TY  - THES
T1  - Analysis of factors driving differences in intensification and income from agriculture among smallholder farmers in Northern Vietnam
A1  - Ufer,Susanne
Y1  - 2022/06/22
N2  - Challenged by difficult topography, remoteness, high ethnic diversity, low levels of infrastructure, high poverty, and high dependency on upland farming systems the well-being and incomes of ethnic minorities and the poorest in the Northern Uplands of Vietnam, like in many upland areas in Southeast Asia, are still predominantly linked to agricultural productivity. At the same time agricultural commercialization, the introduction of modern agricultural technologies, higher input use, and stronger agricultural specialization have increased the demands on households’ ability to adequately invest in agricultural intensification and to protect themselves from agricultural income risks. In the Northern Uplands, one of these extensively cultivated, highly commercialized, and highly specialized upland crops is maize. A crop that needs high levels of inputs and is predominantly grown for cash income from poor and non-poor farmers alike. 
Given the low asset levels of households, it is therefore of particular interest which differences in challenges farmers may face regarding the improvement of maize production dependent on their wealth level. Yet, no detailed research exists that analyses how the level of asset endowments with natural, human, physical, financial, and social capital as well as risk aversion impact technology adoption, maize intensification, and maize productivity of farmers of different household wealth. This doctoral thesis seeks to fill these knowledge gaps by investigating the following research topics: (1) the level and short-term changes in agricultural input use and productivity in maize production by household wealth, (2) the impact of risk aversion on fertiliser use in maize production by household wealth, and (3) the impact of household asset levels and the return to assets on productivity differences in maize production between households of different wealth.
Research analysis builds on a quantitative dataset collected from a random panel sample of 300 rural households in Yen Chau district, Son La province, in the Northern Uplands of Vietnam in the period from 2007 to 2010. The research area is relatively poor, ethnic diverse, with a high dependency on upland agriculture, crop income, and maize income in particular. Econometric analysis is carried out firstly by organizing households by wealth through a composite asset-based indicator derived from principal component analysis (PCA) and secondly by applying extensive descriptive analysis, regression analysis, and econometric decomposition-based techniques to the so differentiated dataset. 
Results from the first research topic (1) “Level and short-term changes in agricultural input use and productivity in maize production by household wealth” show that average numbers on maize input use, maize productivity, and maize income hide important wealth-related differences. While adoption rates of modern maize seeds and mineral fertilisers are widespread and very similar by wealth terciles over time, input use intensity of fertilisers, maize yields, and maize incomes differ significantly between the poorest tercile and the wealthier maize farmers. While a substantial share of the poorest household tercile uses fertiliser quantities well below recommended levels, households of the middle and wealthiest terciles are about twice as likely to apply fertilisers according to average or above-average fertiliser recommendation levels. Moreover, between approximately one-tenth and one-quarter of households from all wealth terciles overuse fertiliser, too. The poorest tercile further buys despite the lower use of fertilisers more often seed and fertiliser inputs on-loan than households from higher wealth terciles. Consequently, the poorest households have to pay relatively higher input costs at otherwise mostly similar market prices for seed and fertiliser inputs and maize output. Yield, input price, and output price risks are high in the research area for all households. However, the poorest tercile of households suffers somewhat more from risks due to fluctuations in yield, output price, and maize income.
Results from the second research topic (2) “Impact of risk aversion on fertiliser use by household wealth” show that the fertiliser quantity applied to maize is affected by the risk aversion of the household head in the poorest one-third of households and not affected by the risk aversion of the household head of the wealthier households. The results remain valid when different empirical risk aversion measures (i.e. self-assessment scale and lottery game) are considered, when instead of total fertiliser quantity, the quantity of NPK, urea, or total nitrogen are considered, as well as when different measures of household wealth are considered (i.e. asset-based wealth index, household per-capita income, and household per-capita consumption expenditure).
Results from the third research topic (3): “Impact of household asset levels and the return to assets on productivity differences in maize production between households of different wealth” show that quantity-based assets effects are more important for the size of the maize income gap per hectare between the poorest one-third of maize farmers and maize farmers of higher wealth than the return-based assets effects. Quantity-based asset effects significantly account for more than two-thirds of the entire maize income gap, while return-based assets effects are on the contrary not statistically significant at all. From the quantity-based assets effects, credit limit, ethnicity of the household head, and upland land value have large and significant effects. Farm size, value of buffalo and cattle, and household head age have smaller and less constant effects. Credit limit is the most important and most consistent driver by size, showing that financial access for poor minority households is still a defining obstacle to agricultural productivity. Ethnicity of the household head has a strong and positive but decreasing impact on the maize income gap, showing that ethnicity-related soft factors, such as differences in location, agricultural practices, and traditions, should be taken more into acknowledgement. From the significance of the upland land value, we conclude that environmental factors, such as the prevention of soil erosion, the protection of soil fertility, and considerations of long-term sustainability, should gain more emphasis. Other minor factors driving the maize income gap are discussed in more detail in the thesis.  
From the results of the doctoral thesis, we draw the following main conclusions for the development of pro-poor strategies for the improvement of upland agricultural productivity. 
Firstly, agricultural research and extension should pay more attention to identifying bottlenecks households face dependent on their wealth level to avoid overseeing the specific obstacles poor and non-poor farmers face that may result in the perpetuation of poverty traps, increased inefficiency, and the waste of resources. Consequently, more emphasis is needed on research and extension that incorporate wealth-related agronomic, financial, and risk-related aspects, that better account for the efficient use of inputs, and that place increased emphasis on the adaptation of technological innovations to the farm systems and needs of upland minorities.
Secondly, more emphasis should be placed on helping poor households to deal with risks. While all households face yield, input and output price risks, risk aversion affects input intensification decisions in the poorest one-third of households. Hence, especially for the poorest more emphasis should be placed on the expansion of risk management options that account for the lower risk-bearing capacity and higher risk aversion of the poor, and the often higher transaction costs. This includes measures such as improving the financial literacy of households, allowing households to borrow for consumption, linking credit with insurance or saving options, decreasing transaction costs and the complications linked with successfully applying for loans, as well as, if possible, developing innovations like improved seeds or agricultural management systems that have the capacity to lower the downside risks of production.
Thirdly, assets should be considered an important driver of agricultural productivity differences between poor and non-poor farmers. Based on the findings that the quantity-based effects of assets are the major and only significant driver of the maize income gap between the poor and non-poor maize farmers, we conclude that the poorest households can have the same productive agricultural potential as the non-poor if they would have the same level of assets. Supporting households to get access to and accumulate assets should be therefore an important policy goal. This is particularly true about assets that can be influenced by agricultural research, improved agricultural extension, and targeted development policy. Based on our research findings this concerns the improvement of access to credit, the support of ethnic minority households, and efforts to increase the sustainability of upland agriculture. 
Fourthly, investing in the long-term suitability of maize production and upland farming systems is recommended. While maize can be a very profitable cash crop in the short run, the results of this doctoral thesis also show that a strong specialization in maize production may embody multiple short-term and long-term economic as well as sustainability related livelihood risks. This includes the possibly limited extent of the profitability of additional credit and inputs as well as a strong dependency on the quality of natural resources. In consequence, policies should focus on directly improving the conditions for maize productivity as well as improving the conditions for long-term development. Such measures include supporting households to invest more in profitable on-farm and off-farm diversification, the improvement of infrastructure and extension to increase the profitability of unused cropping choices, the development of profitable soil protection measures, and investments in education and the development of off-farm job alternatives.
KW  - Intensivlandwirtschaft, Hochland, Risikoverhalten, Landwirtschaftliches Einkommen, Armut, Vietnam
CY  - Hohenheim
PB  - Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
AD  - Garbenstr. 15, 70593 Stuttgart
UR  - http://opus.uni-hohenheim.de/volltexte/2022/2020
ER  -