@jakeanders.uk and his team show that PISA slightly changed its wording for a question about truancy in 2015, which substantially changed responses (more students report that they "missed" than "skipped"). Useful broader insight: wording will affect how families/educators describe attendance issues.

P.S., here is some data from MI and GA showing how the rise in absenteeism is driven by moderate increases in days absence. Most schools have attendance rates equivalent to ~2-4 extra absences. (This graphic is slightly truncated for clarity; some schools have better attendance.)

Nat is really focused on the question of why absenteeism rates increased since the pandemic, and why those increases have largely persisted. He rightly suggests that a broad shift in norms and expectations may play a role. But... 2/3 www.aei.org/research-pro...

This week @ballotpedia.org covered @sarahlenhoff.bsky.social and my work on chronic absenteeism alongside Nat Malkus from AEI. While are arguments are framed as opposing, what they really represent are the answers to two different questions. 🧵 1/3 news.ballotpedia.org/2026/07/15/h...

Some corroborating evidence here: we found that even in Detroit, where the majority of students are chronically absent, students who were chronically absent in the prior year had 9x higher odds of being chronically absent in the current year. journals.sagepub.com/doi/10.1177/...

…Detroit has very low proficiency rates but when you look at growth they do very well (on average) as a district! edopportunity.org/trends/explo...

Second nitpick, way too negative on the impact of schools and teachers in disadvantaged contexts. Here, I ultimately end up agreeing with his main point, but a few detours, especially since he draws on Detroit as (rightfully) a key illustrative example!

Plenty of reasonable ideas here but the idea that MI has “ spent more and got less for our educational dollar for at least two decades” is ridiculous. Adjusted for inflation, we’ve only gotten back to pre-Recession levels of spending in the past couple years. epicedpolicy.org/wp-content/u...

Relatedly, in our recent report, we did find that the small number of schools (~10%) where principals reported using attendance-based grade retention had higher attendance value-added scores on avg. (but those using court-based prosecution had lower scores on avg) detroitpeer.org/wp-content/u...

Really interesting evidence from national policy in Norway where high school students don't get credit for class if they miss >10% hours w/o medical excuse. Useful bc we have little evidence about "punitive/expectation" mechanisms. Meaningful but notably modest effect: ~2 fewer absences per year!

NBER working paper version (maybe easier for folks to access?) of the authors' paper is here: www.nber.org/system/files...

Featuring our latest study on attendance value-added and school attendance practices! detroitpeer.org/wp-content/u...

Finally, for school/district leadership, we found a modest positive association between value-added and (a) school leader involvement (especially those frequently engaged in monitoring attendance initiatives) and (b) district guidance, especially when school leaders perceived it as very helpful.

I apologize! Got sidetracked and didn't finish the thread. Here it is: detroitpeer.org/wp-content/u...

We also found that the presence of several staff positions--particularly those that provide schools with additional administrative capacity and which focus on providing students and families with greater support--were associated with higher attendance value-added.

...some strategies are (at least modestly) associated with higher attendance value-added scores on average. For home visits (shown here), you can see meaningful differences in the distributions of value-added scores. In most cases, however, the overlap is great, will little differentiation.

Notably, we found that attendance value-added was positively correlated with economic disadvantage. This is likely due to: (a) schools with more low-income students implementing a broader set of strategies and (b) school-based efforts being even more impactful for disadvantaged students.

First, our value-added estimates. We found meaningful variation. If the avg student moved from lower value-added school (i.e., 25th percentile) to a higher value-added school (i.e., 75th percentile), they would be expected to attend ~7 additional days of school. These estimates are stable over time.

NEW: We connected school attendance value-added estimates (for 2022-23 through 2024-25) to statewide survey data on school-based attendance practices. Our goal was to identify effective attendance strategies (e.g., specific practices, organizational systems, staffing, leadership). What did we find?🧵

I think this study by Dan Hamlin is a good starting point (not perfect). He looked at transition from MS to HS & relative change in climate. He finds a modest association here (e.g., +1 day for entering a safer climate, -2 days for entering less safe climate). journals.sagepub.com/doi/abs/10.3...

The Educational Opportunity Project now has updated data for school years 2022-2025. It’s a powerful illustration of why it’s important to look at growth, and not just achievement, to understand how schools are doing. edopportunity.org/trends/explo...

Michigan has made progress in the past few years thanks to increased education spending, but we have a long way to go to achieve adequate funding for high-poverty schools and districts.

...that incorporating measures of local partisanship (Trump vote share in 2020) and a proxy for union strength (though this adds less explanatory power than partisanship) substantively attenuated but did not eliminate the association between funding adequacy and school reopening.

Compelling new evidence on the positive impact of the community schools model from @oppinsights.bsky.social. Substantial positive impact on academic outcomes and later-in-life earnings, especially for lower-income/higher-risk students. opportunityinsights.org/paper/cis/

I testified today about inter-district choice in front of the MI House of Reps education and workforce subcommittee. I was proud to spotlight the excellent and robust scholarship from MI on this topic! 🧵

“State context explains roughly 33% of the variation in daily attendance” — I think part of this has to do with weather! (whereas other macro factors vary more w/in state) We found that temperature was pretty strongly correlated with absenteeism in major cities journals.sagepub.com/doi/abs/10.1...

Now posted on @annenberginstitute.bsky.social EdWorkingPapers: What are schools doing to improve attendance? Based on school leader surveys in MI and GA, we describe the prevalence of various attendance practices and organizational systems. We find little variation, both between and within states.

One thing that's really interesting here is the travel standard limits, which I think are reasonable! They don't want students to have to travel more than 30-40 minutes one way. But it shouldn't just be a rural standard. Note that in urban school choice contexts, students routinely travel that far.

I'm skeptical about their claim that ADA-based funding is improving student attendance! That basis for recommending sticking with ADA-based funding seems thin to me. edpolicyinca.org/sites/defaul...

Though the negative effects are significant and persistent, the overall impact on MI's post-pandemic attendance is modest, mainly because few students were in districts with remote-only instruction for more than 3 months.