Every Year You Postpone Keyboarding Costs One Cohort

Delaying keyboarding does not cost your district money. It costs cohorts. At 400 students per grade, every year you wait sends 400 more students past the elementary instruction window, and if you catch them up later your launch year grows from 1,200 students to 1,600. Meta Typing Club charges no license fee while you run the arithmetic.

TL;DR: At 400 students per grade, a keyboarding block launched this year serves 1,200 students in 300 to 720 supervised hours. Wait a year and the launch year serves 1,600 in 400 to 960 hours. Cumulative hours over any horizon are identical. Delay costs 400 students a year who leave the window untaught. Meta Typing Club carries no licensing cost as of 2026.

The line item that never appears in the budget book

A postponed initiative enters the budget as a zero. Your finance office can price a program you approve. It cannot price one you defer, because that cost lands in a later year under someone else's name.

Keyboarding is the cleanest example. It draws little open opposition yet rarely survives master scheduling: the case for it sounds qualitative, the case against it is a bell schedule with no gaps. We are Meta Typing Club, a typing platform, so check every figure against your own file. This block costs about one percent of the instructional year, the same claim nearly every pitch on your desk makes. The percentage carries no weight until someone names the minutes and the adult who gives them up.

A decision nobody has priced loses every budget argument it enters.

One cohort exits the window every June

New York's Education Department ties a keyboarding ladder to its ELA standards: students "should be introduced to keyboarding" in grade 2 and "should receive instruction in keyboarding, with a focus on technique over speed" in grades 3 and 4 (NYSED, Guidance on Keyboarding Instruction, 2017).

That window is three grade-years wide, and if your state sets a different ladder, substitute it. While the program is idle, one cohort exits untaught every June into a grade 5 that expects a foundation nobody built. Those students stay teachable, and this model charges the same hourly rate at twelve as at eight. What gets harder later is finding the minutes in a schedule that is already full.

The live question is how many cohorts leave the window untaught while the decision stays open.

What the model assumes and what you supply

Two rows below are settled by published guidance; the rest come out of your own files. The 400 is illustrative, so substitute your enrollment and every figure here rescales.

Input	Source	Value here
Students per grade	Your enrollment	400, illustrative
Grades in the window	NYSED ladder	Grades 2 to 4
Hours per student, one block	Wisconsin DPI	6.25 to 15
Instructional hours per year	Your calendar. ECS counts 31 states plus DC requiring at least 180 days	1,170, from 180 days at an illustrative 6.5 hours
Students per section	Your class sizes	25
Keyboards in those grades	Your inventory	Assumed sufficient; verify
What the block displaces	Your master schedule	Not modeled; you name it

Wisconsin's Department of Public Instruction tells schools to "set aside a regular time for keyboarding (ideally 15 to 30 minutes/day for five to six weeks)" (Wisconsin DPI, Keyboarding at the Elementary Level). Fifteen minutes a day across five weeks is 6.25 hours per student; 30 minutes across six weeks is 15 hours. Against a 1,170-hour year, one block takes 0.5 to 1.3 percent of a single year. One caveat: this model charges one block per student, while the NYSED ladder implies instruction in grade 3 and again in grade 4, so these hours are a floor. Name the displacement yourself. If it is the fifteen minutes at the front of grade 3 ELA that now run as independent reading, your board needs that in writing.

Three start dates, one district

Launch day serves every grade inside the window. At 400 per grade that is three cohorts, 1,200 students, 48 supervised sections of 25, and 300 to 720 hours of teacher time. Each section runs 6.25 to 15 hours, or 1.25 to 2.5 hours a week for five to six weeks. Start later and the launch year carries the stranded cohorts too.

If you start	Launch-year students, catch-up included	Sections of 25	Supervised hours that year	Left grade 4 untaught first
This year	1,200	48	300 to 720	0
Next year	1,600	64	400 to 960	400
In three years	2,400	96	600 to 1,440	1,200

Over any horizon, delay adds no supervised hours. A district that waits three years skips three years of steady state, then repays it all in the launch bulge. By the time a 2029 start finally launches, 2,400 students have had the block either way.

The honest reading of our own arithmetic: delay is free in hours only to the extent you are willing to leave cohorts untaught. Serve them and you pay the same total in a worse shape: a launch year up to twice the size, in whichever year has the least slack. Flat enrollment is assumed. Waiting concentrates the same hours into one oversized launch year and spends them on older students in a fuller schedule.

Give the first block to your fourth graders

If you cannot launch across grades 2 through 4 at once, do not begin with grade 2. Begin with grade 4. Youngest-first sequencing is defensible on every ground except the one that decides this: second graders are still inside the window next year, while fourth graders are in their last one. Default to youngest-first and you strand the 400 with the least time left.

Phase the rollout downward from the oldest grade in your window, the only one that cannot wait.

What the delay does to those students

A cohort that leaves the window untaught still meets typed assessment on schedule. Smarter Balanced runs in grades 3 through 8, and its ELA performance tasks require an essay written in the computer-based delivery system (Smarter Balanced, 2022-23 Summative Technical Report).

A 2019 meta-analytic review links handwriting and keyboarding fluency, mainly on speed (Feng, Lindner, Ji and Joshi, Reading and Writing, 32(1), 33-63, 2019). That is a correlation, not proof that a block raises scores. The closest direct evidence is a 2024 trial in which 51 fourth graders given a typing and word processing program improved on typing and composition measures against a control group (Reading and Writing, 2024). We omit the authors and the effect size because the publisher blocks retrieval of the record.

One figure is missing. No published study prices the unlearning premium, the extra hours a seventh grader with four years of hunt-and-peck needs against a third grader starting clean. Practitioners assert it, nobody measures it, so this model charges none. If that premium is real, it is the only mechanism by which waiting adds hours a district would not otherwise owe.

What this actually costs to run

Each student needs the block once, 6.25 to 15 instructional hours. Each launch wave costs 300 to 720 hours of adult time at 400 per grade, the labor line your plan must name. After the wave clears, one new cohort enters the window every year: about 16 sections and 100 to 240 supervised hours. That recurring row is why the delay totals cancel.

Classroom teachers run it unaided: classes by invite code, homework with due dates, per-student speed and accuracy in the dashboard. Across Meta Typing Club's 10,000+ learners, daily practice averages roughly 10 words per minute of gain per month early on. That is our own average, and it flattens near a student's ceiling. We point teachers to short daily sessions for best results.

The licensing line is zero as of 2026. Meta Typing Club is free to use with no subscription and no per-student charge, with more than 2,000 structured lessons in multiple international languages and correct right-to-left layouts for Pashto and Dari, which matters for newcomer enrollment from Afghan families. Our terms reserve the right to introduce paid features later while committing that core practice stays accessible. The minutes are the durable cost; the license is the likeliest line to change.

The best case for waiting deserves an answer

The benefit side of this analysis is thin. One small trial and a correlational review are not the evidence base you demand before committing 720 hours, and a model that prices delay in ranges while leaving the payoff qualitative is doing what vendors do. That objection wins. The arithmetic measures exposure to instruction, not improvement in any outcome you are evaluated on.

Wei Roy Hua's 2026 arXiv preprint argues that AI dissolves typing's instrumental role by shifting cognitive labor from composing text toward verifying machine-generated text (arXiv:2604.17023). It is single-author, not peer reviewed, and about adult work rather than schools. The direction still deserves weight.

The asymmetry survives. Starting and being wrong costs about one percent of a year's instructional minutes and no license fee. Waiting and being wrong costs a cohort its place in the window. Buy fluency until the keyboard stops being the bottleneck for a child's ideas, and treat speed targets beyond that as optional.

Key takeaways

• Delay is priced in cohorts: at 400 students per grade, every year of waiting strands 400 students past the window.
• Over any horizon, delay adds zero net supervised hours. The delayer skips steady-state years and repays them exactly at launch.
• Delay changes the shape of the bill: a launch this year covers 1,200 students, a launch in three years covers 2,400 students, 96 sections and 600 to 1,440 hours.
• Delay is free in hours only if you accept leaving cohorts untaught. Serving them later costs the same total in a worse shape.
• The model charges no unlearning premium and only one block per student, though the NYSED ladder covers grades 3 and 4. These hours are a floor.
• Verify two inputs before you present: physical keyboards in those grades, and which minutes the block displaces.
• Efficacy evidence is thin: one 2024 trial of 51 fourth graders and a correlational review. This model prices exposure only.

Questions superintendents ask about this model

Our enrollment is nowhere near 400 per grade. Does the model still hold?

Everything rescales linearly. A district with 90 per grade strands 90 students a year. Start now and the launch year is 270 students and 68 to 162 hours. Wait a year: 360 students, 90 to 216 hours. Wait three: 540 students, 135 to 324 hours.

Can we pilot in one building instead of committing district-wide?

A pilot answers schedule fit and whether classroom teachers can run it unsupported. It will not settle the outcome question inside one year, and every other building spends that year adding its grade to the stranded count.

Who owns this if we have no keyboarding teacher?

Classroom teachers, with the platform handling sequencing and record-keeping. What matters is which minutes the block occupies, and principals settle that in a scheduling meeting. Our help pages cover the teacher workflow. If the answer requires a new hire, it will not clear the budget cycle.

How do I show the board progress inside one school year?

Report participation and skill movement separately: completed lessons and practice days per class, then words per minute and accuracy per student. Neither is an outcome claim about writing or test performance, and presenting them as one is how districts lose credibility.

Is there real evidence that starting earlier beats starting later?

Not directly. Nobody has published a head-to-head comparison of early versus late starts in keyboarding, and we would rather say so than manufacture a figure. The case for acting now rests on window arithmetic and on which error is cheaper.

What about the students already past grade 4?

They remain worth serving. No published study covers students taught outside the window, so we will not claim the grade-4 result transfers. What is harder in grade 7 is finding the minutes, not the rate. Tell the board you are paying a bill an earlier decision would have avoided.

If Meta Typing Club is free now, what is our budget exposure in three years?

The exposure is that a paid tier appears and the free version narrows. Our terms reserve that right while committing that core practice stays accessible. Ask every vendor the same question, plus what happens to student data if you cancel.

What to do before your next board meeting

Open the enrollment file and change one number. Put your students-per-grade figure into the table of model inputs, multiply by the years already spent deciding, and you have the count of students who left grade 4 without a keyboarding block on your watch. The total hours hold steady. What grows is the launch year, and that count.

Then settle the smaller question: which grade gets the first block, and which fifteen to thirty minutes it displaces. Build it on Meta Typing Club's structured lessons at no licensing cost today, or on any platform that gives teachers assignment and progress visibility. For a district-scale rollout, write to us.