Why the defaults are thin and the college fields are blank
Every published average for what a child's education costs is built by pooling households that are not alike. Childcare in a high-cost metro and childcare in a small town differ by a factor of three. A family with a grandparent nearby pays nothing for it. A family with two children pays a sibling-discounted rate that is not double the first. Averaging those together produces a number that describes no household on the list, and the further out the projection runs, the less it describes.
College is worse, because the published cost of attendance and what families pay have drifted apart. Institutional discounting means a large fraction of students at private colleges pay materially less than the sticker figure, and the discount is set per student by a formula the college does not publish. Quoting an average here would make this page confident and wrong. The college fields are empty on purpose: fill them from the published cost of attendance of the specific schools in play, which is a real number attached to a real institution.
How the years are built
The calculation walks one year at a time from your child's current age to the last year of college. Each year is assigned to one stage, and the cost of that year in today's dollars is assembled from the lines that apply to it.
| Stage | Runs from | What it counts each year |
|---|---|---|
| Childcare | Now until the age you set | Twelve months of the childcare figure |
| School years | End of childcare to the start of college | The annual school-year figure, twelve months of activities, and private tuition in the age range you set |
| College | The start age, for the number of years you set | Cost of attendance for the path selected, less expected grants and scholarships |
Each year's total is then multiplied by the inflation factor for the number of years away it is, and rounded to the cent. The running total in the table is the sum of the rounded figures, so the column adds to the headline number exactly rather than approximately.
The inflation rate is doing a lot of work
Over an eighteen-year run the difference between three percent and five percent is not marginal — the final year's cost differs by roughly forty percent between them, and the total by around twenty. Education costs have historically not tracked general inflation, and the relationship has not been stable enough to extrapolate confidently in either direction.
The defensible way to use the field is to run the page three times: once with a rate you would call optimistic, once with one you would call pessimistic, and once in between. The spread you get is a more honest answer than any single number the page could give you, and it tells you something the point estimate hides, which is how much of the risk sits in the assumption rather than in the costs.
The monthly saving figure and what it assumes
The saving number is the level monthly contribution which, growing at the return you entered, exactly funds every year's cost at the moment it falls due. It is a present-value calculation: each future year's cost is discounted back at the return rate, the discounted amounts are summed, and that sum is converted into a level monthly payment over the years remaining until the last one.
Three things it assumes and you should not: that the return arrives smoothly rather than in the order markets actually deliver it, that you never stop contributing, and that the cost estimates hold. A portfolio that is up for fifteen years and down twenty percent in the year the first tuition bill arrives has funded less than the arithmetic says. That sequence risk is the standard reason college savings are shifted toward safer assets as the date approaches, and this page does not model that shift.
What belongs in a conversation instead of a calculator
Two things sit outside anything a page like this can compute. The first is that borrowing exists for college and does not exist for retirement, which is why financial planners generally treat retirement saving as the prior claim. That is advice with a reason attached, not a rule, and how it applies depends on your ages and what you already have. The second is that the number this page produces is a projection of a decision that has not been made yet — which school, whether four years, whether at eighteen. A total that looks alarming eighteen years out is a reason to keep planning, not a verdict, because almost every input to it will change before it arrives.
Questions people ask
Why does the calculator not fill in an average cost of college?
Because the number that gets published is the sticker price, and a large share of students do not pay it. Colleges discount their own list price through institutional grants that are set student by student, so two families at the same school with similar incomes can face different bills. Any average that ignores that is misleading in one direction, and any average of what families actually pay hides the enormous spread. The figure you should use is the published cost of attendance for the schools you are actually considering, which each institution has to publish, and then adjusted by whatever aid estimate that school gives you. Running the page twice with the highest and lowest school on your list is more informative than any single average.
Should I include the child's own earnings, or loans?
Earnings from a summer or term-time job can go in the grants and scholarships field if you want the page to net them off, though they are usually small next to the other lines and they vary year to year. Loans should not go in that field. A loan does not reduce what the education costs, it moves the payment later and adds interest, so treating it as aid makes the total look smaller than it is. If you expect to borrow, use this page for the full cost and then handle the borrowing separately in the loan calculator, where the repayment schedule is what you actually want to look at.
How do I use this for more than one child?
Run it once per child and add the totals, rather than doubling one run. The overlap matters: two children three years apart have three years where childcare or college costs land simultaneously, and that is the pressure point a combined average hides. If you want to see the collision, run each child separately, write the year-by-year columns next to each other by calendar year rather than by age, and add across. Sibling discounts on childcare and private tuition are real and worth applying to the second child's figures rather than copying the first child's.
The total looks impossible. Is that right?
It is arithmetic on the numbers you supplied, compounded over up to two decades, so it is right in the sense that it follows from the inputs. Whether it is a useful prediction is a different question. Almost nothing in an eighteen-year projection survives contact with the actual years: incomes change, childcare ends earlier or later, activities are dropped, aid arrives. The value of the number is in comparing scenarios against each other — public against private, four years against two plus transfer, saving from now against starting in five years — rather than in treating any single run as a forecast.