Two 2026 Recalls, One Household Ledger: What Registering Your Products Is Worth Over Five and Ten Years

Updated September 27, 2026. Every number below was read on that date from the agency, regulation or statistical release named at the end of this page. No retailer price or store rating appears anywhere in it.

Photo: Wikimedia Commons, CC BY-SA 3.0

Two recalls from one August, and the question they put to a household budget

In the space of ten days last month, two things a typical American family owns were pulled back by the federal government. One was a bag of lettuce. The other was an infant car seat. Neither event is remarkable on its own; between them they raise the only recall question that has a dollar answer, which is whether the small, dull habit of registering things and reading recall notices is worth what it costs you in time.

Start with what each recall actually was, in the agencies' own numbers.

The lettuce. CDC and FDA both opened outbreak pages for a multistate Cyclospora outbreak linked to iceberg lettuce. CDC's final update, dated September 11, 2026, records 12,883 illnesses, at least 570 hospitalizations and 2 deaths, both in Michigan, across 21 states, with illness onsets running from June 14 to August 17, 2026. The implicated product was iceberg lettuce from Taylor Farms de Mexico sourced in central Mexico, voluntarily withdrawn on July 17, 2026. CDC's wording on closure is exact: "This outbreak has ended. The number of recent infections linked to the outbreak has significantly declined." FDA's own investigation page gives a first onset of June 22 rather than June 14, and its traceback work at the growing and processing sites was still described as open when this page was written. The epidemiological outbreak is closed; the regulatory investigation is not. Both statements are true at once, and the nine-day difference in the first onset date is a live discrepancy between two federal pages, not a typo on one of them.

The car seat. This is the correction the earlier version of this page needs. Child restraints are regulated by the National Highway Traffic Safety Administration, not by the Consumer Product Safety Commission, and the campaign number is 26C006. Britax Child Safety's defect information report, filed with NHTSA on August 20, 2026, covers 114,660 potentially affected units across 74 model variants of the B-SAFE Gen2, B-SAFE 35, B-SAFE Ultra and Endeavours lines, manufactured between August 17, 2020 and May 16, 2023. The defect is not the handle as such. In the filing's language, "extended use and certain wear conditions may cause a screw in one side of the car seat's hub assembly to loosen, allowing for a partial detachment of the carry handle," creating "a fall hazard and potential risk of injury to the child." Britax estimated the actual defect rate at 1 percent of the affected population, which is about 1,147 seats. The remedy, a bracket kit for the hub mount, was still in development when the report was filed, and the reimbursement window for consumers' own out-of-pocket repairs ran through September 24, 2026 — three days before this page was updated.

The clock the regulations set, and the gap inside it

Most people assume a recall is an announcement. It is a sequence with legal deadlines, and the deadlines are asymmetric in a way worth knowing.

• 24 hours. Under 16 CFR 1115.14, a firm must report to CPSC within twenty-four hours of obtaining information that reasonably supports the conclusion that its product contains a defect creating a substantial risk. Weekends and holidays are excluded from the count, and the firm's own prior investigation "should not exceed 10 days" without justification.
• 3 business days. Under 49 CFR 577.13, a vehicle or equipment manufacturer must notify dealers and distributors within three business days of filing its defect report with NHTSA.
• 60 days. Under 49 CFR 577.7, the letter to the owner must be furnished no later than sixty days from the date of that same filing.

Subtract the second from the third and you get the number that matters to a parent: the retailer can know up to 57 days before the letter is legally required to reach you. Nothing in that is improper, and the manufacturer may well move faster. But the regulation only guarantees you the slow end of it, which is the entire argument for not relying on the letter.

On the food side the classification, not the clock, is the thing to read. Under 21 CFR 7.3, a Class I recall means "a reasonable probability that the use of, or exposure to, a violative product will cause serious adverse health consequences or death"; Class II covers effects that are "temporary or medically reversible" or where serious harm is remote; Class III is "not likely to cause adverse health consequences." A Class I notice is the one that justifies interrupting your evening.

Five years of the government grading its own recall system

The question of whether recall vigilance is worth your time turns on a number most people have never seen: how many recalled products ever actually get fixed, returned or discarded. Both agencies publish it, and neither number is flattering.

Measure	Figure	Source and period
CPSC recall response rate, FY2021 to FY2025	32% / 16% / 24% / 51% / 44.5%	CPSC FY2025 Agency Performance Report; FY2025 target was 43%
Durable infant and juvenile products returned or repaired	33% average across 21 recalls, more than 2 million units	CPSC report to Congress under Section 104(d), February 2016, covering Sept 2010 to May 2015
Registration cards actually returned for those products	13% average, range 1% to 45%	same CPSC report
Vehicle recall completion, weighted average, recalls begun 2012 to 2022	65.8%; about 87% for vehicles 0 to 3 years old	NHTSA Report to Congress on Improving Vehicle Safety Recall Completion Rates, April 2026

Two warnings belong with that table. CPSC's own footnote says the calculation method for the response rate was changed at the FY2024 midyear review to an "average correction rate," described as "less variable and more representative than the prior calculation method" — so the jump from 24 percent to 51 percent between FY2023 and FY2024 is partly a change in the ruler, and the five figures should not be read as a clean trend. And NHTSA's completion-rate dataset explicitly excludes tires, child seats and equipment, which means there is no published federal figure for how many recalled car seats get fixed. The 44.5 percent and the 33 percent are the closest available proxies, and they are proxies, not measurements of this.

The Government Accountability Office has put its own caveat on top. GAO-21-56 (November 2020) found that CPSC's units-corrected metric likely overstates effectiveness, because a consumer who simply throws a recalled product away is invisible to it, and that only 61 percent of firms filed more than three quarters of their required monthly progress reports for recalls closed between February 2016 and May 2020. GAO-24-106356 (January 2024) separately found NHTSA's research into what actually drives repair rates had lapsed.

Plan A and Plan B, priced over five and ten years

Here is the comparison this page exists to make. Plan A is the registered household: fill in the card for every durable product, subscribe to the federal alert lists, skim them weekly, and check the pantry whenever a Class I food notice lands. Plan B is doing none of that and dealing with problems as they arrive.

Plan A costs time, and time has a published price. BLS reports average hourly earnings for all private employees of 37.75 dollars for August 2026, and median usual weekly earnings for full-time workers of 1,251 dollars for the second quarter of 2026, which is 31.28 dollars for a forty-hour week. Both are used below, because the choice between them moves the answer by about 17 percent and pretending otherwise would be dishonest.

The time budget, built from the actual tasks: four registration cards a year at five minutes each is 20 minutes; a two-minute weekly skim of recall email is 104 minutes; four pantry checks a year at ten minutes is 40 minutes. That is 164 minutes, or 2.73 hours a year.

	Per year	Over 5 years	Over 10 years	Note
Plan A time cost at $37.75/hour	$103.06	$515.30	$1,030.60	average hourly earnings, all private employees
Plan A time cost at $31.28/hour	$85.40	$427.00	$854.00	median full-time weekly earnings, divided by 40
Plan A cash cost	$0	$0	$0	every federal alert list is free
Plan B cost when nothing goes wrong	$0	$0	$0	this is the honest base case and it is the likely one
Plan B cost of one avoidable foodborne case	$543 to $16,668 depending on pathogen, 2023 dollars			USDA ERS per-case cost estimates
Plan B cost of one missed child-restraint defect	not quantifiable from any published federal figure			see the section below

The multiplication in that table is mine; the hourly rates, the per-case costs and the response rates are the agencies'. What the table shows is that Plan A costs a household somewhere between 427 and 1,031 dollars of time over a decade, and that the entire case for spending it sits in two rows that cannot be filled in with a single number.

What the avoided-loss column can and cannot be filled with

USDA's Economic Research Service, in its cost estimates last updated June 16, 2026, puts the total annual US cost of foodborne illness from 31 major pathogens at 74.7 billion dollars in 2023 dollars. Its per-case figures span four orders of magnitude: norovirus 543 dollars, STEC 2,865 dollars, Campylobacter 13,405 dollars, nontyphoidal Salmonella 16,668 dollars, Listeria monocytogenes 2,504,296 dollars. CDC, using a separate methodology, estimates roughly 48 million illnesses, 128,000 hospitalizations and 3,000 deaths a year nationally.

ERS publishes no per-case figure for Cyclospora, which is the pathogen in the recall this page started with. So the single most relevant cell in the table cannot be filled from the best available federal source. The honest way to bound it is the outbreak's own arithmetic: 570 hospitalizations out of 12,883 illnesses is a 4.4 percent hospitalization rate, and two deaths in 12,883 is about 1 in 6,400.

One more caution, because it will trip up anyone checking this. ERS's own website still hosts a 2016-vintage chart page giving very different per-case figures for the same pathogens — Salmonella at 3,568 dollars rather than 16,668, and an aggregate burden of 15 billion dollars rather than 74.7 billion. That is not two agencies disagreeing; it is one agency's old page sitting alongside its current data product, both indexed and both looking official. Check the update date before quoting either.

The case where being careful costs you money

Recall vigilance is not free of downside, and the two August recalls illustrate both kinds.

The food kind is waste. BLS puts average household spending on fresh vegetables at 355 dollars a year and on food at home at 6,224 dollars (Consumer Expenditure Survey, 2024, published December 2025). A household that discards lettuce on every leafy-green headline rather than on a matching brand, lot and best-by date can spend a meaningful fraction of that 355 dollars throwing away food that was never implicated. The Taylor Farms withdrawal was specific: named brands, named pack sizes, best-by dates from July 18 to August 3, 2026. Reading the specifics is the difference between a recall costing nothing and costing a week of produce.

The durable-goods kind is worse, because it can leave you without the thing. When Britax filed on August 20, the repair kit did not yet exist in final form. A parent acting immediately on the notice had no remedy to accept, and the reimbursement route for their own out-of-pocket fix closed on September 24. "Act tonight" is good advice for checking whether you are affected. It is not always available as advice for fixing it.

The numbers nobody publishes, which is why this ends in a judgement rather than a total

Four gaps kept the ledger above from closing. There is no federal figure for what share of recalled child seats are repaired or replaced, because NHTSA's completion dataset excludes them. There is no official US government statement of what a replacement car seat costs; grant programs publish bundled totals that mix seats with installation clinics and outreach, which is not a per-seat price, and this page will not quote a retail one. NHTSA's public recalls API returned nothing for campaign 26C006, so the equipment recall could not be cross-checked against the agency's own machine-readable index, only against its filed PDF. And the injury count itself is unsettled: the filing extract gives three reports of minor injury since 2016, while at least one consumer-testing organisation reporting on the same recall says four.

What the published figures do support is narrower and firmer than a dollar total. Between a third and a half of recalled consumer products are ever corrected. Thirteen percent of registration cards for durable infant products come back. The regulation lets your retailer know up to 57 days before your letter is due. Those three facts together say that the return on Plan A is not the expected dollars saved — on the numbers, that is probably less than the time costs — but the removal of a specific failure mode: being in the 55 percent who never hear, on the one occasion when the product in the house is the one with the loose screw. Buying that for under 3 hours a year is a bet on the tail, not on the average, and it should be made with that understood rather than on a promise of savings.

A one-evening setup, in the order the regulations justify

1. Subscribe to the federal lists first, because they cover products you have not bought yet. Recalls.gov carries CPSC, NHTSA, FDA, USDA and EPA in one subscription; NHTSA's own list is the faster route for vehicles and vehicle equipment.
2. Register the child restraints tonight. Under 49 CFR part 588, the card exists specifically so the manufacturer can reach you during a recall campaign, and it asks only for the model name and number, month and year of manufacture, and your name and address.
3. Write the manufacture date somewhere you will keep it. Part 588 requires manufacturers to retain registration records for six years from the date of manufacture. A seat is often in service longer than that, so the record that matters most in year seven is yours, not theirs.
4. Match on lot and date, not on headline, for any food notice. The recall notice names brand, size and best-by range; nothing outside that range is implicated.
5. Check the classification before reacting. Class I under 21 CFR 7.3 is the tier that carries a reasonable probability of serious harm.
6. Re-check open recalls once a year, on a date you already remember. Completion rates fall as products age, and an unrepaired recall does not expire.

Tip: The pantry half of this checklist is easier when you can actually verify your storage temperature rather than guess at it, since a matching lot number is only half the question. A simple refrigerator thermometer costs less than one discarded grocery run. (These are Amazon Associate links — we may earn a small commission on qualifying purchases.)

Photo: Wikimedia Commons, CC BY-SA 3.0

Related on this site

• The 2026 measles surge: a working parent's back-to-school checklist
• The Meta teen-safety settlement: a parent's practical plan

The federal pages these figures were read from

• CDC, Cyclospora outbreak linked to iceberg lettuce, final update September 11, 2026 — 12,883 illnesses, at least 570 hospitalizations, 2 deaths, 21 states.
• FDA, Investigation of multistate outbreak of Cyclospora illnesses, iceberg lettuce, July 2026 — the recall detail and the differing first-onset date.
• NHTSA, Part 573 defect information report, campaign 26C006, filed August 20, 2026 — 114,660 units, the model and date ranges, the defect language and the 1 percent estimate.
• CPSC, FY2025 Agency Performance Report — the recall response rate series and the footnote on the changed calculation method.
• CPSC, Report to Congress on recall notification effectiveness, February 2016 — the 33 percent and 13 percent figures.
• NHTSA, Report to Congress: improving vehicle safety recall completion rates, April 2026 — 65.8 percent, and the exclusion of child seats.
• GAO, GAO-21-56 and GAO-24-106356.
• eCFR, 16 CFR 1115.14, 49 CFR 577.7, 49 CFR 577.13 and 21 CFR 7.3.
• USDA Economic Research Service, Cost estimates of foodborne illnesses, updated June 16, 2026 — 74.7 billion dollars and the per-case figures, in 2023 dollars.
• CDC, Food safety facts and statistics — 48 million illnesses, 128,000 hospitalizations, 3,000 deaths.
• Bureau of Labor Statistics, average hourly earnings, August 2026; usual weekly earnings, second quarter 2026; Consumer Expenditure Survey, 2024.

This page summarises published federal documents and is not personalised legal, medical or financial advice; it is written by no one holding a law licence, a clinical credential or a child passenger safety technician certification. Recall status, remedies and reimbursement windows change, and the federal rules cited apply in the United States as of the dates given.