Abstract: | We investigate the effect of employment systems on the layoff‐performance relationship. We construct a typology of two types of HPWS (Calculative or “hard” HRM and Collaborative or “soft” HRM) and two non‐high performance systems (Traditional HRM and Low HRM). We use attribution theory as a framework from which to draw hypotheses. We examine survey responses from two waves of panel data. We employ cluster analysis to identify distinctive configurations of employment practices used in UK workplaces. We use the cluster outcomes as explanatory variables in moderator regression analysis. Following layoffs, we find that Calculative workplaces experience lower subsequent performance than Collaborative workplaces. Over the next five years, Calculative and Collaborative workplaces experience equivalent amounts of performance change but Calculative workplaces fail to make a full recovery. |