a wetland region of the Xe Champhone river, an important tributary of the Mekong. The landscape comprises rice paddy (accounting for about 80% of the cultivated area), lakes, swamp forests, freshwater marshes, and numerous dry season standing waterbodies (Claridge, 1996; Martin, Lorenzen, Arthur, Kaisone, & Souvannalangsy, 2011). The rice field production systems in Champhone are all composed of lowland rainfed paddy of which some is also irrigated. Localized seasonal flooding affects the type of farming possible so rainfed, non-irrigated paddy may be further divided into land which floods during the wet season and land which is situated on higher ground or further from the river so does not flood. Land quality affects agricultural productivity and food security through its impact on both food supplies and farmers’ incomes, which in turn, influences farmer decision-making. (b) Rapid rural appraisal A rapid rural appraisal (RRA) was conducted as part of an initial exploratory analysis to provide general background information about rural households and their livelihoods. Focus groups were asked to identify sources of household income, village resources, seasonality of resource use, and indicators of socio-economic groupings. Groups of participants ranged in size from six to 40 people, with a mixture of ages and gender. Validation by triangulation was used in all ranking exercises which involved establishing three separate focus groups to discuss a topic simultaneously in order to validate responses. Methods used included matrix ranking, village resource mapping, and seasonality diagrams of livelihood activities. These were all open-ended response questions, so where the relative importance of different livelihood activities was considered, matrices were left blank for respondents to create the relevant categories. (c) Participatory wealth ranking (PWR) A definition of wealth was sought to distinguish between households of different socio-economic groups. Wealth ranking is a participatory method of ranking households based on income, wealth, and other local measures of well-being to reflect the local definition of wealth (Be´ne´, Mindjimba, Belal, Jolley, & Neiland, 2003). This was carried out using a card sorting method in which a list of all households in the village was provided by the headman (Mukherjee, 1993). Households were then ranked by three separate groups of village informants (for triangulation purposes) by their own ranking criteria. Informants were later asked to name the criteria which were most influential in their decisions regarding wealth groups, on which socio-economic indicators for a household survey were based. (d) Household survey

Figure 1. Map of the study area, Savannakhet Province, Laos.

A formal household survey was conducted using a structured questionnaire asking questions regarding asset wealth and occupational activities. The indicators of wealth used were developed based on a preliminary analysis of the PWRs, RRAs, and findings of Garaway (1999) and Bush (2004). These included: size of household, amount of land, number of buffalo, number of cattle, ratio of dependents, rice production, television ownership, and additional income. An available adult member of the household was invited to

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

4

WORLD DEVELOPMENT

participate in the survey. Respondents were questioned about household demographics, their levels of self-sufficiency, and assets and were asked to describe all activities carried out by the household. A sample size of 107 per socio-economic group was calculated as necessary to be 80% certain of being able to detect a significant difference (a = 0.05) in the socio-economic status of the households (Sokal & Rohlf, 1969). A stratified sample of ten villages was selected, representing a variety of distances from fishing grounds and markets, chosen with the help of the local knowledge of Livestock and Fisheries Department district staff, previous RRA information, and a map. Stratification within villages was carried out by selecting an equal percentage of each wealth group classified by the wealthranking exercise; ‘‘rich”, ‘‘middle”, and ‘‘poor”, so that the sample was composed of a proportion of each wealth group consistent with the proportion of that wealth group within the total village population. Households were then selected randomly from each of the three groups. Sampling continued until each wealth group had a minimum of approximately 107 households resulting in a total of 519 households surveyed. Surveys took place from October to December 2008. (e) Metrics (i) Wealth index Household income and consumption expenditures are routinely used in measuring long-term poverty and welfare (Filmer & Pritchett, 2001); however, in developing countries data on household income or expenditure are often unavailable or unreliable as measures of economic status due to the difficulty in assigning a monetary value to everything in a predominantly subsistence-based economy. Therefore, assetbased indexes provide a suitable alternative as a measure of economic status (Hargreaves et al., 2007; Houweling, Kunst, & Mackenbach, 2003; Howe, Hargreaves, & Huttly, 2008). In the study villages, direct income was not considered a major criterion of wealth, a similar finding to other inland fisheries (Be´ne´ et al., 2003), but instead various assets such as rice fields were considered important. This result can be related to the fact that the majority of households are still heavily involved in a subsistence-based economy where rice farming, livestock farming, and fishing represent major pillars of the system (Martin, Lorenzen, & Bunnefeld, 2013). Therefore, rather than an analysis of income and consumption expenditure, an asset-based approach was used to develop an index of wealth using data from the household survey. Household survey data on the selected indicators were aggregated into a wealth index through the application of a Principal Components Analysis (PCA) (Montgomery, Gragnolati, Burke, & Paredes, 2000). A PCA was used to transform the set of correlated variables into a set of uncorrelated ‘‘components”. The first principal component was selected as the linear index of all the variables that captured the largest amount of information common to all the variables which was then used as the wealth index (Cordova, 2008; Filmer & Pritchett, 2001; Mckenzie, 2003; Vyas & Kumaranayake, 2006). Variables were selected based on a literature review of assetbased indicators from which factors such as family size, number of animals, farming tools, and ownership of land were selected (Filmer & Pritchett, 2001; Vyas & Kumaranayake, 2006), as well as context-specific indicators such as ‘‘food sufficiency” derived from previous studies in Savannakhet

(Garaway, 1999) and using information from the PWR. A wide range of assets were chosen in order to avoid the problems associated with clumping/clustering and truncation (Mckenzie, 2003). Sixteen assets were used as indicators for the asset index (Table 4), and produced very similar classifications when different subsets of variables were used in its construction, suggesting that the final choice of assets did not substantially impact the results. (ii) Asset diversity The Sustainable Rural Livelihoods (SRL) framework was used to investigate the impacts of assets and vulnerability on diversification. In the livelihoods approach, ‘‘assets” refer to resources owned or accessed by household members and are often classified into five categories: human capital (skills, education, health), physical capital (produced investment goods), financial capital (money, savings, loan availability), natural capital (land, water, trees), and social capital (networks and associations). This approach regards the asset status of the poor as fundamental to understanding the options open to them, the strategies they adopt for survival, and their vulnerability to adverse trends and events (Allison & Ellis, 2001). Many different definitions and categorizations have been used in studies investigating assets, rural livelihood activities, and diversification (Anderson & Deshingkar, 2005; Barrett & Reardon, 2000; Perz, 2005b). The SRL classification was used in this study as the groupings had the least potential for overlap among categories yet still included labor, which is regarded as a critical component in determining household livelihood strategies (Barrett et al., 2006; Smith et al., 2005) (Table 1). (iii) Occupational diversity Quantifying the diversity of household activities has also been undertaken using a variety of different methods and a variety of activity categories (Anderson & Deshingkar, 2005; Cinner, Daw, & Mcclanahan, 2008; Cinner et al., 2010; Perz, 2005b). These are often confounded by incorporating a mixture of spatial, sectoral, and functional categories. Here, the diversification variable termed ‘‘occupational diversity” was defined as the total number of household productive activities using sectoral classifications rather than functional and spatial groupings, due to the decreased ambiguity (Barrett & Reardon, 2000). The categories and sub-categories of activity were based on those described by Ellis (2000), modified to ensure the basis was purely sectoral and relevant to the local situation in Laos (Table 2). There are a number of different definitions of what constitutes farm and non-farm activities, but Ellis (2000) distinguished between farm, off-farm, and non-farm activities by classifying them as: Farm—own-account farming, whether owner-occupied land, or land accessed through cash or share tenancy. Includes livestock and crops. Off-farm—typically wage or exchange labor on other farms. Includes labor payments in kind, e.g., harvest share systems. Also includes local environmental resource collection, e.g., wild plants, firewood. Non-farm—non-agricultural activities including non-farm rural wages or salary or self-employment, e.g., a business. Urban to rural remittances within national boundaries and urban transfers to rural houses, e.g., pensions and international remittances. ‘‘Farm”, ‘‘non-farm”, and ‘‘off-farm” are often used as separate categories, however, as ‘‘off-farm” is a spatial indicator

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS

5

Table 1. Asset categorizations according to the SRL framework and indicators of each used in this study Human  Labor (adults of a working age)  Education

   

Physical

Financial

Motorbike Shop Boat Tractor

   

Cows Pigs Poultry Buffalo

Social

Natural

 Migrant relatives  Government connections

 Ponds  Rice land  Other land

Table 2. The categories and sub-categories* of activities used to define occupational diversity Natural resource-based activities Collection (forests, aquatic resources) Cultivation (food) Cultivation (non-food) Livestock Non-farm natural resource use e.g., brick making, weaving, carpenter Fishing *

Non natural resource-based activities Rural trade/business Services/profession (e.g., teacher, health visitor) Rural manufacture Skilled labor (e.g., mechanic, sewing) Vehicle driving Industrial laborer

Remittances not used as they are a spatial classification.

whereas the other two are sectoral, the two categories of ‘‘farm” and ‘‘non-farm” were used alone in this study, where the term ‘‘farm” also incorporates ‘‘off-farm”. (f) Statistical methods To assess the consistency of socioeconomic groupings defined by the PWR exercises among villages and the mean PCA-derived wealth index, the two were compared. Being composed of directly measurable components, the wealth index is free from the relative subjectivity of the participatory wealth rankings within each village and so can validate the rankings across villages. Nevertheless, the wealth rankings include indicators which may not have been measured or may not be possible to measure and are important to the local definition of wealth, so as well as each providing a means of validation for the other, the two methods are also complementary. A constant was added to the calculated wealth index and this was cube root transformed to ensure normality of errors and constancy of variance and a one-way ANOVA was used followed by a Tukey test. In order to test the theories of distress-pushed and progressive diversification, the relationship between wealth and occupational diversity was analyzed. With distress-induced diversification, increased wealth would be expected to reduce the vulnerability of the household and therefore lower the need for insurance through diversification. Under this hypothesis, a negative correlation would be expected between wealth or remittance value and occupational diversity. In contrast, with a progressive view of diversification, increased wealth, and an increased value of remittances would be expected to positively affect occupational diversity by enabling investment in a broader range of livelihood activities. An ANOVA was used to examine the relationship between wealth and occupational diversity, where occupational diversity was defined as the number of different sectoral employment categories and wealth as the normalized household wealth index. Y ¼ b0 þ b1 X þ e where Y and X are indexes of wealth and occupational diversity respectively. The wealth index comprised a combination of both the quantity and diversity of assets, so to explore the relationship

further, the role of household assets in shaping diversification strategies was analyzed through investigating the relationship between the quantity and diversity of household assets with occupational diversity separately. The quantity of assets was standardized as the number of individual household assets (number of ponds, buffalo, pigs, poultry, paddy land, and cows) per working member of the household (household members minus the number of dependents: >60 or <16). This was analyzed using a glm with poisson errors. The relationship between the diversity of household assets (based on Table 1 classifications) and occupational diversity was also explored using a glm with poisson errors. An overview of the typical values for household asset diversity is given in Table S6. The type of diversification undertaken among households was also explored to investigate differences between livelihood strategies undertaken by different socioeconomic groups. The percentage of farm and non-farm activities undertaken by different wealth groups was assessed using a glm with binomial errors. The number of farm and non-farm activities undertaken by different wealth groups was also analyzed, using an analysis of deviance with poisson errors and means of each wealth group were compared using Tukey contrasts. The hypotheses that land poverty and lack of education are entry barriers to undertaking non-farm activities were also assessed using a glm with poisson errors with the explanatory variables: area of rice land and highest education level attained within the household. The relationship between wealth and migration was also investigated. The proportion of poor, middle, and richranked wealth groups receiving a remittance were compared using an analysis of deviance with quasibinomial errors on the wealth-ranked groups. The hypotheses that land poverty, wealth, and education form potential barriers to migration were investigated in a logistic regression using wealth index, area of land per household worker, and highest level of education in the household as explanatory variables and a binary variable indicating whether or not the household received a remittance. To determine whether migration is fueled by wealth, or wealth is a product of migration, the relationship between the value of remittance provided and wealth status of the household was also assessed. If wealth is a product of receiving remittances, it may be assumed that there would be

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

6

WORLD DEVELOPMENT

a significant relationship between the amount of remittance and level of wealth, whereas if migration was dependent on wealth, the value of the remittance would not necessarily be expected to correlate well with the level of household wealth. This was explored using an analysis of deviance to compare the level of remittance received by different wealth groups. 3. ANALYSIS OF RESULTS (a) Socioeconomic status Table 3 provides the main reasons for wealth group classifications as determined by the PWR. Important factors in determining a household’s status as ‘‘poor” included owning little or no rice land, owning poor quality rice land, and the inability to produce enough rice to feed the household. Lack of equipment required for farming, a high number of dependents and the inability to obtain loans hinders progression of this group. Conversely, richer households were characterized by possession of items such as tractors, televisions, and motorbikes (bought outright as opposed to through creditagreements). They also owned large areas of rice land, had alternative sources of income from shops and/or remittance payments, and were more highly educated. Ownership of productive assets such as large numbers of livestock and multiple fish ponds, access to irrigation, and harvesting rice surplus to household requirements were also important indicators of more wealthy households. Inheritance (or lack of) was cited frequently as a cause of wealth or poverty as was a subjective indicator of a household’s work ethic, which included the ability and willingness to work, the level of entrepreneurism shown, and attitude toward saving. These attitudes have been noted in other studies, indicating that they are widely-held views (Bush, 2004). Income level was not mentioned as a criterion of wealth. Households classified as ‘‘middle” wealth were most frequently cited as having sufficient (but not surplus) rice and this group formed much of the range between the two

extremes of ‘‘poor” and ‘‘rich”. However a clear distinction arose from a description (reported separately in different villages) of these households representing a transition phase from poor to rich. These households were cited to have received little inheritance, yet through informal loans and social networks and a strong work ethic were managing to increase their standard of living, productive assets, and family prospects. This description is synonymous with the definition of a ‘‘stepping up” lifestyle described by Dorward et al. (2002). The PCA used to develop a wealth index resulted in a first principal component which explained 21% of the variance. The rice sufficiency index had the highest factor weighting (Table 4). Within the calculation of the wealth index, three variables (area of non-paddy land owned, annual quantity of rice harvested, and number of boats owned) were allocated a negative factor weighting from the PCA. This implies that all other assets being held equal, a household with a boat will be ranked lower in terms of socio-economic status than a household without one. Negative weightings are not uncommon in PCA-calculated wealth indexes, and a reason for such a result may be due to ownership of a boat being more strongly correlated with variables that are expected to be associated with lower socio-economic status, which is likely, given that a higher proportion of poorer households fish in large open-access lakes and rivers (Martin et al., 2013). These variables all had very small weightings (0.026, 0.005, and 0.009) so had little influence on the resulting index. The percentage of variance explained by the first principal component was fairly substantial and within the range generally observed for household socio-economic indicators; 12–27% (Houweling et al., 2003; Vyas & Kumaranayake, 2006). Results from the ANOVA comparing the two measurements of wealth indicated that the mean wealth index was significantly different between poorer, middle, and richer wealthranked groups (p < 0.001) (Figure 2). This suggests that both the wealth rankings obtained from the PWR exercise and the PCA-derived wealth index are good indications of socioeconomic status, and that inter-village heterogeneity for the rankings is small.

Table 3. Frequency of statements made to describe wealth groups during participatory wealth ranking in Savannakhet Theme

Statement

Total

Poor

Farming Food Family Family Farming Farming

Own little or no riceland (<1 ha) Harvest insufficient rice to feed the family Large family (many young children) Not enough labor (few family members working age) Only harvest rice in one season Own no agricultural equipment (might borrow/rent buffalo to work in fields)

8 4 4 3 3 3

Middle

Inheritance/work ethic

3

Food Family Family Farming Farming Money

Were previously poor and worked their way up through borrowing equipment from family and friends and working for other people rather than from inheritance Harvest sufficient rice for family (but no surplus to sell) Sufficient labor (family members of a working age) Large family (many dependents) Own buffalo Own cows Can borrow money from bank to buy fertiliser to farm dry season rice to sell

3 2 2 2 2 2

Family Farming Family Work ethic Assets

Inheritance from parents/grandparents Owns a large area of riceland (>5 ha) Family member working abroad Hard working High-asset ownership and savings (money and rice)

8 4 4 3 3

Rich

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS

7

Table 4. Factor weights for each variable of the first principal component from asset information collected from the household survey Household variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Factor loading

*

Rice index 0.425 Area of rice land owned 0.347 Number of hand tractors 0.331 Number of poultry 0.331 Number of cows 0.325 Number of motorbikes1 0.314 Number of private ponds 0.291 Number of TVs 0.276 Number of pigs 0.202 Irrigation access 0.157 Income from migrant relatives 0.142 Number of buffalo 0.138 Shop 0.093 Area of non-paddy field land owned 0.009 Annual quantity of rice harvested 0.005 Number of boats 0.026 h i ðaþbÞ * Rice index ¼ 5 þ 0:5 where a = 0.5* number of yrs not enough rice, and b = 0.5* yrs of surplus rice within the previous 5 years (Garaway, 1999). 1 Complete ownership, not including those purchased through credit agreements.

Figure 2. Participatory wealth-ranking categories compared to the wealth index as defined by a PCA of household survey asset data.

(b) Livelihood diversification Results from the household survey indicated that in terms of the number of households involved in each livelihood activity, rice farming was the most common activity, undertaken by almost all households (94.2%), followed by poultry farming (87.3%), cattle farming (71%), and fishing (57.9%). (i) Distress vs. progressive diversification The relationship between household wealth index and occupational diversity was positive, corresponding to the theory of progressive diversification (Figure 3). The mean wealth indexes at different levels of occupational diversity were significantly different (p < 0.01, n = 519). Multiple comparison Tukey test indicated that the wealth of households with 5 and 1, 6 and 1, and 4 and 1 were all significantly different. The positive relationship between wealth and occupational diversity did not correspond to a distress theory of diversification, in which a negative relationship would have been expected. This may indicate that distress is not the primary

Figure 3. Occupational diversity and wealth index (p < 0.01).

reason for diversification. However, while a negative relationship between wealth and diversification would have led to a fairly clear rejection of the progress hypothesis, the relationship was not so clear. While there was a lack of richer households with very few livelihood activities, some poorer households still had fairly high levels of occupational diversity. This suggests that increasing levels of diversification may be the first step in ‘‘virtuous spirals of accumulation” similar to those described as taking place in east Africa (Ellis & Freeman, 2004). But this also means that those poorer households with very low occupational diversity may simply be too poor to be able to diversify. Another factor to consider is that despite the relative socioeconomic distinctions, wealth was low throughout the entire study population. Therefore, even the better-off households may still be in need of some level of diversification as an extra insurance measure to mitigate against the high levels of natural and economic risk. If insurance is a normal good, then increased wealth is likely to lead to increased demand and so there may still be some form of risk avoidance behavior contributing to the diversification seen in the richer households that are better able to finance it. Even if there may still be some risk avoidance mechanisms at play, diversification is clearly a positive, preferred livelihood status with all richer households having a relatively high level of occupational multiplicity. This indicates that there is some form of progress-pull, and is indicative of the increased opportunities available through diversification as a means of raising total household income. Due to the high seasonality of rice farming, the main reason for diversification could be to utilize household resources more productively through year-round employment. Some farms are unable to cultivate rice during the dry season due to lack of irrigation facilities, while others are unable to farm at times when the flooding is so high that their paddy land is swamped and so seeking alternative, higher income-generating employment at these times when productivity is known to be low would have the effect of increasing annual household income. Other complementarities between livelihood activities may also exist such as the synergies that exist between fishing and farming (Martin et al., 2013), such as passive net fishing undertaken while actively farming in the interim. Qualitative information on the opportunities available to households within different villages and the access to these resources also supported the progress theory of diversification. Livelihood strategies vary seasonally with changes in labor availability depending on the type of farm production system

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

8

WORLD DEVELOPMENT Table 5. Village system characteristics (RRA results) Village

Farm production type

Livelihood system

System response

Huameuang

Low-lying, near Xe Champhone, poor crops in rainy season due to flooding

Near Kengkok market. Land scarce

Dorn Yeng

Low-lying, near Xe Champhone, poor crops in rainy season due to flooding Low-lying, near Xe Champhone, poor crops in rainy season due to flooding

Few job opportunities Few job opportunities

Rice farming and fishing dominate activities. Poor villagers rely heavily on fishing

Tan Soum

Low lying, far from Xe Champhone. Good rainy season and dry season rice crops

Most households carry out rice farming activities all year round

Paleng

High-lying, far from Xe Champhone. Good rainy season and dry season rice crops High-lying. Good rainy season and dry season rice crops with large irrigation system

Rice farming dominates opportunities. Cheap irrigation from other villages due to lie of the land Mainly rice farming, small pump for irrigation. Small forest area, close to Kengkok market Rich soils, large forested areas, so a number of livelihood options are available: rattan, vegetables, coconuts, farming, fishing. Near Song Khong market Good farming opportunities with irrigated paddy and forested area Little land ownership due to resettlement. Near Kengkok town and market. Clay soils, so small clay-pot making business Lots of riceland available but not much irrigation access so hardly any dry season rice. Clay soils, so village supports 3 small factories making clay stoves No forest cover

Jobs in local district government offices, farming, and fishing. Those with no land fish and work on other peoples’ land. Clay pot making More fishing takes place during the dry season when there is very little rice farming High employment by stove factory, rice farming

Lam Phan

Low Haukham

Dorn Keng

High lying, far from Xe Champhone. Good rainy season and dry season rice crops with large irrigation system

Dong Nong Kheun

Low-lying, far from Xe Champhone. Good rainy season rice crops, but little access to irrigation in the dry season

Vatthana

High lying, far from all waterbodies. Good rainy season rice crops, but little access to irrigation in the dry season High lying rice land. Good rainy season rice crops, but little access to irrigation in the dry season

Buek Thong

owned. Access to resources played a large part in the opportunities open to people and the types of livelihood activities they would become involved with. This is illustrated in Table 5 which provides summaries of the options available to particular villages and illustrates their dependence on the resources in the area. In Buek Thong, clay soils enabled the establishment of a stove-making industry, and in Low Haukham, the fertile soils made other activities possible such as rattan and other non-rice crop farming. The positive correlation between occupational diversity and wealth may also be due in part to the high work ethic of the wealthy (Table 3), enabling them to undertake more activities by working longer hours. Both higher work ethic and complementarity of activities fit with the findings of Anderson and Deshingkar (2005) where even diversification at the individual level was associated with higher overall earnings. While the metric used in this study was the household, investigation into diversification within households by individuals is an area for further research as the benefits of diversification as opposed to specialization at the individual level may be very different. The finding that diversification appears to be progressive within these communities corresponds with the main documented trends in artisanal and small-scale fishing communities where diversification is often deemed to be ‘‘for accumulation” (Coulthard, 2008). Similar findings have also been observed for agricultural livelihoods in India where the effect of diversification on poverty was generally positive, however, only upto

Many fishing households to compensate for poor rice harvests. Little land. Other non-farm employment takes place Many households fish but are gear limited

Most households carry rice farming activities all year round with some fish farming Many activities undertaken. Dried rattan sold to international market (USA)

Rice farming and rattan farming predominate

a certain threshold due to constraints such as labor availability (Birthal, Roy, & Negi, 2015). In Vietnam, diversification was also found to be a positive livelihood strategy (Thulstrup, 2015) as well as in parts of east Africa (Ellis & Freeman, 2004). Niehof (2004) suggested that in growing rural economies, diversification reflects the ‘‘good” type of diversification, whereas in stagnating rural economies diversification reflects poor people coping with risks. As Laos is becoming more monetized, more liberalized, and more integrated (Bouahom et al., 2004), the results fit this hypothesis and illustrate how these changes are affecting livelihood strategies at the local level. (ii) The role of assets in diversification strategies Quantity of assets. As the wealth index is derived from asset ownership, this supports the asset-based view of progressive diversification. However, when individual assets were considered, increasing levels of asset per worker did not correspond to an increase in household diversification. Lack of a significant relationship between the quantity of a single asset and occupational diversity was apparent with all of the assets considered. This same result was observed in the Andes where a high number of an individual asset (sheep), which could easily be liquidated to mitigate against losses following shocks, was not associated with a reduction in livelihood diversification (Valdivia, Dunn, & Jette´, 1996). This suggests that despite the increased investment opportunities gained from owning

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS

9

Figure 4. Effect of the number of different household assets on occupational diversity and total number of livelihood activities (larger circles represent a greater number of data points). (a) Occupational diversity, (b) total number of livelihood activities.

much of one asset, it does not positively impact diversification, but it also does not reduce the likelihood of diversification even though the household is more able to mitigate against the impact of shocks after the event. Diversity of assets. The number of different assets a household possessed was significantly correlated with occupational diversity (p < 0.02) and highly significant (p < 0.01) when the total number of livelihood activities was used instead of occupational diversity (Figure 4), indicating that livelihoods are more diverse with increasing asset diversity. This suggests that diversity of assets is more important than the quantity for enabling livelihood diversification. At the village level, those with access to a variety of assets were associated with greater wealth through the increased opportunities for households to undertake multiple occupations. Low Haukham was described as one such village with high diversity through good access to forests resources, rattan farming (with an export market to the USA), close proximity to a large reservoir for fishing, established fruit trees, irrigation facilities, and fertile farm land (Table 5). This village was described as being wealthy due to the ‘‘many jobs” held by the villagers. This higher occupational diversity among the asset diverse has also been found in South America and Africa (Iiyama et al., 2008; Perz, 2005a).

riers to many high-productivity segments of the rural nonfarm economy, such as the availability of high financial, human, and political capital. The presence of these barriers suggests the poor may be left dependent on increasingly lowreturn farm or off-farm activities, leading to an increase in wealth inequality (Iiyama et al., 2008; Perz, 2005b; Reardon et al., 2000). The fact that poorer groups still took part in the non-farm economy, albeit to a lesser extent than the wealthy, may be due to the high availability of non-skilled wage labor jobs. Even though wages may be low, non-skilled, nonagricultural labor work can reduce vulnerability to poverty by freeing villagers from the seasonal constraints of agriculture, allowing regular year-round work and income (Rigg, 2006b). The protecting and raising nature of non-farm activities means that deagrarianization, in any form, is closely linked to depeasantization (Bouahom et al., 2004; Rigg, 2005). While Rigg (2006a) proposed a de-linking between wealth and land due to the increase in non-farm activities, the prevailing importance of agriculture in rural southern Laos was evident in this study. Rice land ownership was considered the most important determinant of wealth status (Table 3) while some form of agricultural activities was undertaken by almost every household in the study. In addition to undertaking a

(c) Types of livelihood diversification by socioeconomic group and potential barriers (i) Non-farm activities Results showed no significant difference in the proportion of farm and non-farm activities undertaken by different wealth groups, however, richer households undertook a significantly greater total number of non-farm activities than poorer households (p < 0.05) (Figure 5). Ellis and Allison (2004) suggested the nature of diversification differs greatly between better off and poorer households. The better off tend to diversify into non-farm business activities, as found in parts of Africa and Indonesia (Barrett et al., 2006; Freeman, Ellis, & Allison, 2004; Schwarze & Zeller, 2005), reducing their dependence on agriculture. For poorer households, they tend to diversify in the form of casual wage work such as off-farm labor and are often left highly dependent on agriculture (Reardon, 1997). This may be attributable to the inability of the poor to overcome important entry bar-

Figure 5. Number of non-farm household livelihood activities in each wealth group.

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

10

WORLD DEVELOPMENT

Figure 6. Number of farm-based household livelihood activities in each wealth group.

higher number of non-farm activities than the poor, richer households also undertook a higher total number of farm activities (Figure 6). Multiple comparisons between groups using Tukey contrasts indicated that the difference between rich and middle wealth groups was not significant, the difference between poor and middle groups was significant (p < 0.05) and the difference between rich and poor groups was highly significantly different (p < 0.001). This suggests that despite the increase in non-farm activities in the region, livelihoods and welfare are still very much dependent on agriculture and linked to land ownership (in terms of both quantity and quality), across all socioeconomic groups. Agriculture is clearly still crucially important for households with all types of capabilities, resource access, and options and has not just been left to the ‘‘stranded” poor but is also actively pursued by the better off. The combination of both farm and non-farm activities may be highly complementary, providing positive reinforcement. The high seasonality of agriculture means that there is a long labor for rice

which is not realized until the end of the season. Income from non-farm sources in the form of liquid cash may be important for the timely purchase of farm inputs such as irrigation pumps, or through the ability to hire wage labor, leading to improved cultivation practices and higher farm productivity (Ellis, Kutengule, & Nyasulu, 2003; Ellis & Mdoe, 2003). Nevertheless, the lower uptake of different farming activities by the poor may indicate that they are also constrained in the development of their farming activity portfolio. This has been observed in many other developing rural economies. Reardon et al. (2000), suggested land-poverty is another type of barrier to non-farm employment, constraining activity by limiting the capacity to borrow, and by limiting cash revenues from farming needed to start non-farm businesses or support migration; and Smith et al. (2005), postulated that agriculture is likely to be the main driver in the creation of alternative employment opportunities and microenterprise. The results here illustrate the greater opportunities for richer households which are dominant in both agricultural and non-agricultural sectors of employment. (ii) Remittances Livelihood diversification in southern Laos also had a spatial component with a large number of households receiving part of their income from migrant remittances. The destination of many migrant workers was heavily concentrated in neighboring Thailand, likely due to the close proximity of Savannakhet Province and the recently constructed ‘‘friendship bridge” between the two countries. A higher proportion of wealthier households received migrant remittances than poorer households (Figure 7a) with a significant difference between middle and richer households combined compared with poorer households (p < 0.01). As the majority of households receiving a migrant remittance in Savannakhet are wealthy, migration is more likely to be encouraged by wealth and prosperity than driven by poverty. The probability of a household receiving a remittance also increased significantly with increasing wealth (p < 0.05), however the level of education and area of land owned were both removed as nonsignificant using nested Chi-squared tests (Figure 7b). This suggests that while poverty might therefore be a barrier to migration, it is less likely that education and land ownership

Figure 7. (a) Proportion of households receiving a remittance in each wealth category (95% CIs); (b) families with migrant income and those without compared to wealth.

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS

11

for the poor to lower-skilled, lower-paid employment prospects. As with agricultural and non-agricultural diversification, opportunities to diversify through migration are also unequal. 4. CONCLUSIONS

Figure 8. Wealth status of household and amount of migrant remittance received in the previous year (four outliers omitted), n = 136.

are directly related. Nevertheless, the low variability in education level of the population under study (Table S6) could be the reason for the lack of any apparent trend. Comparison of the value of remittance received among wealth groups indicated that the higher remittances were received by richer households. Differences between groups were all highly significant (p < 0.001) (Figure 8). While higher levels of migration among the rich may be viewed as either a cause or effect of wealth, positive correlation between the value of remittance a household received and wealth suggests it is more likely that migration fuels prosperity rather than being driven by it, albeit an initial level of wealth is usually required to initiate migration. These findings conflict with results from Uganda, where remittances and other transfers are important for the livelihoods of the poor but hardly feature for the relatively well-off (Ellis & Bahiigwa, 2003). While remittances were lower, they were still present in the income of poorer households, potentially due to the relative ease of access of neighboring Thailand and the availability of low-skilled jobs there. This suggests there is a difference in the type of migration undertaken by different wealth groups. Wealth is therefore not likely to be a barrier to all migration, but may limit the type of migrant opportunities

Livelihoods in rural southern Laos are highly dependent on agriculture which remains an important cornerstone of the rural economy. Livelihoods are highly diverse and occupational multiplicity is a progressive strategy undertaken by wealthier households with a diverse asset base and associated with accumulation and advancement. Diversification strategies commonly include a combination of agricultural and non-agricultural activities as well as migrant remittances across all socioeconomic groups, however, the form this takes is also dependent on wealth. The increased asset status of households extends to enabling higher income-generating migration opportunities and may facilitate transition into non-agricultural employment, further strengthening livelihoods by reducing the inherent risks associated with agriculture, as well as into a wider variety of agricultural employment activities. The effect of this is that it leaves the poor, with lower levels of diversification, at most risk to natural or economic disasters or other shocks (Thulstrup, 2015). Rural development is central to the development and poverty eradication efforts of the Laotian government. The development strategy outlined in the Eighth National SocioEconomic Development Plan includes a focus on expanding and modernizing agricultural production and developing value-added goods (MPI, 2015), however, results from this study suggest that recognizing the positive aspects of livelihood diversification for rural poverty reduction and economic development are important. Poverty reduction policy in rural southern Laos needs to be broadened and address mechanisms for supporting and encouraging diversification and mobility rather than focus on one sector. Likewise, pro-poor development initiatives that focus on increasing the diversity of assets (rather than the quantity of any one single asset) of the poor are more likely to be successful in supporting livelihood diversification and reducing vulnerability.

REFERENCES Allison, E. H., & Ellis, F. (2001). The livelihoods approach and management of small-scale fisheries. Marine Policy, 25, 377–388. Anderson, E., & Deshingkar, P. (2005). Livelihood diversification in rural Andhra Pradesh, India. In F. Ellis, & H. A. Freeman (Eds.), Rural livelihoods and poverty reduction policies (pp. 408). Oxon: Routledge. APCAS (2012). Result of analysis on Lao agricultural census 2010/11. In 25th session of the Asia and Pacific Commission on agricultural statistics (APCAS 25). Lao Statistics Bureau. Barrett, C. B., & Reardon, T. (2000). Asset, activity, and income diversification among African agriculturalists: Some practical issues. Project report to USAID BASIS CRSP. Barrett, C. B., Marenya, P. P., Mcpeak, J., Minten, B., Murith, F., Oluoch-Kosur, W., ... Wangila, J. (2006). Welfare dynamics in rural Kenya and Madagascar. Journal of Development Studies, 42, 248–277. Bebbington, A. (1999). Capitals and capabilities: A framework for analyzing peasant viability, rural livelihoods and poverty. World Development, 27, 2021–2044. Be´ne´, C., Mindjimba, K., Belal, E., Jolley, T., & Neiland, A. E. (2003). Inland fisheries, tenure systems and livelihood diversification in Africa:

The case of the Yae´re´ floodplains in Lake Chad Basin. African Studies, 62, 187–212. Birthal, P. S., Roy, D., & Negi, D. S. (2015). Assessing the impact of crop diversification on farm poverty in India. World Development, 72, 70–92. Bouahom, B., Douangsavanh, L., & Rigg, J. (2004). Building sustainable livelihoods in Laos: Untangling farm from non-farm, progress from distress. Geoforum, 35, 607–619. Bruge`re, C., Holvoet, K., & Allison, E. H. (2008). Livelihood diversification in coastal and inland fishing communities: Misconceptions, evidence and implications for fisheries management Working paper, Sutainable Fisheries Livelihoods Programme (SFLP). Rome: FAO/DFID. Bush, S. R. (2004). A political ecology of living aquatic resources in Laos (Ph.D. thesis). Sydney: School of Geosciences, University of Sydney, 348 pp. Chianu, J. N., Ajani, O. I. Y., & Chianu, J. N. (2008). Livelihoods and rural wealth distribution among farm households in western Kenya: Implications for rural development, poverty alleviation interventions and peace. African Journal of Agricultural Research, 3, 455–464. Chinvanno, S., Souvannalath, S., Lersupavithnapa, B., Kerdsuk, V., & Thuan, N. (2008). Strategies for managing climate risks in the lower

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

12

WORLD DEVELOPMENT

Mekong river basin: A place-based approach. In N. Leary, I. Burton, J. Adejuwom, J. Kulkarni, V. Barros, & R. Lasco (Eds.), Climate change and adaptation. London: Earthscan. Cinner, J., Daw, T., & Mcclanahan, T. R. (2008). Socioeconomic factors that affect artisanal fishers’ readiness to exit a declining fishery. Conservation Biology, 23, 124–140. Cinner, J., Mcclanahan, T. R., & Wamukota, A. (2010). Differences in livelihoods, socioeconomic characteristics, and knowledge about the sea between fishers and non-fishers living near and far from marine parks on the Kenyan coast. Marine Policy, 34. Claridge, G. (1996). An inventory of wetlands of the Laos. IUCN, The World Conservation Union. Cordova, A. (2008). Methodological note: Measuring relative wealth using household asset indicators. In M. A. Seligson, & E. Zechmeister (Eds.), Americas Barometer Insights. 6 (pp. 9). Vandetbilt University. Coulthard, S. (2008). Adapting to environmental change in artisanal fisheries – Insights from a south Indian lagoon. Global Environmental Change, 18, 479–489. Dorward, A. (2002). Understanding small stock as livelihood assets: Final technical report. In Understanding small stock as livelihood assets: Indicators for facilitating technology development and dissemination (pp. 7). London: Imperial College, ZC0167. Dorward, A., Anderson, S., Nava, Y., Pattison, J., Paz, R., Rushton, J., & Sanchez Vera E. (2002). Hanging in, stepping up and stepping out: Livelihood aspirations and strategies of the poor. Livestock Production Programme: R7823. DFID, Imperial College. Ellis, F. (1998). Household strategies and rural livelihood diversification. Journal of Development Studies, 35, 1–38. Ellis, F. (2000). Rural livelihoods and diversity in developing countries. Oxford: Oxford University Press, 273 pp. Ellis, F., & Allison, E. H. (2004). Livelihood diversification and natural resource access Livelihoods diversification and enterprise development sub-programme. LPS working paper 9. Rome: Food and Agricultural Organisation of the United Nations, 46 pp. Ellis, F., & Bahiigwa, G. (2003). Rural livelihoods and poverty reduction in Uganda. World Development, 31(6), 997–1013. Ellis, F., & Freeman, H. A. (2004). Rural livelihoods and poverty reduction strategies in four African countries. Journal of Development Studies, 40(4), 1–30. Ellis, F., Kutengule, M., & Nyasulu, A. (2003). Rural livelihoods and poverty reduction in Malawi. World Development, 31(9), 1495–1510. Ellis, F., & Mdoe, N. (2003). Rural livelihoods and poverty reduction in Tanzania. World Development, 31(8), 1367–1384. Evans, H. E., & Ngau, P. (1991). Rural-urban relations, household income diversification and agricultural productivity. Development and Change, 22, 519–545. Filmer, D., & Pritchett, L. H. (2001). Estimating wealth effects without expenditure data – Or tears: An application to educational enrollments in states of India. Demography, 38, 115–132. Freeman, H. A., Ellis, F., & Allison, E. (2004). Livelihoods and rural poverty reduction in Kenya. Development Policy Review, 22(2), 147–173. Freeman, H. A., & Ellis, F. (2005). Synthesis and policy implications. In H. A. Freeman, & F. Ellis (Eds.), Rural livelihoods and poverty reduction policies. Oxon: Routledge. Garaway, C. (1999). Small waterbody fisheries and the potential for community-led enhancement: Case study in Laos (Ph.D. thesis). London: Centre for Environmental Technology, Imperial College, 414 pp. Haggblade, S., Hazell, P., & Brown, J. (1989). Farm-nonfarm linkages in rural sub-Saharan Africa. World Development, 17, 1173–1201. Hargreaves, J. R., Morison, L. A., Gear, J. S. S., Kim, J. C., Makhubele, M. B., Porter, J. D. H., ... Pronyk, P. M. (2007). Assessing household wealth in health studies in developing countries: A comparison of participatory wealth ranking and survey techniques from rural South Africa. Emerging Themes in Epidemiology, 4. Houweling, T. A., Kunst, A. E., & Mackenbach, J. P. (2003). Measuring health inequality among children in developing countries: Does the choice of the indicator of economic status matter? International Journal for Equity in Health, 2, 8. Howe, L. D., Hargreaves, J. R., & Huttly, S. R. A. (2008). Issue in the construction of wealth indices for the measurement of socioeconomic position in low-income countries. Emerging Themes in Epidemiology, 5.

Iiyama, M. (2006). Livelihoods diversification patterns among households and their implications on poverty and resource use: A case study from a Kerio river basin community The land use change, impacts and dynamics project. Working paper number 51. Tokyo: University of Tokyo, 31 pp. Iiyama, M., Kariuki, P., Kristjanson, P., Kaitibie, S., & Maitima, J. (2008). Livelihood diversification strategies, incomes and soil management strategies: a case study from Kerio valley, Kenya. Journal of International Development, 20, 380–397. Lohmann, C., & Liefner, I. (2009). Location, non-agricultural employment and vulnerability to poverty in rural Thailand. Erdkunde, 63, 141–160. Marschke, M. J., & Berkes, F. (2006). Exploring strategies that build livelihood resilience: A case from Cambodia. Ecology and Society, 11, 42. Martin, S. M., Lorenzen, K., Arthur, R. I., Kaisone, P., & Souvannalangsy, K. (2011). Impacts of fishing by dewatering on fish assemblages of tropical floodplain wetlands: A matter of frequency and context. Biological Conservation, 144, 633–640. Martin, S. M., Lorenzen, K., & Bunnefeld, N. (2013). Fishing farmers: Fishing, livelihood diversification and poverty in rural Laos. Human Ecology, 42(5), 737–747. Mckenzie, D. J. (2003). Measuring inequality with asset indicators. Journal of Population Economics, 18, 229–260. Montgomery, M. R., Gragnolati, M., Burke, K. A., & Paredes, E. (2000). Measuring living standards with proxy variables. Demography, 37, 155–174. MPI (2015). Five year national socio-economic development plan VIII (2016–2020). Ministry of Planning and Investment, Vientiane Capital, 187 pp. Mukherjee, N. (1993). Participatory methods and rural knowledge. In N. Mukherjee (Ed.), Participatory rural appraisal: Methodology and applications (pp. 160). New Delhi: Concept Publishing Company. Niehof, A. (2004). The significance of diversification for rural livelihood systems. Food Policy, 29, 321–338. Pauly, D. (2006). Major trends in small-scale marine fisheries, with emphasis on developing countries, and some implication for the social sciences. Maritime Studies, 4, 7–22. Perz, S. G. (2005a). The effects of household asset endowments on agricultural diversity among frontier colonists in the Amazon. Agroforestry Forum, 63, 263–279. Perz, S. G. L. (2005b). The importance of household asset diversity for livelihood diversity and welfare among small farm colonists in the Amazon. Journal of Development Studies, 41, 1193–1220. Reardon, T. (1997). Using evidence of household income diversification to inform study of the rural nonfarm labor market in Africa. World Development, 25(5), 135–741. Reardon, T., Taylor, J. E., Stamoulis, K., Lanjouw, P., & Balisacan, A. (2000). Effects of non-farm employment on rural income inequality in developing countries: An investment perspective. Journal of Agricultural Economics, 51, 266–288. Rigg, J. (2005). Living with transition in Laos: Market integration in Southeast Asia. London and New York: Routledge. Rigg, J. (2006a). Forests, marketization, livelihoods and the poor in the Laos. Land Degradation and Development, 17, 123–133. Rigg, J. (2006b). Land, farming, livelihoods and poverty: Rethinking the links in the rural south. World Development, 34, 180–202. Rigg, J. (2007). Moving lives: Migration and livelihoods in the Laos. Population, Space and Place, 13, 163–178. Salas, S., & Gaertner, D. (2004). The behavioural dynamics of fishers: Management implications. Fish and Fisheries, 5. Schwarze, S., & Zeller, M. (2005). Income diversification of rural households in Central Sulawesi, Indonesia. Quarterly Journal of International Agriculture, 44, 61–73. Smith, L. E. D., Nguyen Khoa, S., & Lorenzen, K. (2005). Livelihood functions of inland fisheries: Policy implications in developing countries. Water Policy, 7, 359–383. Sokal, R. R., & Rohlf, F. J. (1969). Biometry: The principles and of statistics in biological research. New York: W. H. Freeman and Co, 887 pp. Thulstrup, A. W. (2015). Livelihood resilience and adaptive capacity: Tracing changes in household access to capital in central Vietnam. World Development, 74, 352–362.

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018

LIVELIHOOD DIVERSIFICATION IN RURAL LAOS Valdivia, C., Dunn, E., & Jette´, C. (1996). Diversification as a risk management strategy in an Andean agropastoral community. American Journal of Agricultural Economics, 78(5), 1329–1334. Vyas, S., & Kumaranayake, L. (2006). Constructing socio-economic status indices: How to use principal components analysis. Health Policy Plan, 21, 459–468. Wilen, J. E., Lockwood, D., & Botsford, L. (2002). Avoiding surprises incorporating fisher behaviour into management models. Bulletin of Marine Science, 70, 553–575.

13

APPENDIX A. SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.worlddev.2016.01.018.

Available online at www.sciencedirect.com

ScienceDirect

Please cite this article in press as: Martin, S. M., & Lorenzen, K. Livelihood Diversification in Rural Laos, World Development (2016), http://dx.doi.org/10.1016/j.worlddev.2016.01.018