First, reconnecting job quality to economic growth will likely require big steps. The policies simulated in this paper would all qualify as major policy initiatives, yet none would create a sufficient number of good jobs to employ even half of the U.S. workforce.

 Second, eliminating bad jobs appears to be easier than creating good jobs Most of the proposals examined especially universal programs such as universal health care or a universal retirement plan do more to reduce the share of bad jobs than they do to increase the share of good jobs. (Our

classification system divides jobs into three categories: good jobs, bad jobs, and jobs that fall in between.)

 Third, a combination of complementary policies appears to be significantly more effective than if any one of the policies is enacted on its own. Separate implementation of a universal retirement plan or health insurance would both greatly boost the share of workers in good jobs, but the

simultaneous implementation of both policies woul d raise the good jobs share by more than the sum of the two distinct policies.

Fourth, gender pay equity would go a substantial way towards eliminating the large good jobs gap between men and women. By our calculations, a policy of pay equity for women and men with the same educational qualifications would reduce the gender good jobs gap by about 90 percent.

 Finally, increasing unionization appears to be substantially more effective than a comparable expansion of college attainment. Given that increasing college attainment is a long and expensive process, these findings suggest the importance of emphasizing unionization as much or more than college attainment as a key path to improving job quality.