"The Journal of Labor Research 20th anniversary symposium review of What Do Unions Do? offers a unique opportunity to examine how the claims made in the book have fared in ensuing research and to ponder what parts of the book I would change if I could. This paper responds to the 18 critical essays in the journal. It recognizes three major errors of omission: failure to take account of unionism outside the US; failure to analyze public sector unionism; and failure to analyze the effects of unionism on economic growth; and the problem of determining the "optimal level of unionism" on the basis of estimates of what unions do. Ensuing research has found that What Do Unions Do? correctly identified union effects on turnover, fringe benefits, earnings inequality, political action, profits, managerial flexibility and human resource management, and that wage effects vary widely. Estimates of the union effect on productivity tend to be positive but modest, ruling out negative effects on average, but not conclusively establishing positive effects. Critical comments from some of the symposium panelists notwithstanding, I believe that the bulk of the evidence supports the What Do Unions Do? claim that management opposition has been a major factor in the decline in union density in the US"--National Bureau of Economic Research web site.