The effects of large-scale, racially-charged violence on labor productivity and racial identity: The riots of 1967–68 and outcomes in Major League Baseball

•Dozens of major race riots occurred in the United States in 1967 and 1968, the vast majority of which affected the Major League Baseball seasons of those years.•Event-study estimations using stacked difference-in-difference methodologies and difference-in-difference estimations with two-way fixed effects show that batters, but not pitchers, saw worsened productivity after major riots near their teams’ cities.•Two-way fixed-effects estimations show strong and significant evidence that pitchers saw worsened productivity after major riots occurred near their birthplaces, but stacked difference-in-difference event-study estimations show only weak evidence of this.•Stacked difference-in-difference event-study estimations show no evidence that riots affected hit-by-pitch rates, but two-way fixed effects estimations show that batters playing for teams whose cities experienced major riots were hit by fewer pitches, with white batters playing for such teams being hit by fewer pitches from white pitchers, suggesting a growing racial identity among white pitchers. Many major race riots occurred in 1967–68. We investigate whether these riots affected individual labor productivity by studying 240,408 plate appearances from Major League Baseball games in those years. Stacked difference-in-difference and two-way-fixed effects difference-in-difference estimations show that batters, but not pitchers, experienced significantly worsened productivity after major riots in their teams’ home cities. In addition, two-way-fixed effects difference-in-difference estimations show evidence of worsened productivity of pitchers after riots near their birthplaces, and that white batters whose teams’ cities had undergone major riots were hit by fewer pitches from white pitchers. However, these latter results are not robust to stacked difference-in-difference estimators. We discuss these differences in the paper.