Social experiments are powerful sources of information about the effectiveness of interventions. In practice, initial randomization plans are almost always compromised. Multiple hypotheses are frequently tested. "Significant" effects are often reported with p-values that do not account for...

Older women's patterns of labor supply over the past forty years have differed markedly from those of younger women. Their labor force participation declined sharply during a period of rapid increase for younger women, and then increased significantly while younger women's plateaued and even...