Why Product Recalls Make You Less Safe

Genuinely dangerous products should be pulled from the shelves. But government recalls sometimes punish manufacturers for vague problems and blatant consumer misuse, actually reducing public safety. Even the word "recall" turns out to be defective.

By Dan Koeppel

Elmo was in danger. In the video Bryan Dussault posted on YouTube in February 2010, the Sesame Street favorite is strapped into a child-safety seat. Dussault tightens the five-point harness and then abruptly yanks the unit's shoulder straps, which are supposed to be snug against the infant passenger. The harness straps loosen. Dussault pulls the shaggy toy away as if it were being tossed, unprotected, in a high-speed collision. To anyone viewing the clip, but perhaps especially to anyone who travels in a car with a child, the scenario is startling and scary.

The series of events that led Dussault to film the sequence began on a freezing Chicago day earlier that winter. At the time his son was 3 years old. It was Grandma's turn to pick up the boy at preschool. But there was a problem. "My mother-in-law called me," Dussault says, "and said she couldn't get the car-seat straps to tighten up."

Dussault drove to the school and took his child home himself, along with the car seat that had apparently failed. He owned the same item—the $79 Vantage, purchased at Walmart and sold under the Safety 1st brand name. In his garage Dussault tested his own seat; again, the straps loosened. "I knew that if I got two doing that," he says, "there were probably thousands out there doing the same thing."

Dussault contacted the seats' manufacturer, Dorel Juvenile Group. Based in Columbus, Ind., Dorel is the world's largest maker of children's car seats. It sells about 8 million child-restraint systems—the official term for car seats—each year. But shoppers won't find the Dorel name in stores. Instead, the company's goods appear under a range of recognizable brands including Disney, Cosco, Eddie Bauer, and Maxi-Cosi.

Dorel offered to exchange Dussault's seats. But when the replacements arrived, Dussault says they failed his test as well. He filed a complaint with the National Highway Traffic Safety Administration (NHTSA). Meanwhile, Dorel asked him to return the original seats so the company could test them, standard procedure when an exchange is offered. Dussault was hesitant. "They were hot on my tail to get these seats out of my hands," he says. The company, he believed, intended to "sweep things under the rug."

To prevent that from happening, Dussault made the video. He contacted the news media. A local NBC affiliate dispatched a crew. The demise of Elmo went viral, and in 2011 Dorel recalled 800,000 car seats. A win for Elmo and kids everywhere?

That's debatable.

Dussault describes his struggle in David and Goliath terms, a concerned parent confronting a faceless corporation whose interests are driven more by sales than by safety. When it comes to products designed to protect children, that position isn't difficult to understand. I'm the father of a 16-month-old, and even before our son arrived, my wife and I did what most expectant families do: We went shopping.

I was amazed by the number of products available and dismayed by how many appeared to be defective—car seats, strollers, toys, cribs, nursery monitors. It seemed impossible to pick a category that wasn't prone to recalls. When my wife and I debated whether to allow our infant to sleep in the same bed with us, we were told that the only safe way to do so was to buy a "positioning device," a product that would prevent a sleepy parent from rolling over and smushing the poor tyke (never mind that parents have been cuddling their children to sleep for millennia). We rejected the advice, and good thing: On our next visit to Babies "R" Us, the positioner aisle was empty. It turned out the device itself could lead to suffocation if a child were pressed against it the wrong way.

At first my impression was similar to Dussault's: a rogue industry. A little research seemed to back that up. NHTSA recalled more than 100 different models of car seats from nearly a dozen manufacturers during the past decade. And media accounts almost always emphasize the profit-trumps-all angle when it comes to product defects. Particularly notable is a 2007 Chicago Tribune story that chronicled Dorel's attempts to prevent an earlier recall. Dorel had argued—correctly but unsuccessfully—that the straps on the car seats weren't up to standards because there were no clear standards when they were manufactured in 2000 and 2001. Other companies, it pointed out, used materials of similar strength and design and hadn't been penalized. The reporters painted the company as uncaring, even conspiratorial—quoting an internal email from a Dorel executive who wrote: "Why? It still sells." The series of investigative reports that included the story won a Pulitzer Prize. And after losing an appeal to the NHTSA in 2010, Dorel was forced to recall about 4 million car seats.

But eventually I began to question whether product defects could really be so rampant. Most of the recalls I had read about appeared to be for items that caused no injuries or where injuries could only be vaguely attributed to a specific problem. As I began to search my home for recalled products—not just stuff aimed at kids, but appliances, tools, electronics, and everything else modern shoppers own and covet—I quickly became overwhelmed. There were so many recalls and so many sources of information on recalled products that I could barely keep track. Did I own dangerous items? Were they really dangerous or just potentially dangerous? And what action should I take, if any?

Trying to sort through it all was impossible, so I gave up. And in that way I became an even more typical American consumer: one unable to make my home truly safer, because the system by which we judge, identify, and correct broken gear is as defective as anything found on store shelves.

In the United States, product recalls are overseen by six federal agencies. The majority come under the jurisdiction of the Consumer Product Safety Commission (CPSC), which handles everything from toys and power tools to appliances. Recalls connected to automobiles—cars, tires, and child-restraint systems—are within the NHTSA's purview. Boats and nautical items are regulated by the Coast Guard. Products containing chemicals, such as house paint and pesticides, are Environmental Protection Agency territory; the U.S. Department of Agriculture (USDA) and the Food and Drug Administration (FDA) handle edibles, pharmaceuticals, and cosmetics.

The division of federal oversight—car seats to the NHTSA, strollers to the CPSC; burgers to the USDA, fries to the FDA—is just the beginning of the knot consumers need to untangle if they're hoping to learn whether a product they own poses a threat. Though there's ostensibly a one-stop database of product alerts, recalls.gov, the website's utility is marred by confusing design and poor information retrieval. Clicking on the Search for Recalls button opens a page with six additional search boxes, each tied to a different agency or category. If you don't know what jurisdiction a product falls under, you'll have little success, which may be just as well: Sometimes products are identified only by inventory-control codes useful to retailers and manufacturers. Photographs are often unavailable. If you do manage to discover a potentially relevant recall, you're given the investigation number assigned by the overseeing agency but not always a link to the investigative details themselves.

Details that are included don't generally provide context for injuries or useful consumer information. For example, the notice for a Task Force electric log splitter, recalled in July 2011, states that two injuries had been reported after operators placed their hands on the handle while the splitter was in operation. The notice instructs consumers to stop using the product and send for free warning labels. "That has got to be one of the weirdest recalls we've ever seen," wrote Everett Snyder on ProToolReviews.com. "The recall itself doesn't do too much to explain the problem or deliver specific (detailed) instructions for safe use of the tool... No, for that you need to wait for the stickers to show up."

This muddle is made worse by the scale and number of recalls. During the past five years, more than 150 million children's items, 110 million household goods, and 9 million pieces of sporting equipment have been recalled by the CPSC alone, according to the recall-tracking group WeMakeItSafer. There are so many recalls—nearly 1500 by all six agencies in 2011—that many consumers have simply stopped paying attention to them. A 2010 Consumer Reports survey found that just under 25 percent of respondents ever bothered to research whether a product they owned had been subject to a recall; of those who knew they owned a recalled product, only about 30 percent actually did something about it. There's even a new term for the phenomenon of ignoring product-safety warnings: recall fatigue. "So many recalls are announced in so many ways that when you hear it on the news, it just doesn't register," says Craig Wilson, vice president of quality assurance at Costco Wholesale Corp. "The perception is that there's a lot of crying wolf going on."

A day after Elmo's mishap hit Chicago airwaves, the NHTSA's Office of Defects Investigation opened case PE10-009—a preliminary evaluation, which doesn't indicate that a recall is necessary or even being contemplated. Though the agency receives about 35,000 consumer complaints each year, just two—both Dussault's—had been registered for the Vantage, which had been highly rated in Consumer Reports.

Preliminary or not, opening an investigation also opens floodgates. The manufacturer of the product in question is required to deliver thousands of pages of data: test results, transcripts of customer-service calls, descriptions of manufacturing procedures, sales figures. The NHTSA announces the inquiry to the public, an evidence-seeking strategy that can lead to hundreds of additional complaints. For the Vantage, it yielded only six new reports, none of which involved injuries. Despite this, the agency moved the inquiry to the next level: engineering analysis. On July 2,2010, case EA10-005 was opened: Dorel was asked to provide complete details on how the seat was built and tested, and whether any changes had been made at the factory that might have coincided with the complaints of strap slippage. Over the next couple of months, Dorel submitted thousands more pages of documentation.

In an 11-page letter to the NHTSA dated Sept. 21, Terry Emerson, Dorel's director of quality assurance, stated that no slippage had occurred in the Vantage's initial rounds of testing—2900 on crash sleds and 2000 mechanical strap pulls. The public investigation attracted 40 additional complaints, representing 0.00005 percent of seats sold. No injuries or deaths were reported. Dorel, Emerson wrote, "does not believe the subject units contain a safety-related defect."

But the company did find that some center front adjusters (CFAs)—the locking and release buttons that secure the straps—could, if excessively dirty, not tighten as well as a new product's (though that could be remedied by making sure the CFA is actually pressed firmly into place). And so, on Feb. 14, 2011, Dorel initiated a voluntary recall, offering what consumer-safety specialists call a sealed fix—a repair kit rather than a full product swap. The kit consisted of a tiny tube of food-grade lubricant; the primary ingredient, canola oil.

The main difference between Dorel's remedy and a squirt of something from a typical pantry is a sticker that comes with the kit, indicating the fix has been applied. But there is no evidence that the tube-of-lube solution addresses what caused Dussault's seats to fail. When Dorel finally got those car seats back, it couldn't reproduce the problem. Nor could outside testers, including Consumer Reports, which takes an aggressive stance toward product safety. Though Dorel won't disclose how much the recall cost, its annual reports indicate the juvenile division has spent more than $50 million in the past five years on product liability expenses.

Barry Mahal, Dorel's executive vice president for child-restraint systems, says the company agreed to the recall out of "an abundance of caution." What Mahal didn't add was that Dorel also likely agreed to the recall because it knew it couldn't win; it had already been down that road with the recall that led to the Pulitzer Prize—winning investigation. Though the Tribune reporters quoted a Dorel attorney pointing out that a particular car crash, not a product defect, may have caused a child's injuries, the story ended with a depiction of that child rocking back and forth in her kindergarten class, unable to speak or seemingly comprehend.

The interplay between tragedy and recall makes for gripping narrative; there's little doubt that such stories can compel normally glacial federal agencies to react quickly. The most prominent example of this is the fairly well-known—but still poorly understood—story of Toyota and the phenomenon called unintended acceleration.

Accounts of Prius, Lexus, and Camry models unexpectedly gathering speed began to dominate the headlines in 2009. That was when the audio of a horrific 911 call—made from the car of off-duty California Highway Patrol officer Mark Saylor—garnered hundreds of thousands of online plays. Saylor and three members of his family died when their Lexus spun out of control while traveling more than 100 mph. The incident was ultimately attributed to the wrong floor mat installed by a dealer, but experts, including Joan Claybrook, former head of the NHTSA, insisted the problem had to be an electronic defect. In February 2010, Department of Transportation secretary Ray LaHood recommended that Americans stop driving their Toyotas. Three weeks later, the company's CEO, Akio Toyoda, issued a public apology. Toyota ended up recalling more than 8.5 million vehicles. The damage to the company ran into the billions of dollars: Its U.S. market share dropped from 17 percent in 2009 to 12.6 percent in 2011, putting it in third place behind General Motors and Ford.

In an effort to find the alleged bug, the NHTSA enlisted NASA to conduct the largest automotive defect investigation in history. In February 2011, the agencies concluded that no flaws in Toyota's control systems could be found. The most likely culprit, the report revealed, was "the driver's unintended application of the accelerator, rather than, or in addition to, the brake." In other words, human error.

Brian Lyons, safety and quality communications manager at Toyota USA, says that the company didn't realize how quickly media reports would snowball. "The evidence said that there wasn't anything wrong with the electronics systems," he says. "The battle was getting the truth out there."

Though Toyota's market share is recovering, the company faces nearly 200 lawsuits. In the first of those cases, Toyota argued (in accordance with NASA findings) that driver error was the main cause of the accident—a defense that resulted in a storm of condemnation. Many of those who chastised Toyota for blaming the victim wrote for blogs operated, with varying degrees of transparency, by law firms looking to identify such victims and file claims on their behalf. These firms use technological savvy to game Web results, either by purchasing keywords and advertising or by optimizing their sites to register prominently with Google. Enter "car seat defect" into the search engine and of the first 50 results, 43 are law firms.

"What we've ended up with is a protection racket," says Michael Krauss, a professor of law at George Mason University who specializes in product liability. "These lawyers are on a never-ending search for clients. Paying them off has become part of the cost of doing business."

Krauss says that excessive litigation has skewed the entire way products are built, sold, and marketed today. Companies hesitate to deal openly with flaws because they fear litigation; consumers, noticing this, mistrust manufacturers. And the U.S. legal system has become an enabler, Krauss says, because liability cases here—unlike in most other countries—are decided by juries rather than judges, allowing room for emotion to swamp facts. Multimillion-dollar settlements are far more likely, Krauss says, "because a company is going to be very averse to being in a courtroom where aggrieved parents are holding up pictures of a dead or disfigured child while the jury hears the evidence and deliberates."

Companies like Dorel work in such a charged atmosphere that they can't even obtain product-liability insurance; instead, they keep tens of millions of dollars reserved for private lawsuit settlements. Dave Campbell, an analyst with the Juvenile Products Manufacturers Association, says that consumers pay a premium of up to 5 percent on most kid-oriented products in order to fund settlement pools. For smaller businesses, operating closer to the margin, high liability insurance can push them out of the market—or dissuade an entrepreneur with a smart innovation from launching a company in the first place.

Government has a role in ensuring product safety. But the bottom line is that a recall—whether or not the product is dangerous—is useless if nobody hears about it. Most current efforts to spread the word seem to do little but add to the clamor. Federal agencies now offer more mobile apps, more text alerts, and more website widgets, all drawing on the same vast stores of raw data. The most recent addition to this well-intentioned deluge is saferproducts.gov, mandated by the Consumer Product Safety Improvement Act (CPSIA) of 2008. The site duplicates much of the information on recalls.gov, though with a better-looking interface and a new feature: a public forum for reporting defects.

As valuable as such a tool might seem, one result has been to turn the site into an electronic complaint department. (Recent incidents included a man who was burned by fireworks and another who felt nauseated after shaving.) CPSC commissioner Anne Northup objected to the use of agency resources to investigate such claims in Congressional testimony in March 2011. "Many believe the public database, if left unchanged, will be useful only to trial lawyers or advocacy groups that will be able to populate it with unverifiable, secondhand information for their own purposes," she stated. (CPSC chairwoman Inez Tenenbaum disputes the reports are unverifiable.) More promising is that the CPSC recently opened its database and underlying code. This should enable private software developers to build custom, highly targeted applications, says agency spokesman Alex Filip.

Both the NHTSA and the CPSC have also been plagued by an investigative process that hasn't kept up with technology. In congressional hearings following the Toyota recalls, NHTSA administrator David Strickland said the agency employed just five electrical engineers and one software engineer out of its 125 engineers working on automotive investigations. The CPSIA mandated that the agency improve its technical facilities so it could conduct more authoritative product tests. In a 2011 report to Congress, the CPSC said that it was unable to adequately accomplish that goal because it "lacked the necessary infrastructure to directly accredit the testing laboratories." In both internal documents and public statements, the federal agencies admit to being as overwhelmed as consumers. "We've got very, very full plates, and we don't move at the speed we'd like to," an NHTSA official told me. "But we do what we can with what we have."

Most observers say the solution lies in both addressing the recall process and rethinking how recalls are communicated. That includes examining the terminology. In the United Kingdom, the word "recall" is used only for products that have caused, or are likely to cause, deaths or very serious injuries. Lesser cases are termed corrective actions; they can be addressed via consumer education or minor fixes, like the tubes of oil offered by Dorel. The adoption of more nuanced wording has been proposed repeatedly in the U.S., but government officials involved in recalls say that such a system is too risky because it might prompt consumers to further ignore product-defect notices. "The problem is that we end up making the judgment, when that should be left to consumers," the NHTSA official says. "If they think the risk is too minor, they won't take advantage of the remedy."

If scare tactics worked—if they actually made people comply with recalls—then perhaps a little hyperbolic horror would be a good thing. Unfortunately, says William K. Hallman, a psychologist and director of the Rutgers University Food Policy Institute, fear turns out to be a lousy motivator. "Scaring the crap out of people doesn't work," says Hallman, who is working with the FDA to help improve recall communications. "When people are that terrified, the outcome is usually no action at all."

As I read through the accounts of various recalls, I found myself wondering: Regardless of whether the issue is a foot on the wrong pedal or a Roman candle aimed in the wrong direction, how much responsibility do manufacturers bear for operator error? In the case of Mark Saylor, someone was clearly at fault: the dealer who installed the floor mats in his loaner vehicle. But what about when consumers purchased floor mats that weren't made by Toyota? The company's corrective action, which included a pedal redesign to make mat entrapment less likely, was the right thing to do; a shopper should be able to reasonably expect that a product will be safe under standard usage. That said, many cases I looked at seemed to revolve around good people making honest mistakes—bad mistakes, sometimes—and being unable to recognize or admit to the fact.

One company that attempted to both do the right thing—fix a problem that wasn't necessarily a defect—and point out that customers had a responsibility to use its product correctly is Maclaren, one of the world's largest makers of baby strollers. In November 2009, the U.S. subsidiary of the British firm recalled 1 million umbrella-type folding baby carriers. Twelve children had suffered injuries, including partial amputations, when their fingertips got caught in the product's folding mechanism. Those injuries never occurred when children were actually in the strollers—only when they became entangled as parents were folding the equipment. The problem also wasn't unique to Maclaren strollers.

Despite clear instructions on proper usage, Maclaren agreed to a recall. It was preparing to announce the effort when the process suddenly went awry, according to company CEO Farzad Rastegar, writing in the January 2011 Harvard Business Review. Maclaren had planned to issue a press release announcing the recall on Nov. 10, 2009, but one day earlier, word of the recall leaked to the New York Daily News, setting off a media frenzy. "For several months," Rastegar wrote, "we had worked with the CPSC on a plan to make owners more aware of the danger and to provide protective hinge covers... But the news story provided none of this context. Nor did it explain that practically all strollers on the market have similar hinges."

Within 24 hours TV crews showed up at the company's Connecticut headquarters and hundreds of mommy blogs issued frantic warnings. Panic ensued. A Brooklyn mother told The New York Times that the two biggest threats her children faced were "swine flu and Maclaren strollers." The company's website and email crashed—and even that was portrayed as an example of Maclaren's ineptness.

"We had hoped that the recall would build awareness about the wider risks of operating a stroller, not just about hinges," Rastegar wrote. "Instead, we would have to start defending our brand."

In the end Maclaren distributed more than 300,000 hinge covers. Even so, by the end of 2011, 149 additional injury reports had been filed, indicating typically low recall compliance. Even as it reissued notices, Maclaren insisted the process was flawed. Its website pointed visitors toward a document calling for all stroller manufacturers to offer hinge covers. (Several brands with similar designs have been recalled, but I was able to find five folding strollers with unprotected hinges in less than an hour of shopping.)

CPSC's Filip says the piecemeal approach to umbrella-stroller recalls could change; the CPSC could issue a category-wide product-defect order, as it did for drop-side cribs in 2011. But in order to do that, the agency would need "to make a strong case for intrinsic danger," he says. So far, such a case hasn't emerged.

Just after Christmas 2011, Maclaren USA filed for bankruptcy. Among the listed liabilities were "unknown" claims (meaning the amounts were undetermined) by seven families and law firms suing the company, along with the CPSC, for ongoing expenses related to the stroller recall.

Products need to be user-friendly in the real world, where people are in a rush, where crevices get jammed with gunk and goo, where—in the case of car seats—Mom's vehicle might have deep buckets and Dad's truck might have a bench. We're a one-car family, but even so, I was surprised at how difficult it was to install our car seat (it didn't help that I was doing it for the first time while my wife was in labor—don't ask). Statistics show that as many as 80 percent of existing car seats—depending on the seat type, the age of the child, and the vehicle—are improperly installed or misused. It's been that way for decades despite efforts to teach parents the correct process. There are more than 25,000 certified car-seat-installation technicians in the U.S., most of whom offer their services for free through local hospitals and police and fire departments.

Maybe the expectation of ease is unreasonable. A car seat is a complex product that has a very critical function. As much as I wish that my car seat would simply click in and always be secure, I know better: I've got to learn how it works, use it correctly, and maintain it.

By the end of this year, Dorel will submit a report to the NHTSA outlining the results of the Vantage recall. Most likely, compliance will be low, and most likely, it will be difficult to determine if any injuries have occurred because of the recalled product or if any have been prevented by the recall. What's certain is that there will be more recalls. Parents and manufacturers will continue to make mistakes—something I became personally aware of when, using Urban Apps' Recalls for iPhone, I made a second pass through my house.

One of our son's earliest favorites was a bright yellow baby chair called a Bumbo. I discovered it had been recalled in 2007 after reports that children pitched themselves out of it. Though there were some design flaws, the bigger issue seemed to be misuse: Serious injuries occurred when the product was placed on a table or counter, resulting in a fall from height.

I was well aware of this problem; I'd almost let it happen myself. We were having dinner, and I'd placed the Bumbo in the center of our dining room table. Our son wriggled forward, and suddenly he was out of the chair. He didn't fall off the table, but it was close—and I felt terrible about the bump on his head.

With the Bumbo I broke two major rules when it comes to product safety. First, I didn't follow the instructions. That was because—here's the second broken rule—I didn't have instructions. We bought the product used, as more and more parents are doing these days; reselling a recalled product is illegal, though such goods can be readily found at garage sales and on auctions sites like eBay. But mostly I didn't use common sense.

Personal responsibility isn't the sole answer. It needs to be backed up with an approach to defects that's evidence-based, and with communications that get genuinely bad products off the market. One such model can be found in membership clubs. Costco's Wilson says that traditional recall notices have been so ineffective that his company created its own system, maintaining a database of everything a customer buys. When a recall is announced, the company calls everyone who has ever purchased the product, then follows up with a letter. It also has an internal recall process that moves far more quickly than a government agency's, issuing notices for products it believes are unsafe based on negative customer feedback. Wilson says that Costco's effectiveness rate for reaching consumers is more than 90 percent.

Could that Costco model be applied to consumers who don't buy at membership-based stores? Widespread use of debit cards, transactions that automatically register serial numbers for inventory control, and the fact that most major retailers now have loyalty programs that track purchases indicate the data may be available. (CPSC's Filip says such a program is not within the agency's current mandate.)

Several companies are developing search-engine-like technology that trolls federal recall sites. WeMakeItSafer, for example, reformats and consolidates the information into useful data, then notifies registered users directly when there is a relevant recall, or works with retailers and manufacturers, who can direct the recall toward a specific database of customers. That way, consumers will receive information only about products they own. "The answer isn't telling the world there's a recall in the hope of reaching the few people that own a product," CEO Jennifer Toney says. "It is about using technology to home in on just the people who need to know they have issues with something they own and need to get it fixed."

I want—for myself and my family—to be protected from truly defective products. But I also want to see a system that is honest about identifying those products, that assesses risk properly, that advocates responsibility as much as—or more than—it assigns blame. I want to know that decisions are being made carefully enough that I'll be likely to hear about them—and likely to care about them. I'd be delighted if the government did that; I'd also be delighted if that was something that could happen via new technology or through my favorite retailer or anyone else who could figure it out. I'd even be happy to get the message from Elmo. I just want the message to mean something.