I’m excited to announce the launch of No-Spin Evidence Review - a new online publication of the Coalition for Evidence-Based Policy. We summarize recent program evaluations & explain what the evidence really shows. See example below & check out No-Spin at

Terrific & compelling remarks by Australian MP Hon. Andrew Leigh at the OECD, outlining why randomized trials are the latest major reform in the history of public administration, and articulating the ultimate goal of evaluation: Finding interventions that work!

See our new No-Spin report on an RCT of Communities that Heal - a community engaged, data driven program to reduce opioid overdose deaths. Large, high quality RCT finds no discernible impact on such deaths over 1y follow-up:

Note that the study received uncritical coverage in NY Times & elsewhere - coverage that largely repeats the study abstract's inaccurate claims:

See our new No-Spin report on the evaluation of a California program to train K-3 teachers in the science of reading. Despite the study abstract's claimed positive effects, the study found no impact on reading/English achievement in its primary analysis:

Recently released (Jan 2025): 4 new reports of No-Spin Evidence Review – our online publication that summarizes recent program evaluations and explains what the evidence really shows. See

See our No-Spin report on the Sit-D Police Training RCT published in QJE: Although the study claims effects on police performance, it doesn’t report results for most of its preregistered primary outcomes, preventing reliable conclusions re program impact.

See our new No-Spin report on the RCT of ASSISTments - a low-cost online tool to support 7th grade math learning: High-quality RCT finds sizable gain (~31% of a grade-level) in math scores on the state test at 2y follow-up.

See our No-Spin report on the NC Early College RCT (giving students opportunity to earn college credit in high school): Despite claimed large effects in the study’s abstract, its main finding was a modest 2.4 % pt gain in bachelor’s completion at 10y mark.

See our No-Spin report on the RCT of Roca (a program for justice-involved young men to prevent recidivism): High-quality RCT finds no discernible effects on reincarceration or employment over 5y; is able to rule out sizable but not modest positive effects.

@weswinham Hi Wes, there was no discernible intent to treat impact, but the researchers’ nonexperimental analysis suggests positive impacts when teachers gave students >35 min/week of practice time with the program. Seems like a good hypothesis to test in a future RCT.

See our new No-Spin report on the Choose to Change RCT (group therapy & mentoring for at-risk teens): Despite the study's claimed effects on youth crime, it found no discernible impact on the preregistered primary measure of crime - # of arrests - over 3y.

See our new No-Spin report on Project QUEST's occupational training of low-income adults for well-paying healthcare jobs: High-quality RCT finds a 15-20% gain in annual earnings at the 14y mark - an extremely promising (but not quite definitive) result:

Just released: Our No-Spin report on the RCT of City Year's Tier 2 services (e.g., tutoring, mentoring) for at-risk middle schoolers. High-quality study finds Tier 2 yields no added benefit to students over 1-1.5y, vs City Year's schoolwide assistance:

RT @JohnArnoldFndtn: Great to see a Denver city councilman quoting an RCT of a basic income pilot meant to reduce homelessness that didnt w…

@AnyaSamek Thanks - I agree on the value of such correlational findings. But this paper also reports on the study's prespecified primary outcome (grade-school test scores) & exploratory outcomes (discipline, grades), & portrays results as positive even though the primary impacts were null.

Here's the forthcoming paper (ahead of print) in Journal of Political Economy: Disclosure: my former employer (Arnold Ventures) helped fund this study.