Community Resilience to Climate Change: Theory, Research and Practice

145 so that they would not be affected by floods like the 2000 flood is determined by their well-being or capacity to cope with floods. Those who can upgrade their house are more resilient to the impacts of floods. The third characteristic of household resilience is about their interest in learning and doing new creative things. This is consistent with the third property of resilience, namely, the capacity to transform and innovate (Folke et al. 2002, Walker et al. 2004, Marschke and Berkes 2006, Marshall and Marshall 2007). However, what are the new creative things in this context? Many farmers developed an innovative way of living with floods by exploiting the flood benefits as well as carrying out flood-based farming activities. These farming practices allow them to improve their household income and create jobs for local laborers to maintain their livelihoods during flood months. These emerging flood-based farming practices not only provide income and food security for rural households, but also help to maintain agricultural sustainability. The perception of floods transformed them from natural disasters into beneficial resources for livelihood development. The subjective well-being approach of measuring households’ resilience is used to reflect the actual capacity of households to cope with flood events. Because resilience is a multidimensional concept, the use of multiple items can help to capture a wide range of factors that contribute to households’ resilience in a specific context. The use of standardized data yielded better underlying factors than the nonstandardized data in this context. While the former approach captured nine items forming three factors, the latter approach only obtained five items forming three factors (Appendix A 1). A standard factor analysis cannot deal with dichotomous variables, but the standardized data approach allowed a factor analysis to be carried out with dichotomous responses. This approach was well validated by using both SPSS and MPLUS software (Appendix A 2). By combining qualitative and quantitative analytical methods, new aspects of resilience among farmers in the MRD have been highlighted. The responses and adaptive behaviors enable some of these farmers to sustain their livelihoods during floods and recover quickly afterward. However, the use of the subjective well-being approach to obtain the perceived capacity at the household level to cope with historical flood events may be prone to some limitations in explaining the resilience of households to future large flood events in the context of climate change. Although item 6 in Table 2 reflects the capacity to evacuate in a future extreme flood event, it does not capture all the dimensions of resilience in future large flood events. For example, with regard to the level of interest in learning and carrying out new flood-based livelihoods, this measure does not take into account future large flood events or very small flood events that may occur due to climate change, which may exceed the capacity of households. This is a key limitation of this research. Therefore, it is important for future research to integrate climate change scenarios into the questionnaires for assessing the resilience of households in the MRD. It is also argued that the determinants of resilience may include demographic, social capital, cultural, economic, and political, aspects of the natural hazards, information, and the geographic setting of places (Gaillard 2007, Norris et al. 2008). For example, information and social networks are very vital for people in disaster prone areas to make behavioral decisions. But these factors may be variable at different scales of analysis. The social capital of households is important for accessing resources to cope with annual flood events. However, some of these factors are not included in the current measures of households’ resilience to floods within the scope of this research. The current measures of households’ resilience were more likely to focus on conceptualizing the concept of resilience in the real context of “living with floods” in the MRD. Some factors such as social capital are often treated as exogenous variables (Narayan and Pritchett 1997). The resilience properties obtained from the factor analysis from this research will be used as latent variables to investigate their relationships with social capital and socioeconomic variables of households in further analysis. At present, the current measures focus on the experience or perceptions of households in coping with past flood events, but do not permit interpretation of the results in the context of climate change. Further study should be carried out to improve the current measures of resilience in future flooding under predicted climate change scenarios.