Subjective questions on welfare, such as satisfaction with particular welfare dimensions, are increasingly being used to measure changes in well-being during crises. Although subjective questions on welfare have well-known limitations, it is possible that analyzing changes in response to large events can address some of these concerns. However, this paper illustrates a new difficulty in interpreting changes in such measures in tumultuous circumstances. Specifically, crises can impact both the scale with which households report their subjective assessments and the behaviors from which deprivations are traditionally inferred, and it is unclear how well subjective measures align with traditional welfare metrics in such a setting. This paper demonstrates the importance of this issue following the onset of the conflict in the Republic of Yemen, which caused large declines in nearly all traditional measures of well-being. However, the findings show that households reported a large increase in satisfaction in the same welfare dimensions where deprivations increased.