I evaluate the degree to which different wage-setting mechanisms in labor market search models can fit the aggregate facts on labor’s share. I find that staggered bargaining in nominal wages best allows the model to plausibly match the negative relationship between labor’s share and lagged productivity growth and inflation. I also evaluate the role of labor’s bargaining weight—a low bargaining weight seems plausible but by itself, it cannot generate the patterns observed in the data. Adding a standard sticky-price mechanism to the model actually degrades the match between the model and the data—in the data, labor’s share is countercyclical, while it is procyclical in the sticky-price model. Theory and data both agree that wage stickiness is relevant at the micro and macro levels.