Methodology

Who writes this

I build and maintain Maternity Tally, and I check the numbers myself. Every wage cap, replacement rate, eligibility threshold and week count on this site was read off a state statute, an agency rate announcement, or an agency's own benefit estimator — not copied from another calculator or a summary article.

I am not a lawyer and not an HR professional, and nothing here is legal advice. I mention that plainly because the alternative — implying a credential I do not hold — would be worse than having none. What I can defend is the method: primary sources only, a documented schedule for re-checking them, no invented figures when an agency has not published one yet, and a public record of the errors I find, including my own.

That last part matters more than it sounds. These figures move on schedules that differ by state — some in January, some in July, one on the last Sunday in October — and agencies sometimes leave a superseded number published on one page while the current one appears on another. Noticing that is most of the work; writing down what changed, and when, is the rest. Every correction I have made to this site is dated and listed on the corrections page, including the ones where the site was wrong.

Corrections are welcome and acted on. If a figure here looks wrong, send it with a link to the primary source — that is the fastest way to get it fixed, and the correction will be listed with its date on the corrections page.

Where the numbers come from

Every value in the calculator traces back to an official primary source. Federal rules come from the Family and Medical Leave Act and the federal labor regulations that implement it. State programs come from each state's own department of labor or workers' compensation board — California's EDD, New York's Workers' Compensation Board, New Jersey's Department of Labor, and so on for every state we cover.

When you finish the calculator, the "How we calculated this" section at the bottom of your results lists every program that contributed to your timeline alongside a link to its primary source — so you can verify any number yourself.

The data set covers all 50 states plus the District of Columbia. As of 2026, sixteen jurisdictions have full state-specific data: California, New York, New Jersey, Massachusetts, Washington, Connecticut, Minnesota, Colorado, Oregon, Rhode Island, Delaware, DC, Maine, Pennsylvania, Texas, and Florida — chosen to span the full policy spectrum, from Oregon's industry-leading 100% replacement rate for low earners and Rhode Island's 30-week medical leave (the longest in the US, since 1942) to DC's unique 100% employer- funded program (workers pay $0), Delaware's $900 weekly cap (the lowest in the US, with FMLA-equivalent eligibility), and Maine (the newest program, with benefits starting May 1, 2026) to Texas/Florida operating at just the federal floor. Massachusetts, Connecticut, Minnesota, Colorado, Oregon, and Maine all have built-in state-level job protection at any employer size; Washington, Rhode Island, Delaware, and DC's state job protection is narrower (20-50+ employees, mirroring or close to FMLA). Two more states have enacted paid family leave programs that aren't paying benefits yet, so they're not in the calculator: Maryland (benefits January 2028) and Virginia (benefits December 2028). Virginia in 2026 became the first Southern state to enact one.

How the calculation works

You answer a short series of questions: which state you work in, your role in welcoming the child, how long you've been at your job, how many hours you've worked in the past year, your employer's size, your salary, your pay frequency, your delivery type, and whether you're carrying multiples. A few of those only apply if you're the birthing parent; the calculator skips what doesn't fit.

From those answers, the calculator works out which state and federal programs you qualify for, applies each program's specific rules, and produces a week-by-week timeline showing each phase of your leave and how much you'll be paid each week, dated action items (when to file your FMLA paperwork, when to file your state claim, when to add your baby to insurance), and a "How we calculated this" section citing the primary source for every program contributing to your timeline.

The calculator honors each state's specific rules. California's tiered structure pays low earners 90% of wages and higher earners 70%. New York's short-term disability program is capped at $170 per week — a number unchanged since 1989. New Jersey replaces 85% of wages up to $1,119 per week. When a state's weekly cap would limit your benefit below the percentage replacement, the calculator shows you the actual capped amount, not the theoretical percentage.

How the data stays current

State paid family leave parameters typically update annually around January 1, when new wage caps and contribution rates take effect — but mid-year legislative changes happen too. To keep stale numbers out of the calculator, an automated monitor re-reads the state government pages behind them on a schedule.

Every page that carries a benefit cap or minimum is re-checked weekly. That is the tier that can hurt you if it goes stale, so all twenty-three of those sources run on a seven-day cycle. Pages that carry no dollar figures — statutes, eligibility explainers, program overviews — are checked less often, because they change on the timescale of legislation rather than of a rate announcement.

Some state websites refuse to be read by a program at all. Massachusetts returns a “this page is forbidden” response to an ordinary automated request, and New Jersey's civil-rights pages sit behind a security service that answers with a challenge page instead of the content. Both are reached by driving a real browser from a separate machine on a residential connection, which is what those sites will accept.

The more important half of that is what happens when it does not work. A blocked page still returns something — a short error page, or a “checking your browser” interstitial — and a naive monitor will happily record that as the page's new content, then report “no change” every week thereafter while the real figures move underneath it. That is a monitor that has failed while showing a green light. Ours refuses to accept such a page as content at all: it is recorded as a failed read and surfaced for a human, never as confirmation.

When a change is detected, it's never applied to the calculator automatically. It's flagged for human verification against the primary source first. This trades some freshness for accuracy — a real numeric change might take a few days to land after the state publishes it, but the alternative (silently propagating an unverified change) could send thousands of users the wrong numbers.

Seventy-five government sources are monitored this way — statutes, agency rate announcements, benefit estimators and the rates tables that carry the real numbers. Several of those pages carry no dollar figures at all, so the monitor also watches the press-release indexes where each year's announcement appears, because agencies publish every year's release at a brand-new address.

Scheduled audits, timed to when each state actually changes. The obvious assumption — that benefit figures reset every January 1 — is true for California, New York, New Jersey, Massachusetts and Washington, and false for a lot of the map. Oregon's minimum and maximum turn over in late June on a benefit-year basis. Rhode Island's and Maine's move on July 1. The District of Columbia has two separate knobs on two different dates: its bend point follows the minimum wage each July 1, while its maximum is reviewed each October 1. Minnesota's maximum is keyed to the last Sunday in October. So the audits run six times a year rather than annually: December 1 and January 5 for the January cohort, June 1 and July 15 around the midsummer turnovers, and September 28 and October 19 for the autumn ones.

A known change gets an alarm, not a reminder. Legislated changes are often verified months before they take effect, which creates its own hazard: the research is done, the date is known, and then the date quietly arrives. Every future-dated figure on this site is therefore registered with the date it becomes operative, and that register is checked daily. A page-monitor cannot catch this on its own, because nothing on any government website has to change for a law to take effect.

When a number is not published, it stays blank. Minnesota's next maximum is a live example: state law requires the commissioner to calculate it by June 30, but not to publish it, so for several months the figure legally exists and appears nowhere public. A plausible number derived from wage growth would look just as authoritative on this page as a real one. We leave it unset instead and say so, because a figure that looks verified and is not is the worst thing a calculator like this can hand you.

Who the question flow is built for

The calculator is built around four parent roles: birthing parent, non-birthing parent or partner, adopting parent, and fostering parent. State paid leave laws use eligibility rules that map cleanly onto these four roles regardless of family structure — the underlying laws rarely distinguish between, say, a married heterosexual couple and a same-sex couple, or between an opposite-sex partner and a domestic partner. The flow is built to honor that: choose the role that matches your situation and the math is the same.

Concretely:

• Same-sex couples where one partner carries the pregnancy: the carrying partner selects "birthing parent" and the other selects "non-birthing parent / partner". Both qualify for the appropriate state programs — the birthing partner for medical recovery + bonding, the non-birthing partner for bonding leave in states that have it.
• Donor-cycle families (IUI, IVF, donor egg or sperm): the person carrying the pregnancy is the birthing parent for benefits purposes regardless of the genetic relationship to the child. State programs evaluate based on who carries and gives birth, not on biological parentage.
• Intended parents via surrogacy or gestational carrier: select "adopting parent". State paid family leave programs that cover bonding generally cover intended parents under the same rules as adoptive parents, even when no formal adoption is required. California PFL and Massachusetts PFML, for example, both cover intended parents for bonding leave.
• Single parents by choice or circumstance: the role options work identically; nothing in the calculator assumes a two-parent household.

State-specific copy uses "you" and "birthing parent" rather than "mom" or "the dad". The state-program names themselves (e.g., "Paid Family Leave") are used as states officially title them without re-gendering.

Baseline assumptions

State-specific landing pages display a typical case: a $75,000-per-year birthing parent at a 50+ employee company, with 12+ months on the job, full-time hours, and vaginal delivery. This gives a quick sanity check before you customize for your own situation in the calculator.

Recovery weeks default to 6 for vaginal delivery and 8 for C-section — the standard medical recovery periods most state programs accept. If your provider certifies a longer recovery, your actual paid period may extend.

What the calculator doesn't model

• Federal income tax and Social Security withholding. Wage replacement is taxable income (fully or partially, depending on the state). The dollar figures shown are gross — before tax is taken out.
• Employer "top-up" plans. Some employers supplement state paid leave to 100% of your pay. We ask about employer-paid parental leave separately, but the precise interaction with state programs varies by employer policy.
• Multi-state work arrangements. If you live in one state and work in another, the calculator uses the work state's rules — correct for most state programs, but with edge cases for remote workers across state lines.
• Self-employed and 1099 contractors. Most state programs are opt-in for self-employed workers and require enrollment before pregnancy. The current calculator covers W-2 employees.
• Federal employees. Federal employees have their own Paid Parental Leave program (12 weeks at 100% of pay) and don't use state programs. The calculator handles this correctly when "federal" is selected as employer size; a dedicated federal-employee page is planned.

How to report a data error

If a value looks wrong for your state — especially if you have direct experience with the program — please get in touch with a link to the primary source. Public correctness is the whole point.