A child support income calculation derived one way is a guess. A figure that two independent derivations converge on is a position you can hold across the table without hedging, in a caucus, at a hearing, or in a letter you will have to live with. That is the entire method, and most support fights are lost before the guidelines are ever opened because nobody ran the number twice.
The Florida child support guidelines produce an output from inputs. The guidelines do not audit the inputs. Auditing them is your job, and a child support income calculation is where the money in these cases actually moves.
Treat the last sworn financial affidavit as a hypothesis
Start from the most recent sworn affidavit, because it is the baseline both sides and the court will reference. Then stop treating it as a fact. Financial affidavit review means testing the document against the records behind it, not summarizing it. An affidavit is a snapshot taken by an interested party at a moment that has usually passed, often prepared in a hurry, sometimes prepared from memory, and frequently never updated after the thing that changed everything.
Rebuild the income picture before you touch any support number at all. Negotiating from stale affidavit figures spends the advantage of the correction before you have it, because you burn the session arguing about a number you are going to abandon. Worse, it anchors your client to a figure that will not survive contact with the pay records, and clients remember the first number they were told far longer than they remember why it changed.
Year-to-date is the wrong input for a child support income calculation after a mid-year raise
Where a raise, a promotion, a rate change, or a schedule change has landed mid-year, year-to-date figures understate current earning capacity by construction. Year-to-date blends the old rate with the new one. It is an average of a period that no longer exists. The more defensible approach annualizes the current pay period by the number of periods in the year, which describes what the party is actually earning now and going forward, which is the question support answers.
Then state the choice of method and the reason for it inside your work product, in a sentence. The method is what gets challenged. Not the arithmetic. Nobody at the table is going to catch you multiplying wrong, but everybody is going to ask why you used one input instead of another, and an analysis that already answers the question in writing survives the objection instead of absorbing it.
The second derivation has to use inputs the first one never touched
Derive the figure again from unrelated inputs, then compare. One derivation takes the rate on the stub and annualizes it by pay periods. The other takes reported hours per period, multiplies by the rate, and annualizes that. Two paths, and the inputs barely overlap. Where they converge, you assert the number flatly. Where they diverge, the divergence tells you which input is wrong, which is worth as much as the convergence and usually more.
Here is an illustrative example, invented for this article and matching no file. Suppose a stub shows a semi-monthly gross of four thousand one hundred, which annualizes to ninety-eight thousand four hundred across twenty-four periods. The same stub reports eighty-six hours at a stated hourly rate that yields something closer to eighty-nine thousand annualized. That gap is not a rounding artifact. Something in there is a stipend, a differential, a recurring bonus, or an hours figure that does not reflect what is actually being worked, and you now know precisely where to look and precisely what to ask for. A single derivation would have handed you one clean number and no idea it was wrong.
Imputation is a separate argument, not a shortcut in the arithmetic
Keep imputed income in its own lane. Correcting a stale or understated figure using the other side’s own records is an accuracy exercise, and it is nearly unanswerable when the records are theirs. Imputed income is an argument about capacity that carries its own evidentiary burden and its own proof problems. Blur the two and you convert an unanswerable accuracy point into a contested imputation fight, which is a trade nobody should make voluntarily. Fix the record first. Argue capacity separately, if the record still calls for it.
Proportional allocation is the default; an even split is a transfer
Allocate variable child-related costs in proportion to income, not evenly. An even split of uncovered medical and similar recurring expenses looks even and functions as a transfer from the lower earner, who is paying the same dollars out of a smaller pool. If the parties are dividing anything recurring and variable, the percentages should track the income shares the guidelines already established, because using two different ratios in one agreement guarantees the inconsistency gets argued later.
Related precision problem. Where an insurance premium covers only some of the children, attribute the credit to the covered children rather than applying a blanket family figure. Applying the whole premium against the whole sibling group distorts the calculation in one party’s favor by accident, and the error is invisible until someone rebuilds the worksheet.
Settle the principle before you argue the accounts
On debt characterization, force consistency of treatment before anyone argues an individual account. Identify the governing prior agreement first, because a temporary allocation frequently answers the question already, and a party who has been relying on that allocation for months cannot selectively depart from it now. Then state the principle and make each side apply it to itself. My version of the principle is blunt.
Either all the credit card debt is marital or none of it’s marital. It’s not piecemeal.
Then show each side the arithmetic its own principle produces. A party holding the larger balance in a category abandons the position surprisingly often once the math is on the table in front of them, and they abandon it faster when they arrived at the principle themselves. This is the same discipline as counting days rather than trusting labels in a schedule, which I wrote about in the piece on letting an outside calendar define a timesharing period. Principle first, then arithmetic, then the individual items.
Two related habits worth adopting. Demand reimbursement claims as legible itemized lists, with images as corroboration rather than as the primary record. A folder of unreadable photographs proves nothing, will not survive a hearing, and cannot be responded to in good faith. Say so while there is still time to cure it, not in a pretrial filing. And correct a wrong number the other side is relying on immediately, even when the correction helps them, because a settlement built on a figure one side later discovers was wrong does not hold, and you will spend more rebuilding it than you ever gained. The same instinct drives what a proper fee audit checks and in what order.
None of this is legal advice, and the analysis in any given matter is yours to run and yours to sign.
If running a child support income calculation twice, on every case, sounds like something you do not have hours for, that is exactly the work I built the financial systems to carry. Tested inside an active Florida family law firm on real cases with real deadlines before it was ever packaged for anyone else. Bring a case and we will scope it together on a custom build strategy session, so you can make more money and work less.