Comment on “Disasters, Household Decisions, and Insurance Mechanisms: A Review of Evidence and a Case Study from a Developing Country in Asia”

Abstract

The paper by Sawada (2017) on disasters and insurance in Asia is a highly ambitious and useful study, which attempts to assess the extent of the mitigation of consumption reduction caused by disasters by means of insurance based on a literature review and a case study in Vietnam. Since the final version of the paper has successfully incorporated many of the comments made on an earlier version, I do not have many comments except for a few concerns. Sawada (2017, p. 18) states in his abstract that “it is imperative to strengthen market, government, community based insurance mechanisms to diversify aggregate disaster risks at individual, national, and regional levels in Asia.” I wonder if this concluding statement can be justified from the present study because there are costs involved in strengthening the insurance mechanisms, whereas the cost side is not analyzed in Sawada's study. It seems to me that in order to justify this argument, the sources of market failures should be discussed and analyzed more explicitly. More generally, Sawada seems to assume implicitly that “full insurance” or “more insurance” is socially desirable. This may or may not be the case because of moral hazard issues. In the real world where information asymmetry is present, moral hazard is a real issue. I expect that the moral hazard would occur rampantly with a full insurance scheme, which actually implies that full insurance would break down. I definitely do not think that full insurance is socially desirable. I am surprised to learn that “the full consumption risk sharing hypothesis holds for the commune level risk sharing network.” This is not the case for larger networks. I simply cannot believe that full insurance at the commune level is achieved unless commune people help one another based on community ties. Sawada points out that there are six mechanisms for coping with risk: (1) consumption reductions; (2) dissaving; (3) borrowing; (4) public transfers; (5) private transfers; and (5) additional labor income. The question is which mechanisms for coping with risk are important and unique at the commune level, but not at the level of larger networks. Intuitively, I thought that it would be private transfers as people would know each other well in a community. However, according to Sawada's (2017) Table 2 private transfers do not respond to unexpected losses. So the puzzle remains as to how the commune achieved full insurance. It seems to me that this issue requires further scrutiny.