Due to the absence of unemployment insurance (UI) choices, the traditional approach to estimating the value of UI is to infer it from the observed consumption response to job loss under some assumption on risk preferences. Exploiting the rich data and unique policy context in Sweden, we propose two alternative approaches that relax this assumption and we implement all three methods on the same sample of workers. The first approach considers the difference in marginal propensity to consume (MPC) when unemployed vs. employed, which allows to identify the difference in prices to smooth consumption in the respective states. The second approach exploits UI choices embedded in the Swedish UI system in a Revealed Preference approach. While the drop in consumption expenditures is relatively small (∼ 13 percent), we find that the MPC is around 25 percent higher when unemployed than employed, translating into a marginal value of transfers that is at least 60 percent higher when unemployed than employed. This high value of UI is confirmed by our RP estimates and indicates substantial risk aversion given the relatively small drop in consumption expenditures.