New DP: econpapers.repec.org/paper/zbwi4r... Which AER papers were robust (in terms of the % of robustness checks significant at 5%) was predictable by our survey participants, but the survey participants were overly optimistic and overestimated the degree of robustness reproducibility