This research attempts to explain the relationship between poverty, livelihood difficulties, risk and risk management and vulnerability to poverty of farm households in Northern Thailand. Furthermore, this study proposes a health insurance concept for reducing idiosyncratic risks and poverty of farm households. The survey underlying this study was conducted in Tambol Pong Yang, Mae Rim, which is a mountainous district of Chiangmai province and is representative of the northern mountainous region of Thailand. Nine villages were interviewed in the study area. Four of the villages where populated by Hmong hill tribes. Data were collected in two types of questionnaires: the first
questionnaire was comprehensive and looked at all socio-economic aspects of the households; the second was related to health insurance. Data were collected for the first questionnaire by interviewing two groups of farm households: the hill-tribe Hmong and a local people known as Khon Muang. The random sample consists of 200 households: 142 local northern and 58 Hmong households. Primary data from the second questionnaire on health insurance was collected in the Mae Rim district. The survey covered 200 households, 146 of which are
Thai and 54 Hmong.
The research started with the analysis of poverty and outlined the assessment of the empirical poverty index, using Principal Component Analysis (PCA). After the significant factors affecting poverty were identified, the next procedure applied participatory rural appraisal (PRA) approaches in order to know how households sustain their livelihoods. Then, individual farm households were examined to analyse how they manage risks. If they managed crises well, they were not exposed to severe consumption and income shocks. Health risks were at the center of interest. Apparently, health insurance can mitigate their risks so that demand and supply of health insurance was analyzed.
Thereafter, a classification of the factors responsible for dynamic poverty and vulnerable households was done. This is important for policy makers to propose appropriate health insurance and poverty reduction policies. The analysis proceeded in six steps:
Firstly, the result of the PCA was utilized to determine the important factors affecting household poverty. Furthermore, a poverty index was
developed. The PCA retained 16 out of 65 possible poverty determining variables. The explicit factors relevant for assessing poverty are the dwelling conditions, assets, human resources, and food security, respectively. The factor which can lead the poor to become even poorer is the human resource factor, where e.g. the number of dependents is comprised. The poverty comparison between farm households living in the highlands and lowlands found that Hmong households, which normally live in mountainous regions, are relatively poorer than the local northern households.
Secondly, PRA was used to explore livelihoods, risk and risk management strategies of farm households in Northern Thailand. Results of the

PRA showed that the most pressing problem plaguing households is their debt. Households try to honor their debt repayment obligations, but it appears that the frequent occurrence of income shocks and their low risk management capacities prevent them from doing so effectively. Land issues relate to the second most important problem area. Another pressing problem negatively influencing households’ livelihoods are droughts, which lead to water shortages, higher fertilizer prices and middleman problems.
Thirdly, the risks which households experienced at different time periods were analyzed as well as the cost of risks, and the risk management strategies of households. Results of the risk and risk management analysis found that there are five major types of risks frequently encountered in rural areas: 1) natural risks; 2) theft risks; 3) production risks; 4) life-cycle risks/human risks; 5) health risks. Risk management strategies can be divided into coping and adaptive strategies. The former refer to short-term coping mechanisms and the latter to long-term adaptive strategies that households use in times of crises to maintain their livelihoods. The most often selected adaptive strategy that households use to manage risks is saving in cash, with 21.9%. Most of the risks (58.8% of all) can be managed within 12 months. However, 25.2% of risks are long-run risks.
Fourthly, this study examined health insurance for the poor in order to provide recommendations for reducing health expenditures at the household level. Respondents reported that the burden of health expenses became lower after they had signed up for state-administrated health insurance, the so-called 30 Baht Health Insurance. However, 42% of the respondents stated that the health expenses still represented a relatively high burden to their household budget. According to the logistic regression analysis of household demand for health insurance, the results indicate that the household health risks variable, the number of times that a household consults a doctor variable, the price satisfaction variable, the accessibility to health insurance information variable and the gender variable have a positive relation to the probability of purchasing health insurance.
Fifthly, conjoint analysis on health insurance aimed at providing concepts for new, alternative health insurance products to support the exiting health insurance system in Thailand. The price premium of 30 Baht (or about $0.8 or €0.6) was the lowest, having been selected in one third of all cased when it occurred. A premium of 60 and 90 Baht was less popular, having been selected 17% and 18% of all times it occurred, respectively.
Finally, the study examined the linkages between poverty and vulnerability to poverty by the classification of a vulnerable group of farm
households. The results demonstrated that 42% of the populations in the study area were poor in 2003 and a significant share of these was chronically poor (11% of the population). Almost one-third of the population is transitorily poor i.e., 30.5% of the total population. It is 43.5% of households are in the vulnerable group, while the rest of households (56.5%) are in the non-vulnerable group.

Appropriate health care policies can play a key role in alleviating health risks and poverty. In order to help the government reducing expenditures to support the health care system for the poor, it would be possible to slightly increase health insurance premiums. The premium of the existing health insurance system should not, however, exceed 60 Baht per doctor visit, and hospitals should increase the number of doctors and health staff so that patients receive faster service.
In conclusion, a better understanding of the relationship between poverty, livelihood difficulties, risk, risk management and vulnerability to poverty of households is important to improve the poverty reduction policies of Thailand. Furthermore, the proposal of a health insurance concept can reduce idiosyncratic risks and poverty of farm households.