Frank, I have to disagree with you TOTALLY here. The solution to problems like this is NOT to study them to death. One must first get to the cause and ELIMINATE the cause by some means. The number of these parts that are making it to circuit boards through manufacturer authorized distribution channels in "Western" countries is at present almost nil. As much as I disagree with Deming on many things, I agree with him that testing in quality is a tail-chase.
I would suggest that IC manufacturers provide a cradle to grave tracking system for their components. Any IC manufacturer that doesn't would not be "trusted" by board houses. There would be cost involved but given the potential liability and rework costs it might be well worth it.
Although I'm sure the problem is very real, your assessment of "potential" is a good illustration of extrapolation in action. By your logic, I could say that potentially $165B in automobiles were counterfeit as well, since that is the size of the U.S. market.
Good reporting should require that you put a realistic upper bound on monetary claims, and not just shout that the entire sky is falling.
Karen- I'll defer to Bill Schweber's answer down below.
I would also point out: these are pretty broad product categories. Between the top five categories, you've got the bulk of chip types. I suspect it would be useful to see a more detailed breakdown, but perhaps IHS iSuppli reserves that kind of information for those who pay for their report.
My company - Tiger's Lair - has a solution to prevent counterfeiting of new digital ICs based on modifying their design. A chip produced from a protected design will not work unless it is unlocked in a secure environment. This makes both reverse engineering and overproduction useless since the resulting chips will not be operational.
We are looking for partners interested in exploring this solution.
Oh, and there should be a place to expose every user of fakes. EEtimes can play a role. Let a dedicated editor receive my fakes report along with a copy of their invoice. It'll take some investment from EEtimes, but it will be made up by adds from companies who are able to deliver clean components directly to the assembly floors.
I recently received a product with a faulty LDO. The solution was to buy parts from European distributors and ship them to China myself.
In my opinion the distributors can play a large role. The stolen night-shift fallouts typically don't make it to approved distributors. So part of the process of selecting a subcontractor should be that they use verified distributors and show me the shipping documents for the parts on my BOM.
There is business to be made for DigiKey and others in sending the parts on my BOM to China and mark the rolls only with my item numbers. Shipping CDP to China is something no Western company is happy to do. But if they could sell such a service as an insurance policy, I'd happily buy.
There is some movement toward addressing the twin problem RWatkins identifies. Most optoelectronics manufacturing has moved offshore since the comms industry bust of the early aughts. A member of the U.S.-China Economic and Security Review Commission recently announced plans to create a "trusted" manufacturing facility for optoelectronic components near Dayton, Ohio, and Wright-Patterson Air Force Base. The military is concerned about fake parts as well as erosion of the U.S. industrial base. More "trusted" manufacturing plants is one way to tackle the counterfeit problem. Undoubtedly, other approaches will emerge as the problem gets worse.