In dynamic wage bargaining models it is usually assumed that individual unemployment benefits are a fraction of the average wage level. In most countries, however, unemployment benefits are instead tied to the previous level of individually earned wages. We show how the analysis has to be modified if this fact is taken into account and compare our findings for the wage-setting curve with outcomes under other unemployment compensation schemes. In particular, we show that the widely used vertical wage-setting curve relies on more restrictive assumptions than usually considered. We also demonstrate that a reduction of unemployment benefits of those who get unemployed after the bargaining period leads to higher equilibrium unemployment.