The Evening-Planning Tax: Giving Teachers Their Nights Back

Ask a teacher what they did last night and a surprising number will describe a second shift. Dinner, then the laptop open again. A unit to map. A worksheet to bend for the three students who aren't there yet and the two who raced past it a week ago. A stack of exit tickets to skim. Tomorrow's plan to rebuild, because today didn't land the way the plan promised. The lesson happens in the classroom. The planning happens at the kitchen table — at nine, at eleven, on a Sunday that was supposed to belong to someone else.

That uncounted second shift is the evening-planning tax, and cutting it is the actual point of Claude for Teachers. Not raising test scores. Not dropping a chatbot into the back of the room. Giving time back. Anthropic frames the entire product around a single gap: the distance between what good instruction is supposed to look like — differentiation, mastery-based learning, small-group work — and what a teacher's week actually has room for. The evening is where that gap gets paid, in the one currency a teacher can't manufacture more of.

The tax nobody wrote into the contract

The job people picture is the part with students in it. The job teachers live is mostly the part around it. Planning is invisible labor — it produces no artifact anyone photographs, earns no applause, and appears on no schedule — yet it is the work that decides whether the visible hours go well. And it is elastic in the worst way. There is always one more student to plan for, one more misconception to anticipate, one more version of the worksheet that would serve the kid in the third row a little better. Conscientiousness is precisely what makes the tax bottomless. The better the teacher, the longer the night.

Here is where those hours actually go:

• Adapting one lesson to the room you actually have. A single class is really several classes wearing the same period. Meeting them means producing tiered versions of the same content — below, at, and above proficiency level — without watering any of them down. Done by hand, this is the single most time-expensive move in teaching.
• Finding and vetting raw material. Hunting for a text at the right reading level, a problem set aligned to the standard, a diagram that isn't subtly wrong — then trusting it enough to put it in front of thirty children.
• Re-planning tomorrow because today moved. Formative signals arrive every day. Acting on them means rewriting the next lesson the same night you gathered them — which is exactly when you have the least energy left to do it.
• The administrative residue. Converting a plan into a clean, printable handout. Re-formatting. The last mile between "I know what I'll teach" and "the materials physically exist."

None of that is glamorous. All of it is the evening.

Why the strain lands hardest where resources are thinnest

The tax is not levied evenly. The pressure is heaviest in under-resourced schools — the places with the largest classes, the widest spread of readiness inside a single room, the fewest planning periods, and the least access to ready-made, high-quality materials. A teacher with a curriculum coordinator, a well-stocked resource shelf, and a manageable course load pays a smaller tax than a teacher who is the coordinator, the resource shelf, and the whole department at once.

That is why treating this as a productivity gimmick misses the point entirely. When the scarce resource is a teacher's planning time, anything that returns hours to the people who have the least of them is not a convenience — it is an equity move. The teachers who most need a planning partner have historically been the ones who could least afford one. A tool that is entirely free once you're verified, and that plans alongside you rather than around you, changes who gets to teach well without burning out.

Time is the thing teachers actually value

Ask people why they teach and almost nobody says "to write lesson plans." They describe some version of the relationship — the moment a concept lands, the kid who comes back a year later, the thirty small human transactions a day that are the entire reward. The evening tax is corrosive precisely because it trades that reward for logistics. Every hour spent re-formatting a worksheet is an hour not spent on the part of the job that made it worth doing in the first place.

The person who fought for the privacy guardrails behind this launch put the goal in exactly those terms. Randi Weingarten, president of the American Federation of Teachers, on the "Gold Standard" of K-12 safety and privacy she has been developing with Anthropic:

"We've been working with Anthropic on a Gold Standard that sets out industry best practices for safety and privacy in K-12 education… It's important that Anthropic is committing to these principles in their new Claude for Teachers — a tool designed by and for educators to assist them instructionally and hopefully give them more time for the human relationships at the heart of learning."

More time for the human relationships at the heart of learning. That is the entire thesis of giving evenings back, stated by someone whose whole job is protecting teachers.

What actually comes off the plate

Claude for Teachers attacks the tax at its most expensive points, not its cheapest. You can ask it to plan a lesson from high-quality instructional materials aligned to your state's standards, and it drafts both the plan and the student-facing materials for you to revise — the details of that flow are worth reading in Plan a Week of Lessons from High-Quality Instructional Materials. You can hand it an existing lesson and get tiered versions back, so the differentiation that used to eat a whole evening becomes a review-and-adjust job. You can point it at a folder of class data — roster, diagnostics, notes — and have it build a picture of where each student stands, with you controlling exactly what's shared and nothing shared going toward model training.

And because the product includes Claude Code and Cowork, the work can carry forward without you sitting there. You can schedule the recurring parts — for example, a daily 4pm pass over the day's exit tickets that adapts tomorrow's plan — so it runs while you drive home. A curated helper like lesson-plan-studio or a broader bundle of education agent skills does the same kind of work outside the classroom-verified product. The through-line is simple: hand off the labor that produces the plan, keep the judgment about what the plan should be.

Giving evenings back is a pedagogy decision

It would be easy to file this under "efficiency" and move on. That undersells it. When a teacher gets an evening back, that time doesn't vanish — it flows to the parts of the job a tool can't touch. Noticing the kid who's gone quiet. Rewriting the one explanation that isn't clicking. Being rested enough tomorrow to teach the way you know how. Making small-group instruction feasible instead of aspirational, which is a whole argument in itself over in Small-Group Instruction at Scale.

The evening tax was never a sign that teachers weren't working hard enough. It was a sign that the week was structurally too small for the job. You cannot fix that with a motivational poster or another mandate. You fix it by taking the mechanical, repeatable, high-effort planning work and handing it to something that's good at exactly that — so the finite human budget gets spent on the irreplaceable part. That is what a free planning partner, aimed squarely at time, is for.

---

Part of the Claude for Teachers series. Related: Plan a Week of Lessons from High-Quality Instructional Materials · Small-Group Instruction at Scale. Browse AI tutoring skills or more builder insights.