Skip to main content

Reducing Absences, Capturing Days

Behavioral scientist Todd Rogers on new (and old) strategies to fight absenteeism and get more students back into the classroom.

Chronic absenteeism — when a student misses 10 percent of school days in a year — can hinder student success. It’s an increasing problem, says behavioral scientist Todd Rogers — one to which many school districts struggle to find an effective solution.

Reducing absenteeism is vital, says Rogers, as the Every Student Succeeds Act has led in many states to the increased use of absenteeism as one of the five metrics in districts’ evaluation. “So for the first time the vast majority of kids will be attending districts that are being evaluated by their states based on ability to reduce absenteeism,” he says.

But all absences are not equal. Students miss school for a lot of reasons with peak absent years in kindergarten, first, ninth, 11th, and the 12th year, or what is known as the “transition years,” says Rogers, a professor of public policy at the Harvard Kennedy School and director of the Student Social Support R&D lab. Further, parents aren’t always aware of how deep the problem is.

“In education you find that parents underestimate a child’s absences by a factor of two,” Rogers says, explaining that if child missed 20 days, the parent may think he’s missed only 10.

Traditional interventions, such as emails or phone calls, have not been effective in targeting children and their families. But Rogers’ recent research has shown that an old-fashioned notice in the mail is an effective way to reduce chronic absenteeism, by as much as 10 to 15 percent. The mail-based intervention conveys to parents a simple, straightforward message that targets false beliefs about the frequency of the absences and provides motivation.

“We focus on empowering parents with useful information,” he says. “We get parents more motivated to get the kid to school.”

Seeing the mailings’ success led Rogers to create InClassToday – an easy-to-implement absence reduction program that partners with schools and districts to reduce chronic absenteeism.

In this episode of the Harvard EdCast, Rogers talks about the issue of absenteeism and how to effectively get more students back into the classroom.

Read more about how parents can be allies in reducing absences in Usable Knowledge.


Jill Anderson: I'm Jill Anderson. Welcome to the new season of Harvard EdCast. Chronic absenteeism, when a student misses more than 10% of the school days in a year, is a big issue facing school districts, and a barrier to student success. The reason students miss school varies immensely. So Todd Rogers, a behavioral scientist at Harvard studying the problem, knew it was a question of how to motivate students and families to see the importance of attendance and reducing absences. 

It turns out, though, that what you think might work to target absenteeism, like rewards, or getting called into the principal's office, isn't as effective as a simple old fashioned mailing sent home to parents. I wanted to know more about this. I asked Todd whether all absent days are created equal, or when being absent from school starts to become a real problem. 

Todd Rogers

Todd Rogers: There are different kinds of absences. When a whole class misses, they basically just all fall behind concurrently and in parallel. But when one kid misses, the class moves forward, and that kid falls farther behind. But then when the kid comes back, the class has to slow down to bring the kid back up. And it may be a little bit, there are these hidden costs of absences that we don't have much empirical evidence on, but have to be really consequential. 

Jill Anderson: Right. Your research was looking at chronic absenteeism, which is something I've never thought about, had never heard about. And that's being absent 10% of the school year. So do we have any idea when being absent from school starts to become a problem for a kid? 

Todd Rogers: For now, let's just talk K12. There is not evidence that a certain number of days is a critical threshold. But there has been convergence, just for simplifying, that 10% of days is a meaningful measure. Even though missing the 16th day versus the 15th is not worse than missing the 18th day versus the 17th, right? We don't have any evidence of that. 

But it is a simplifying way of thinking about it. 10% of days seems like a lot. It's easy to organize around. And Every Student Succeeds Act has resulted in 39 states and DC using some measure of chronic absenteeism as one of the five metrics they evaluate districts on. And so for the first time, the vast majority of kids are going to be attending districts that are being evaluated by their states based on their ability to reduce absenteeism. 

Jill Anderson: So do you have any idea how much of a problem this is on a bigger scale? Like is this something you see a lot of? I know you were looking at kindergarten, elementary age. Is this a lot of kids who are missing such huge numbers of days? 

Todd Rogers: The answer is, lots, especially in large urban districts. And we, in my lab, have done a bunch of research on K through 12, not just elementary. And the peak absence years are kindergarten and 1st, and then 9th, 11th, and 12th, right? So basically, the transition years, and then the graduation year. 

Jill Anderson: Why is that? What's going on there that parents aren't sending their kids to school? 

Todd Rogers: It's incredibly varied. It helps to inform why the interventions we've developed have been so effective. Kids miss school because they don't want to go to school because they're being bullied, because they don't like it. Because parents work night shift and have a hard time managing the logistics of getting the kid to school the next day. Because a sibling is sick, because someone else in the family is ill and the kid has to be a part of the family. The family loses their home or gets evicted. They don't feel safe going to school. The transportation issues. I could go on. Every absence has its own story. Kid is sick. Parent thinks the kid is at risk of an asthma attack. Lots of potential reasons. 

But what we find with our interventions is that because it is so varied, unless you are doing an intervention that really gets into the details of a family's life and helps to address the structural challenges that families face, these more general interventions that increase motivation to get the kid to school — we focus on empowering parents with useful information — will get parents more motivated to get the kid to school. And that makes the parent more likely on the margin on every one of those things. Kid doesn't want to go to school, parent pushes back a little bit more. A sibling is sick, parent pushes back a little bit more on the logistics. Parent works night shift, parent's a little bit more likely to plan. 

And so, when it's as varied as it is, increasing parent attentiveness to reducing absenteeism and parent motivation makes them push the margin on all these different causes of absenteeism. 

Jill Anderson: But as a lot of us know, in education, reaching parents, parent-educator relationship isn't always strong or great. So tell me a little bit about some of the interventions, and some of the ways you're moving the needle on this type of thing. 

Todd Rogers: I'm a behavioral scientist, which means I do a cross between psychology, economics, education, political science, sociology, marketing, a cross between a lot of different disciplines. Basically, what we do is we borrow the best insights from different fields. And we focus on turning them into scalable interventions that mobilize and empower families to support kids more effectively. 

And so one of the original insights that we pursued is based on research in energy, where Robert Cialdini and collaborators Noah Goldstein and others did this original research showing that if you tell people that they use more energy than their neighbors by a printed mailing, that it gets people to reduce their energy use. Two Harvard alums, Alex Laskey and Dan Yates, who were friends, read this and started a company called Opower, that implements a version of this, where they send people monthly or quarterly mailings comparing their energy use to their neighbors. And found that these mailings become social artifacts. They get shared in the home. People put them on the fridge. They talk about them. One in five US households now receives them. They reduce energy use the equivalent of increasing the price of energy by about 25%. 

It's super effective. I've studied them over seven years, and the effect gets bigger over time. People don't just habituate to it. An underlying-- and I'm going to get to the education implication-- but underlying it is two false beliefs. People don't realize how much energy they use and they definitely don't realize they use more than their neighbors. 

So we found in education — originally in the school district of Philadelphia, then in Chicago, then in 10 districts in California, and my lab is now done this in lots of districts — that parents underestimate their own child's absences by a factor of two. So if my kid has missed 20 days, I think my kid has missed 10. 

And the second is, if my kid has missed more school than her classmates in her school, the majority of parents think their kid has missed less or the same. Meaning it's like the Lake Wobegon effect, where all the kids are above average. 

So these two false beliefs, we started developing these really easy to understand, mail-based interventions that are modeled after this energy intervention, that conveys to parents, as simply and straightforwardly as possible, written at a fourth grade reading level, easy to consume, personalized graph showing your kid has missed this many days, their classmates have missed this many, when relevant. And we find we've consistently reduced chronic absenteeism by 10% to 15% across randomized experiments now in over 12 districts involving approaching 50,000 or 100,000 families in these randomized experiments. 

And so we developed the intervention, and it was super easy to implement-- for us at least, although it requires data and expertise-- and it can be implemented at scale with fidelity, which is hard in education. 

Jill Anderson: It seems so crazy in 2019 that a postcard could — like a piece of paper is so helpful in something like this. You think about schools, they use a lot of paper, and they're pushing a lot of paper out there. But I'm sure a lot of it gets lost in a child's backpack and it doesn't make it home. So this is just such an easy way to do this. 

Todd Rogers: Right, so we do it through the mail. We intentionally make it so it's written at a fourth grade reading level and 16 point font, so easy to consume. We send it repeatedly, five or six times a year. Target false beliefs. And the reason printed mail, we've actually done experiments with texts. The average family receives 50 to 100 text messages a day. And they're really potent for mobilizing a specific action like, your kid hasn't turned in their homework today, get your kid to do it. 

But if it's a more diffuse action spread over time, it just disappears into the ether. Read the text. What am I supposed to do now? But these pieces of mail become what we call social artifacts. The majority families report putting them on the fridge, or putting them on the kitchen counter. And they report talking to the kid about it. Talking to other adults. They bridge time by having a shelf life in the home. The reports of mail's death is grossly exaggerated. 

Jill Anderson: When you say targets false beliefs, what do you mean by that? 

Todd Rogers: False belief is how many days your kid was absent. Personalized data about your kid, graphically represented. And we do these pilot tests where-- my collaborator on the original research is a guy named Avi Feller. He's a statistician at Berkeley. And we were trying to figure out how to visually display this. And I naturally think we should have up and down graphs, and he originally thought we should have downward graphs. Up and down, but down. More absences is negative, so you should go down. 

We user tested it with the Philadelphia parents and with others online. And it turns out that people have a hard time understanding graphs, period. And that horizontal graphs, graphs that go left to way easier to understand. Like, a ton easier, which I had no intuition on. And so that's the kind of thing that we focus on making it easy to understand. 

So the reason it works is because it can bridge time, because it's a physical artifact. We've seen in experiments, text messaging is nowhere near as effective. And we make it super easy to comprehend, and clear, and focused. Nothing but absences are important. They accumulate. Your kid has missed so many days. 

Jill Anderson: So if there is an educator, an administrator, listening to this and they're thinking, great, how do I get started doing something like this? Where do you begin? 

Todd Rogers: Yeah, so, originally, as I said, we did this in a dozen districts. About a half dozen districts approached me independently and were like, can you help us implement this? We want to do it ourselves. We want to implement it ourselves. It requires competence with the data, being able to produce graphs, being able to manage the mail flow, and then also being able to manage the hundreds and sometimes thousands of parents who call the district with questions about it. 

So I tried with these six districts. None of them managed to do it with fidelity at scale. Some of them rewrote the letters so they were written at a college reading level instead of fourth grade reading level, in 7 point font instead of 16 point font. Basically like a newsletter, which no one reads. Others couldn't do the graph. Others just couldn't pull it off. So we started an organization that actually is now helping dozens of districts and hundreds of thousands of families around the country implement this at scale. It's called In Class Today. 

And one thing that's really cool about them is that because they run lots of optimizing experiments, like you think internet A/B testing, the effect is about 50% more effective now than it was last year. Because they're figuring out who to target, what the optimal message is, what the optimal timing is. 

Jill Anderson: I guess it wouldn't make sense to do this if you only had a handful of students who were chronically absent, right? 

Todd Rogers: So I don't think about it in terms of chronic absenteeism. So the intervention was easy to implement at scale and it cost roughly 1/ 100 as much as the next best kind of intervention, which is like mentor, social workers, truancy officers. They generate about one additional day of attendance per $500 spent. And this intervention generates about one day of attendance per like $5 spent. So it's easily the most cost effective way to reduce absenteeism, and it could be implemented at scale without changing anything teachers or staff do. 

We don't think about it in terms of chronic absenteeism. We think about in terms of capturing all of the days that we can capture. Because reducing a kid from 10 days of absence to 8, it's not at all obvious that that is not as meaningful. That's a tough sentence I just said. But it's not obvious that reducing a kid from 18 to 16 days is better than reducing a kid from 10 to 8 days. Or reducing a kid from 24 to 22. So we think about it as, where can we capture the most days with this universal treatment across the relevant at-risk kids, so we can free the resources up to go after the harder days. 

Because these are the easy days we're getting. It's not a substitute for the more intensive work. Because we'll reduce chronic absenteeism by 10% to 15%. That still means 90% to 85% of kids are chronically absent still. And that's too many. So we don't see it as a substitute, more of a complement for the other kind of programs. Because the other programs are more expensive and hard to actually do at scale. 

Jill Anderson: It makes me think about when we had these awards for not being absent. I don't know if they still do that in schools. 

Todd Rogers: We actually have a paper that was very surprising to us. With 15,000 high school kids, all of whom, across 10 districts, all of whom have had a perfect month of attendance at some point in the fall. In January, we randomly assign them, and we'll talk about two conditions. We randomly assign them to a control group that gets nothing at the end of January. And a treatment group, which gets a really nice embossed award that says, congratulations on a perfect month of attendance. We mailed it in like a big envelope, thick card stock, and it's like fancy, weird calligraphy. It was a one time award. 

We expected that this would reduce absenteeism. What we found was that it increased subsequent absenteeism. Giving students this surprise award for perfect attendance increased absenteeism. So we didn't know why. In follow-up studies, we learned that the way students interpret this, they interpret it as meaning they attend school more than everybody else. And that they attend school more than the school expects them to. And that's why they're getting an award. So it licenses them to miss school. 

Now, consistent with this, which is really — like we didn't expect this, but it's consistent with the story-- the lowest academic performing kids showed the biggest effect. The kids who least want to be in school, when they get this award, they're like, oh, I guess I don't need to go to school as much as I do. 

It's not a huge effect, but it's a real effect. And it really pushes us to think, what are the unintended signals that we're sending when we do these kind of communications and awards? 

Jill Anderson: For you, studying this, what's been the most surprising thing about absenteeism? 

Todd Rogers: One is the robustness of the mail-based absence intervention that our organization now, In Class Today, now administers. In Class Today has done a bunch of experiments and consistently reduces chronic absenteeism at 10% to 15%. I'm sort of surprised, given the we've now studied this across large randomized experiments in more than a dozen school districts. I'm surprised at that because very few things are replicated this much and are that robust. That and I'm surprised that the effect keeps getting bigger as we get smarter. 

The second, I studied that awards thing because I, like other people involved in education, assumed it would work. And, after the fact, it makes sense. When you tell kids that they're attending school more than everybody else, and more than we expect them to, they're like, oh, I guess I don't have to go as much. 

But it really pushes me back to like a first order thing that I assume comes up a lot in this podcast. Just because something seems sensible doesn't mean that it's going to work. And we really do need a higher standard of proof in education for what works. Because it's really hard to know what causes anything unless you do these kind of randomized experiments or other kinds of careful analysis. Often things are correlated. 

So, for example, if you target an intervention at a kid who missed a lot of days, the kid is likely to attend school more subsequently. But that's a thing called regression to the mean. When people have extreme behavior, they tend in moving forward to come back to what is typical behavior. It's not because of the intervention. It's just because that's the statistical pattern. When someone has an outlier event, they tend to come back to what they typically do. And if you don't do a randomized experiment, you would mistakenly think that your brief intervention caused them to come back to typical behavior. 

Jill Anderson: What is it about that graph, do you think? Is it tapping into some competitiveness or something? And parents are keeping up with the Joneses? I mean, I think I feel that way when that electric graph come to my house. 

Todd Rogers: Oh, you get that? 

Jill Anderson: Yeah, it's the uh-oh, I'm using more than my neighbor. It doesn't actually make me change anything, in practice-- 

Todd Rogers: You believe that. 

Jill Anderson: — but I retain it. 

Todd Rogers: Everyone thinks it doesn't change their behavior and the data isn't real. But it appears to robustly and consistently change behavior. So the most impact that we can tell is, I think two things. One, it's hard to keep track of something that accumulates slowly over time. And parents genuinely don't know how many days their kid has accumulated. And when we're like, hey, your kid missed 12 days, parent's like, geez, I thought it was six. I didn't realize they'd missed that many days. Correcting that false belief seems to really matter. 

And the second, which I think is sort of underlying all of it, sending these really focused, easy to comprehend, repeated communications, draws attention back to this changeable behavior that parents have some agency over. And we can see this in the data. Which, I don't know how data nerdish you would like this podcast to be. When you graph the treatment group, the group that gets the intervention, versus the control group, you can see this reduction in absences. So the treatment group has fewer absences. But then as time goes on, the treatment group starts to migrate back to what the control group has. So, basically you've drawn their attention to this issue, they reduce their kid's absences. It slowly trickles back up to where it started. And then you intervene again and you see a spike again in reduced absences. 

So that signals to me that there's some attention issue going on. And parents have agency over whether their kid misses school. They can influence it. And this repeated intervention, repeated communications that are really focused, seems to both correct important false beliefs, while also focusing attention specifically on this in a comprehensive way. 

Jill Anderson: Todd Rogers is a professor at the Harvard Kennedy School, and also director of the Student Social Support R&D Lab. I'm Jill Anderson. This is the Harvard EdCast, produced by the Harvard Graduate School of Education. 


An education podcast that keeps the focus simple: what makes a difference for learners, educators, parents, and communities

Related Articles