Confusing Fast Followers with "IP Theft"
it's illusory to think that automotive technology is protected by patents
I attended the 31st annual Federal Reserve Bank of Chicago automotive symposium in Detroit. This is the first of many posts resulting from that. You might check out Joe White’s “High Speed Rodeo” substack, I sat next to him yesterday, and he posted two updates while I was sitting there. I’m slow and deliberate, academic clockspeed. He’s fast and deliberate, journalist clockspeed.
Let’s get started.
I frequently encounter the claim that the Chinese are stealing our intellectual property. Of course they are – “of course” because our firms do the same. It’s just that with a mature markets in the US, Japan and Europe, dominated by the replacement or refreshing of existing models at a 4-year cadence, it’s not as visible to outsiders. China’s market is still growing, and there are a LOT of new entrants who are still fleshing out their product lineup. New models and upgrades get registered with the government (we have FMVSS where carmakers self-certify and report).
I track that. In CY2024 there were 591 new models, refreshes and minor changes. Of those there were 195 that were new models (including 23 new platforms), of which 75 were new EVs and 75 new PHEVs (and so only 45 were new ICEs). We get a couple dozen in the US, Europe a bit more, and only a handful of new platforms combining the US and EU.1 Since a car company requires a new model or even a new platform, the pace of change is accelerated.
Next week I’ll be in the meeting with 20 other judges to decide the winners of the Automotive News PACE and PacePilot awards. One of our standard lines of inquiry in sorting awardees from great engineering is “how far ahead are they of the competition.” Companies constantly scan each others patent filings, they get feedback from customers and potential customers, and once something is in production they do a teardown (or contract with one of the two big teardown specialists, Caresoft and A2Mac1, for data). Not infrequently, a competitor is already in the market, but in the interim the supplier has picked up conquest business that is then theirs to lose. More often the answer is 6-12 months; at the extreme it’s 3 years for a system that has a innovations across a lot of individual components, and that is only just going into production so rumored (from contracts not won) but unknown to the competition.
Now there is likely the occasional case of a purchaser allowing a customer to walk past a sample room that has their rival’s prototypes on display. The RFQ (request for quote) may have more specificity than needed, given a hint to competitors that they may need to rethink their design. After all, no one likes to be dependent on a single source, and an design or procurement engineer may try to nudge things along. Once a new model is well along in the development process, however, competitors simply won’t have enough time to come up with alternatives until the next model. Car companies are too dependent on suppliers for ideas to piss off one of the big players, and word does spread.
That gets back to my original point, that time to market is the most important competitive variable, and stealing someone else’s ideas puts you too late for the current project. That’s true more generally, as found in a famous 1987 Brookings study, Appropriating the Returns from Industrial Research and Development.2 They surveyed 650 individuals in 150 different lines of business, focusing on executive with an oversight of engineering or R&D (what we would now call the CTO, chief technology officer). What they found was that, while in certain industries patents were central in the chemical industries (such as pharmaceuticals), the key elsewhere was speed to market and sales support. “Lead time” appears in their paper 24 times, often paired with learning curve, that is, using lead time to lower production costs.
Are Chinese “stealing” technology? Since US and European suppliers do that, with the encouragement of Ford and BMW and the other major players, they would be incompetent if they weren’t. The real difference is that their customers are launching so many new products that Chinese suppliers (including global Tier I suppliers operating in China) don’t have to wait 2 years to sell Gen I of an innovation, and they wait another 2 years before they can find a customer for the more mature Gen II version.
Let me close with a concrete example. Tesla was – past tense – a leader in using large aluminum castings for chassis structures. As part of the Fed conference, we saw teardowns of the Model Y across 3 iterations, across 3 years. By US or European standards, such “running changes” are effectively unknown. But we also saw a teardown of a Chinese competitor’s newest model, which made qualitative improvements to what Tesla had accomplished – lighter, less expensive, and stiffer. Using the battery pack in place of a floor stamping is what all EV producers now do in China, because the concept was available, and because unlike Ford (whose EV chassis were also on display) they are on their 2nd or 3rd EV chassis.3
China’s difference is the combination of a faster clockspeed4 and a bigger market.5 The successful EV firms, above all BYD [and the slow-moving Tesla], also have scale. Ditto their suppliers, who pass cost savings on to their customers.
It will be very hard for the US to catch up, absent a faster transition to EVs. Europe is further along, but for now the US, Canada and Mexico are trapped in a vicious circle: absent lower costs, consumer uptake of EVs is growing only slowly. But without faster growth, it’s hard to lower costs. The IRA passed by Congress under the Biden administration aimed to change that. We will see what the new Congress will do.
That is, at most 5, since I have the normal number of digits, surviving my time in factories without losing a finger The rule of thumb when asking someone how long they’d been at Mack Stamping was to ask them to hold up their hands, with one lost finger for each decade. Losing a thumb was uncommon, or at least few who lost a thumb managed to keep on working.
Authored by Richard C. Levin, Alvin K. Klevorick, Richard R. Nelson, and Sidney G. Winter, the first 4 grad school professors of mine at Yale. Nelson moved to Columbia, while Levin went on to become the president of Yale. The lively back-and-forth among Nelson, Levin, Paul McAvoy and a few less regular participants (such as Merton Peck) in the graduate industrial organization seminar at Yale was important in teaching me how to think about problems from multiple angles.
Another barrier is that the sheet metal and interior of the Model Y are largely unchanged. They are limited in how much they can change the chassis because the dimensions and weight distribution of the “top hat” that we see is the same old (and to my eyes, now boring) design.
A nod to Charles Fine of the MIT International Motor Vehicle Program, who pushed that concept.
Current forecasts for CY2025, part of the Chinese New Year news cycle, put the domestic market at 31 million units and exports over 5 million units, putting total production over 36 million units. Google AI states that production in the USMCA in 2024 was 21 million units, including exports. These data include both passenger and commercial vehicles.