Research for Matt and Mike. Not legal advice.
What the law allows, what it does not, why the sequencing idea makes things worse, and the six questions to put to counsel.
Yes, but only if you can prove she would have gone anyway. The law does not give someone on leave a shield. It gives them a floor: they are no worse off than if they had been at their desk. If the position was going regardless, it goes.
馃攽 The catch is who has to prove it. Arukah does. Not her. In an ordinary dismissal the employee must show the employer acted wrongly. Here it reverses: the employer carries the burden of proving the employee would have been laid off during the leave anyway.
No. Three separate rules each stop it on their own, and the third makes the attempt worse than the thing it was trying to avoid. Detail in section 3.
The governing regulation is 29 CFR 825.216. Its first line: "An employee has no greater right to reinstatement or to other benefits and conditions of employment than if the employee had been continuously employed during the FMLA leave period."
If an employee is laid off while on FMLA leave, "the employer's responsibility to continue FMLA leave, maintain group health plan benefits and restore the employee cease at the time the employee is laid off". So a genuine reduction can reach her.
The burden sentence, which is the whole ballgame: the employer "would have the burden of proving that an employee would have been laid off during the FMLA leave period and, therefore, would not be entitled to restoration." Arukah has to prove a hypothetical: that this would have happened anyway. If the evidence is thin, the tie goes to her.
Two Engagement Specialists in Mobile Crisis are on the revised list. If those are the only two, her whole role category in that programme is going, there is no colleague left to compare her against, and the comparison opposing counsel most wants to draw has nothing to point at. That would be close to the best fact pattern available.
馃 We do not know that they are the only two, and the earlier version of this page said we did. The list shows who is leaving. It says nothing about who is staying. If Mobile Crisis has four Engagement Specialists and two are being kept, the fact reverses: a comparator appears, and "why her and not the other two" becomes the case.
馃攽 One question to Ruth settles it: how many Engagement Specialists are in Mobile Crisis in total? Ask before anyone builds an argument on it.
Mobile Crisis is being kept. Matt was explicit that Crisis is not being scrapped. So the programme survives while both people in one of its roles go. That is a defensible structural choice, and it is also the first thing a lawyer will ask about, because keeping a programme and removing a whole role inside it needs a stated reason. Not fatal. It does mean the reason cannot be left implied.
The two have very different tenure. One started in April 2026, the other in April 2024. Eliminating both is cleaner than eliminating one, and it supports the role-category reading rather than undercutting it. Worth knowing before it comes up.
Add it. But a badly built reason column is worse than no reason column, because everything written becomes evidence and none of it can be taken back.
"Funded by a grant that ended 30 April" is provable from the grant file. There is no counter-argument. It is the same reason whoever reads it.
"Not a strong performer" is an opinion, and it opens a fight over whether it is pretext. It drags every performance review that person ever had into the matter. 馃 And if her last review was positive, the reason column has just proved the case against Arukah. A subjective reason converts a defensible restructure into a credibility contest, which is the contest employers lose.
Where a role genuinely was a judgement call, say so in structural terms and stop there: "supervisory layer eliminated, remaining staff report to the clinical director". That is a fact about the org chart, not a claim about the person.
鈿栵笍 Counsel reviews the format before the column exists, not after it is filled in. This will be the most discoverable document Arukah creates this year. Building it and then showing a lawyer is backwards, and unwinding it is not possible once it is written.
Cut the other ten, fall under 50 employees, then she is no longer covered. It is the obvious move and all three of these stop it independently.
29 CFR 825.110(e). Whether 50 employees work within 75 miles is decided when the employee gives notice of the need for leave. Once she is determined eligible, "the employee's eligibility is not affected by any subsequent change in the number of employees", and an employer "may not terminate employee leave that has already started if the employee count drops below 50."
The regulation contains an example that is almost this exact question: an employer with 60 employees in August who expects to be at 40 by December must still grant the leave. Her rights were fixed on the day she gave notice. Nothing done in September moves them.
29 CFR 825.105. An employer is covered if it had 50 or more employees for 20 workweeks in the current or the preceding calendar year. Arukah has been at roughly 55 all through 2026.
So even if the headcount drops to 44 in September and never recovers, Arukah remains a covered employer for the rest of 2026 and for all of 2027. The plan does not buy relief this year, and it does not buy relief next year either.
29 CFR 825.220 prohibits "manipulation by a covered employer to avoid responsibilities under FMLA." The regulation gives examples, and one of them is keeping worksites below the 50-employee threshold for eligibility. Sequencing terminations to get under the threshold is the named conduct, not an edge case.
馃 This is why the idea is worse than the thing it avoids. A straightforward reduction that includes her is defensible and might well be won. A reduction deliberately ordered to strip her coverage creates the intent evidence that the first version lacks. It converts an argument about business necessity into an argument about motive, which is the argument employers lose.
It also costs the good-faith defence. An employer avoids liquidated damages, which double the award, only by showing it had reasonable grounds to believe it was not violating the Act. Once a sequence exists in writing, that defence is gone.
And the practical point. Any plan of this shape has to be discussed with somebody, in an email or a meeting, before it can be carried out. That conversation is discoverable. The instruction to sequence would become the single most damaging document in the case.
Illinois has no state FMLA equivalent for private employers, so the federal rules above are the leave analysis. Four state statutes still touch this.
The IHRA covers disability discrimination at employers with one or more employees, and its definition of disability is broader than the federal one: a determinable physical or mental characteristic arising from disease, injury, congenital condition or functional disorder.
If her FMLA leave is for her own serious health condition, that condition is very likely a disability under the IHRA. That opens a second claim with a different test, a different agency and no 50-employee threshold to argue about. Winning the FMLA analysis does not dispose of it.
Separately, the IHRA's general employer definition was expanded from 15 employees to one or more, so no size argument is available on any protected ground.
The revised list with all eleven, the preliminary list for comparison, her leave dates and notice date, the funding source for each of the eleven roles, and the headcount by worksite. An hour of preparation turns this from an open-ended consultation into a scoped question.
Her position stays out of any document that leaves the building. The board packet describes eleven positions by function, which is accurate and does not identify her. That is deliberate and it should hold until the legal question is closed.
The instinct is to leave her off 2 September and deal with it when she returns. That is the wrong move and it is worth understanding why, because it feels like the cautious one.
The regulation explicitly contemplates a genuine reduction reaching someone on leave. Including her in a company-wide restructure, applied by criteria, on the same day as ten other people, is the defensible version. Pulling her out and deciding in November creates an individual decision about one person who has just come back from protected leave, with nothing to compare it to. That is a retaliation claim rather than a restructure, and it is a harder one to win.
What can move is the timing of the notice, not her inclusion. How and when someone on leave is told, what happens to health coverage, and what Illinois requires on the final cheque are mechanics counsel should set. That is a different question from whether her position is in the plan.
Her position stays out of anything that leaves the building. The board packet says eleven positions across seven functions, which is accurate, sufficient for the board to give feedback, and identifies nobody. No director needs a name to comment on a restructuring plan.
Research, not legal advice. Every regulation cited here was read at source and is quoted rather than paraphrased where the wording matters. An Illinois employment lawyer still has to clear this before anyone acts. The purpose of this page is to make that call short and specific instead of open-ended.
Sources. 29 CFR 825.216, limitations on reinstatement 路 29 CFR 825.110, eligible employee 路 29 CFR 825.105, counting employees for coverage 路 29 CFR 825.220, protection for employees who assert FMLA rights 路 US Department of Labor, FMLA FAQ 路 SHRM, layoffs during leave require adequate documentation 路 Jackson Lewis on the Illinois Human Rights Act expansion 路 Illinois Department of Labor, Paid Leave for All Workers Act 路 Illinois Department of Labor, VESSA