In this study, we examined the effects of welfare and the measurement of social overhead capital for a household. In Japan, a decline in population and the structural straits of public finance give rise to difficulties in maintenance and replacement of social overhead capital that was invested during the period of high economic growth and the subsequent period of fiscal stimulus during the 1990's. The importance of basic services provided by infrastructural services such as public utility services or flood control is increasing because of transformation of regional structures caused by a decrease in population and continual concern for disasters.
To solve the trade-offs between the financial straits for public investment and the increasing needs for safety provided by social overhead capital, we show a model for the evaluation of welfare vulnerability handled by the infrastructure and examine the provision of municipal water. We adopt the concept of “vulnerability” for the model rather than “risk”, because the former describes not only the stochastic state and severity of loss people experience but also the capability of households to reduce risk and the minimum services they receive. In the model, we suppose that households make a choice under ex ante expectations for the probability of a risk event, such as water cut off resulting from drought or facility accidents, while they realize ex post that they should have other expectations. The differences between ex ante and ex post household expectations create vulnerability. The hypothesis of the model for household cognition of risk events and the subsequent choices need to be inspected. We carried out a preliminary questionnaire to propose a process of inspection for municipal water cutoff expectations of households, and their choice under ex ante probability and ex post probability for the loss they may suffer.
The summary of our study follows. First, we show the welfare evaluation model for vulnerability that a household faces. Second, the welfare measurement of vulnerability is developed especially from the viewpoint of theoretical appropriateness of the welfare measure. Third, preliminary questionnaires and their inspection show partially the validity of the application of the model to municipal water cutoff risks and conditions required for evaluation.

JEL Classification: H54, H43, R53, I38