We use novel quasi-experimental variation to (i) test whether firm-specific demand shocks impact wages, and (ii) to disentangle predictions coming from wage bargaining and firm upward sloping labor supply curve (wage posting). We use a unique ...