Editor’s note: This essay is an entry in Fordham’s 2026 Wonkathon, which asked contributors to answer this question: “What rules should govern the uses of digital technology in schools—and who should set them?” Learn more.
Before I outline rules for the use of digital technology in schools, I want to make a distinction between the ideas of “digital instruction” and “digital delivery.” I think that these two things are a bit different and merit explicit definitions in order to come up with coherent rules for the use of digital technology.
“Digital Instruction” refers to using digital technology to teach a lesson that requires the use of digital technology. Looking up digital images in order to study them is a form of digital instruction.
“Digital Delivery” refers to how information is presented and how students engage with the learning. Reading a passage on a screen that might have been a worksheet a few years ago is an example of Digital Delivery.
Take any task being done in a classroom; can it be done with pencil and paper or other analog technology without significantly changing the lesson? If so, then we are discussing the digital delivery of content. If the task cannot be reasonably done by most teachers without the technology then it is a matter of digital instruction.
Instead of trying to say “no kid under a certain age should be allowed to use tech,” we should consider adopting some heuristics that I am borrowing from other industries and applying to education. I also don’t like the idea that we should use a heavy-handed rule like “No more than two hours of tech use a day” because that potentially hampers teacher creativity. More importantly, it then creates a rule that someone has to potentially document or verify.
I would like to see schools and districts impose these heuristics:
1. Default to analog delivery: Education instruction should default to analog delivery as much as possible. Digital technology should have to affirmatively make its case for being the mode of instruction.
2. Minimum tech dosage: If technology is going to be used, a teacher should do their best to decide what is the minimum amount of tech that can be used to get the specific job done.
The easiest way to evaluate this, in my opinion, is the “1990 teacher litmus test,” to help people determine whether these rules apply in a given situation: Could a teacher in 1990 teach the same content without a digital component? If they could teach it effectively, then the present-day teacher should work to teach just as a teacher would have taught then.
I chose 1990 as the cutoff because at that point in time few classes would have computers in the classroom. Computers would likely have been in a separate “computer room” at that point and would have been a class on its own where skills like typing were explicitly taught. Technology would have to be deliberately introduced by the teacher in a 1990 classroom; whether it was wheeling in the giant TV and VCR to show a movie or taking students to the computer room to write an essay, all the tech use would have to have been a deliberate choice for the class. Ideally we should get back to a similar level of deliberate use of technology.
What might a class that passes the 1990 litmus test look like? It would look like a teacher that writes on the whiteboard and expects students to copy information in a notebook. The student with dysgraphia types their notes on a laptop, ideally one that isn’t connected to Wi-Fi automatically so that there are few digital distractions.
The teacher assigns students to research a topic for a presentation that will happen next week. The students open up their laptops for 20 minutes to research their topic. They need to have written a paragraph and gotten two sources. Headphones aren’t permitted for students to listen to music while they work nor can work be done on any other device. The teacher checks to make sure students aren’t on other websites. AI is not the first place students go to find information as they need to find a direct source for their presentation that they can personally verify.
When the students do their presentation based on their research they aren’t reading PowerPoint slides off a board they are talking in front of the class and sharing what they learned. Maybe they made a poster that they put up on the board to help with the presentation but there is no digital interface to the distribution of the information.
The students take a paper-based test which gets graded. The feedback is manually delivered but also recorded on the LMS for parents to see.
The total time students are in front of a digital device would be a fraction of what it is today. Far more importantly, the number of distractions eliminated would be significant.
Why should students have to do work that is likely more difficult and contains significantly more friction in order to accomplish? Because the goal isn’t the completion of a perfect “work product” so much as the goal is for students to refine their thinking through the process of creating work. If a student asks an LLM to generate eight PowerPoint slides about a topic and then reads it out loud to a class has the student learned anything? If that same topic were done by hand on a poster the student will likely have thought more and learned more in the process of completing the task.
Conversely, it is unlikely that a statistics class in 1990 would be able to analyze passenger data from the Titanic. If I wanted my statistics class to download passenger data from the Titanic and teach them how to make Pivot Tables I should do so without concern since I am engaged in digital instruction. A teacher in 1990 could theoretically assign this work by giving students a paper copy of the dataset but the manipulation of that data necessitates a digital interface. It would be unreasonable to assume that this lesson could be done by most teachers back in 1990.
I am not demonizing the use of technology. The example I gave of analyzing Titanic data to teach statistics is a valuable tool for teaching students. It is wonderful that a teacher can do something this creative. I do not want to discourage innovative teaching but I also do not want to re-dress old ways of teaching with digital skins and call it “progress.”
We didn’t decide that the teaching default should be digital. It just kind of…happened. Did we decide to do that because it offers superior instruction or because it feels easier to deliver? The greater difficulty of analog work and the difficulty of switching attention easily leads to the thinking outcomes we want to encourage in students. But if the expectation is that all work that can be done digitally should be done digitally, then there will be resistance in classrooms to doing work on paper or anything that feels “antiquated.”
The ubiquity of digital delivery devices does not mean that it needs to be the default. Setting the default to be what worked for years will go a long way towards improving education outcomes without stifling teachers' technological creativity. The goal of the 1990 litmus test isn’t to bring students back to the bygone era (as much as I’d personally like it) so much as it is to bring students back to a time where technology was a deliberate choice instead of an unconscious default.
