Essays on vulnerability to climate change and child malnutrition in Southeast Asia

Abstract

Southeast Asia is considered as one of the most vulnerable regions to the downside effects of climate change. Extreme weather events such as droughts, floods, unusual heat and rainfall, tropical cyclones have been more frequent and intense in this region. In the meantime, the region’s heavy reliance on agriculture for local residents’ livelihood makes it highly vulnerable to these climate-related risks. Under such circumstances, the perceptions of climate change and adaptation choices made by farmers are important considerations in the design of adaptation strategies by policy makers and agricultural extension services. Furthermore, the exploration for an efficient allocation of farm’s bounded resources among crop and livestock production activities under climate risks is also of a great importance. Southeast Asia has witnessed a rapid economic growth and structural transformation over the last decades. This has helped millions of people move out of poverty. Nevertheless, poverty incidence reflected by the proportion of population living under the $1.25 remains high in the region and unfortunately, the poor are the most vulnerable to climate change. Beside climate risks, child malnutrition is another development challenge confronting Southeast Asia. As a matter of fact, the reduction of monetary poverty does not necessarily translate into a decrease in the malnutrition rates. Insights into the correlation between monetary poverty and child nutritional outcomes might be gained through a comparison between rural and peri-urban areas. In this context, the thesis comprising three essays aims to examine the climate-related risks confronting farmers, perceptions of and adaptation to changing climate, optimal resource allocation and child nutritional outcomes under such risky climate in Southeast Asia. The specific research objectives are: (i) to investigate the farmers’ perceptions of and adaptations to climate change in Thailand and Vietnam; (ii) to identify the farmer’s efficient adjustments of production activities to climate-related risks in Thua Thien Hue province in Vietnam and (iii) to examine the relationship between monetary poverty and child nutritional outcomes with rural versus peri-urban context in three neighboring countries - Thailand, Lao PDR and Vietnam. To do so, the thesis draws on the five-year panel data set of around 4,400 non-urban households in Thailand and Vietnam. The surveys were conducted in 2007, 2008, 2010, 2011 and 2013 in three provinces in the Northeastern Thailand and three provinces in the North Central Coast and Central Highlands of Vietnam. The targeted provinces are assumed to have less developed infrastructural conditions and their inhabitants are likely to be vulnerable to poverty. The survey instruments comprise of household questionnaire including questions on household livelihoods and village questionnaire asking physical and social infrastructure of the community. Beside this main dataset, data from an add-on project in Savannakhet province, Lao PDR in 2013 and 2014 are additionally used in the third essay. The first essay explores the farmers’ perceptions of and adaptations to climate change of rural households in Thailand and Vietnam. The essay, first, investigates the climate-related shocks that farm households experienced, whether they do perceive a change in the longer term climate conditions and which indicators they use to describe climate change. This objective is tackled by means of a descriptive analysis. Second, the essay examines the determinants of the farmers’ perceptions of climate change and their decision to adjust the agricultural production by means of Heckman selection model. Last, the drivers of different adaptation measures taken by farmers are explored by Multinomial Logit model. Results show that the majority of farm households in both countries have experienced climate-related shocks, do recognize climate change and have their own way of describing this phenomenon. Also, the results point out that farmers’ perceptions of climate change are engraved by climate related shocks and in Vietnam are additionally shaped by respondent’s characteristics and location variables. Finally, the factors that drive adaption measures differ among practices, provinces and countries. They are also found in characteristics of respondent and household head. The perhaps most important factors in explaining specific adaptation measures are the three specific climate variables namely the perceptions of rainfall, temperatures and wind. These results can provide important information to policy makers and agricultural extension services who should improve their understanding of the farmers’ interpretation of climate change and the constraints that have so far prevented them from undertaking more and better adaption measures. The effects of climate-related shocks are quantified in the second essay through the estimation of Cobb-Douglas production function by ordinary least squares (OLS) for major crops in the lowland and upland areas of Thua Thien Hue province in Vietnam. The coefficients of risk dummies in the regression of Cobb-Douglas function are then incorporated in the Target MOTAD mathematical model to identify the farmer’s efficient adjustments of crops and livestock production to climatic risks. Findings reveal that flooding, pestilence, drought, coldness and livestock diseases were the most common climate-related risks confronting the farm households in this province. These climatic extreme events negatively affected crop production as well as livestock production. The vulnerability to risks was, however, different among crops. The results of the mathematical programming show that the risk efficient plans differ from the actual farmer's plan especially in case of the lowland farm. Specifically, the farming pattern which ensures income safety for the lowland farm comprises of rice and corn cropping without livestock while rice, corn, cassava and livestock for the upland farm. The exclusion of livestock out of lowland farm’s optimal plan is due to the high “reduced cost” and large income variability. Since there is a trade-off between income safety and income level in livestock production found in the lowland farm, farmers maybe behave with lower levels of risk aversion as assumed by the model in order to have a chance for a higher income. The mathematical results also disclose remarkably abundant labor but constraints in land and cash. Findings overall suggest that implementation of risk management tools; generation of more off-farm and/or non-farm self-employment; more liberated rental land markets and better veterinary services will help farmers to better cope with negative climate effects. Besides climate risks and climate change, child malnutrition is another big development challenge confronting Southeast Asia. This development issue is discussed in the third essay. In particular, this essay examines whether a decline in monetary poverty automatically results in malnutrition reduction in children under age five in Thailand, Lao PDR and Vietnam. It also investigates the peri-urban and rural differences in poverty and malnutrition alleviation in these countries. The analysis is done by means of descriptive statistics and ordinary least squares as well as a district level fixed-effects regression. The results imply that child nutrition remains a problem despite achievements in poverty mitigation in both peri-urban and rural areas in all these countries. Between the two major nutrition outcome indicators, namely underweight and stunting, the latter is the bigger problem. The “severe” undernutrition is even higher than the “moderate” in all three countries. A key insight is that irrespective of income levels rural children are more stunted and underweight than their urban counterparts. Child health primarily drives child nutritional outcomes in both peri-urban and rural areas. This implies that apart from nutrition access to medical facilities for children are vital for development. Moreover, in peri-urban setting, social and extended family networks play a critical role in childcare; while in the rural areas mother`s nutritional outcomes largely determine child health. Overall, findings suggest that the rural and peri-urban child nutritional gaps need to be bridged. However, different strategies for peri-urban and rural should be adopted to improve child nutritional outcomes. Among others, cost-effective childcare services should be established in peri-urban areas and schemes need to be developed to build medical facilities in rural areas.