Nearly every job created since the last recession has gone to someone with a post-high school education.

Judged as of 2019-05-17, using the best US labour-market data available then on net job creation since the 2007–09 recession (e.g. Georgetown University Center on Education and the Workforce, BLS/CPS), split by education level. 'Post-high school education' means at least some college, including an associate degree or postsecondary certificate. True: about 95% or more of net new jobs since the recession went to workers with education beyond high school. Mostly true: 90–95%, or the share is about 95% or more but the data cover a slightly different period or a measure close to the one stated (e.g. net jobs since the recovery began rather than since the recession). Mixed: about 75–90%. Mostly false: a majority but under about 75%, or the near-total figure holds only for a narrow subset (e.g. good-paying jobs only) and is presented as covering all jobs. False: half or less of net new jobs went to workers with post-high school education.

Set before the research began.

1. E1 America’s Divided Recovery: College Haves and Have-Nots - CEW Georgetown Georgetown University Center on Education and the Workforce

   “those with at least some college education have captured 11.5 million of the 11.6 million jobs created during the recovery.”

2. E2 America’s Divided Recovery: College Haves and Have-Nots - CEW Georgetown Georgetown University Center on Education and the Workforce

   “Over 95 percent of jobs created during the recovery have gone to workers with at least some college education, while those with a high school diploma or less are being left behind.”

Georgetown CEW reports that workers with at least some college captured 11.5 million of the 11.6 million jobs created during the recovery, about 99% (E1). It summarises this as over 95% of recovery jobs going to that group (E2). That share is above the 95% threshold. However, the figure covers jobs created during the recovery rather than net change since the recession began, and the evidence does not show the end date of the data. The criteria name this situation as mostly true.