CPSIA - Numbers Don't Lie (Update No. 1)

I have received a lot of feedback on my data on injuries from lead since I published it on Thursday. Some of the comments deserve further exploration.

a. Did anyone have this data previously? I think the answer is NO. I have heard from inside the CPSC that this kind of data analysis is not being done. The data is nowhere to be found, except here. Recent testimony by consumer advocates and Congressional zealots is strangely bereft of details, just long on invective. Don't we deserve better?

The hysteria over lead-in-paint recalls, combined with other recalls that were unrelated created such a lynch mob atmosphere in Congress that the only data that registered was data that supported the mob's POV. The actual data is therefore something of a surprise.

Let's look at the four reported lead injuries for a moment. The one death from lead, the famous incident in Minnesota where Jarnell Brown swallowed a lead charm off a Reebok bracelet, is well-known.

Injury no. 1: L.M. Becker recall (vending machine jewelry, Sept. 10, 2003): "The firm received one report of a child who swallowed the necklace's pendant, which reportedly resulted in high blood lead levels."

Injury no. 2: Four company recall (vending machine jewelry, July 8, 2004): "CPSC has received one report of lead poisoning when a child swallowed a piece of toy jewelry containing lead that was previously recalled. No reports of injury or illness have been received for the recalled products announced today." THIS REFERS TO THE L.M. BECKER "INJURY" ABOVE.

Injury no. 3: Munrie Furniture, Inc. (cribs and matching furniture, December 23, 2008): "Munrie has received one report of a child ingesting the paint. The child was diagnosed with lead poisoning."

Injury no. 4: Allreds Design (bracelets and clips, February 17, 2010): "Allreds Design received one report of a 10-month-old child who was treated by a physician for elevated lead levels."

Do these three injuries (one injury is double-counted) have anything in common? I assert that causation is not proven in these cases. Yes, lead poisoning is alleged, however there is no evidence that the cause has been determined definitively. READ what the CPSC said - the connection to the defective children's product is loose or even conjectural ("reportedly resulted in high blood lead levels"). Remember car seats that gave kids lead poisoning? Toxic car seats were a hoax. The injury data is flimsy at best. And this is all the evidence there is of injury from lead in children's products in the last 11 years of recalls.

The Democrats and consumer groups would rather eat broken glass that admit that the lead poisoning scare is a hoax. They want to run our businesses - so there's no conceding that all this economic damage cannot be traced to anything other than ONE TRAGIC ACCIDENT IN AN ELEVEN YEAR PERIOD. Uno, that's it, in our country of 300 million where thousands of kids die every year for various reasons.

Well, at least we know our companies will die in a valiant cause, to reduce the ten-year death rate from lead in children's products from one to zero . . . .

b. Was the CPSC really broken? This is Mr. Waxman's assertion, as expressed in his opening remarks at the April 29 hearing. So it must be true, right?

I hope to provide more data on this topic soon. In the meantime, I will simply pass along the comments of a friend who is in the CPSC community, namely that Congress underfunded the CPSC for 20 years, leading to severely constrained budgets and hiring. Consider these quotes from a 2007 Businessweek article about the CPSC:

"Yet while the CPSC has never been more vital, through much of its 33-year history the agency has been chronically understaffed and underfunded. Overseeing 400 recalls a year, most at companies' requests, the CPSC's compliance team has less time to initiate its own investigations, which tend to reveal the most serious risks. . . . Growing workload and shrinking resources have left many disheartened. From a peak of nearly 1,000 in 1980, CPSC's head count has fallen to 400. . . . What can be done to help the agency? In a word, money. It's been 17 years since Congress thoroughly reviewed the CPSC's resources and needs, says Nord."

So, let's see, Congress has been tightfisted with budgets for this little agency for many years, starving it of needed resources and headcount, effectively shrinking it over a 17-year period to a withered state, and then after an outbreak of large-scale toy recalls (by and large injury-free), Congress blames the agency for inattention to its mission and severely rewrites the law to punish the marketplace and the agency itself.

After all, why blame the entity responsible for the problem in the first place, Congress? Much easier to blame the agency!

c. Wow, those were a LOT of recalls? Is that the tip of the iceberg? As far as I can tell, the answer is that most of the iceberg is a mirage. Experienced CPSC hands note that the recall notices are prepared by the press office at the CPSC and are meant to attract attention and headlines. Big numbers, if defensible, are best suited to demonstrate that the cop is on the beat. A few tricks of the trade is to add in as many sources of "recalled" items as possible. I believe that as many as 60% (that's no typo) of all recalled units NEVER WERE SOLD. I would simply observe that if they were never sold, they never had the potential to cause injury.

Second, the population of recalled items is always inflated out of an abundance of caution whenever there is ANY doubt as to the identity of dangerous products. In other words, if a company sold one bad lot but also sold nine good lots, all indistinguishable, the recall would be announced for all ten lots, even though there is NO dispute that nine of the lots are absolutely fine.

Without this insight, recall statistics might be alarming, at least in a sense. Actually, the recalls are something of a mirage, an illusion of legions of bad products that really don't exist or were never sold. How can we verify this? Among other things, injury statistics back up this assertion. If we had 300 million units of dangerous products in circulation, the injury statistics of 2381 injuries in 11 years seems pretty low to me. Assuming an average time in the marketplace of three years per recalled item, this implies an annual injury rate of 0.026% (from all causes, not simply lead). If the products are in the market for only one year on average, the annual injury rate is still only 0.077%. In other words, in a worst case scenario, you can safely use RECALLED children's products 99.92% of the time. And you would presumably be even safer with NON-RECALLED products.

Oooo, scary.

Think of it this way: There are about 3 billion toys sold in the U.S. annually, according to Alan Hassenfeld, former CEO of Hasbro. Over 11 years, that 33 billion toys. [Considering that "Children's Products" includes far more than just toys, the pool of 11 years of sales is probably north of 500 billion units.] Were you to assume that all 899 recalls in my data were toys, the pool of 308 million units recalled would represent 0.93% of all toys sold in that period. So, if 0.93% is safe 99.92% of the time, and the rest presumably safe at a higher rate (let's say 99.999%), then the blended safety of all toys is 99.99%. The result is probably higher than that.

Numbers, numbers, do they matter?

We are spending not less than $5.6 billion per annum to "fix" this 99.99%+ safe problem. In an effort to create a much "safer" environment for kids, the helpful folks at the CPSC have produced literally thousands of pages of documents, rules and instructions to govern our businesses down to the tiniest detail. Unfortunately . . . the assertion that anyone will be safer CANNOT be proven as a matter of mathematics.

A neurotic bill administered by people who no longer can assess what is and is not safe is a danger to our society. The data proves it. Who should be held accountable? Congress? The Dems? Inez Tenenbaum? Some or all of the above.