Essays on risk attitudes, knowledge, extreme weather, and farmers' behaviors in rural Southeast Asia

Abstract

Northeastern Thailand and Central Vietnam are two regions where pockets of poverty persist despite overall success in poverty reduction on the national level. While there are profound structural differences between Thailand and Vietnam, a common feature for both countries is that government policies promote the migration of rural labor to facilitate growth in the industrial and service sectors. Furthermore, policymakers in both countries have the vision to transform their agriculture towards large-scale farming, following the model of western agriculture. While out-migration from rural areas has taken place and the share in off- and non-farm income in total household income has been growing, the share of agriculture income in many cases is now less than 50 %. To date, labor rather than land (as in the past) is the main income-generating factor. However, agriculture still plays an essential role in the rural areas of these two countries. Farms are still small, and farm sizes almost remained the same over the past decades. Structural transformation of the rural areas as envisaged by policymakers does not take place. Households keep their agriculture as a backup and safety net and hence small-scale farming continues to dominate. At the same time, farmers in the two regions are increasingly exposed to severe weather events caused by climate change which makes them vulnerable to poverty and food insecurity (ADB, 2009; IPCC, 2014a; Blanc & Reilly, 2017). In this study, it is therefore aimed to obtain a better understanding of farmers' decision-making in agriculture. In particular, the thesis aims to investigate how farmer knowledge and skills and their risk attitudes, on the one hand, and the increasingly occurring extreme weather events, on the other hand, influence their decision-making with regard to farm management decisions. There are three specific research questions to be answered in this research: (1) how do risk attitudes affect household decision-making; (2) what is the impact of agricultural knowledge on agricultural production; (3) how do farmers manage their agricultural inputs in response to extreme weather events. To answer these questions, the thesis draws on two primary data sources. The first is the database of the Thailand Vietnam Socio Economic Panel (TVSEP) project during the period of 2007 to 2017, i.e., six-year panel dataset was collected from some 4,400 rural households in the Northeastern Thailand provinces of Nakhon Phanom, Ubon Ratchathani, and Buri Ram; and in Vietnam’s Central Coastal and Central Highlands provinces of Ha Tinh, Thua Thien Hue (Hue), and Dak Lak. The second data source is historical weather data. We use the monthly high-resolution (0.5) temperature and precipitation data observed from 1948 until 2016 from the Global Precipitation Climatology Centre (GPCC; Schneider et al., 2018), and the Global Historical Climatology Network Monthly - Version 2 and the Climate Anomaly Monitoring System (GHCN + CAMS; Fan & Dool, 2008), respectively. The results of the thesis are presented in three essays. The first essay is “Risk attitudes and implication for livelihoods strategy – evidence from two provinces in Thailand and Vietnam.” Utilizing an Ordinary Least Square (OLS) and a Probit regression with different alternative specifications, the determinants of risk preference for household decision-making are analyzed. Results show that risk attitudes are significantly related to individual characteristics such as age, gender, height, and household wealth. There are correlations between the willingness to take risk and real-life decisions of farm households. The findings show that risk-seeking individuals likely diversify income-generating activities as a cushion against the risk of small-scale farmers in these areas. They invest in self-employment and other non-farm enterprises while still capitalizing in agriculture.

The second essay, named “Farmers’ knowledge and farm productivity in rural Thailand and Vietnam”, investigates the relationship between farmers’ knowledge, skills, and agricultural productivity. This paper uses primary data on agricultural knowledge and skill tests among “TVSEP households” in the provinces of Ubon Ratchathani in Thailand and Hue in Vietnam. A Two-Stage Least Squares (2SLS) approach combining knowledge and skills test results with productivity data of later waves was developed to identify the effects of agricultural knowledge and skills on agricultural productivity. The major finding is that technical knowledge in agriculture is significantly and positively associated with profit but significantly negative with rice yields and cost of input costs. This suggests that knowledgeable farmers may strive for optimal rather than maximum yield and are more judicious in the use of inputs which is good for the economy and the environment.

In the third essay, named “Extreme weather and agricultural input management in rural Thailand and Vietnam: Intensify or de-intensify?” we investigate the impact of extreme weather events, namely drought, on household input management decisions in Northeastern Thailand and Central Vietnam. Eight inputs are captured: land, labor (household labor and hired labor), chemicals (i.e., fertilizer and pesticides), irrigation, machinery, and other agricultural investments. We define two binary drought indicators, namely severe drought and extreme drought, using the Standardized Precipitation Evapotranspiration Index (SPEI) as the criterion. We then use Fixed Effects (FE) model for this paper’s purpose. Results indicate that farmers tend to de-intensify agricultural production in terms of hired labor, pesticides, number of crops grown, and agricultural investments in response to severe droughts. Second, farmers increasingly hire machinery as a substitute for owned equipment and for household labor. Third, the magnitude of the effects increases with the severity of drought. Differentiating the analysis between countries, and upland versus lowland rice production, shows that the level of de-intensification varies. For example, Thai farmers allocate more family and hired labor to agricultural production; Vietnamese farmers invest in agricultural assets. Upland rice farmers focus on several inputs such as pesticides, machinery, and agricultural assets, while lowland farmers focus on available irrigation systems.

All three essays have generated important policy messages for Governments in both countries to consider public support measures to strengthen rural households coping strategies toward extreme weather events and climate change.