When Synergy Becomes Suppression: Mechanisms After Acquisition

The subject here is the decline in an acquired start-up’s measured innovation and the competing processes that could produce it. The practical task is diagnosis: distinguish a change in patenting strategy from a real reduction in inventive or productive work, then assess why the decline is larger in some acquisition settings than others. The available evidence points toward market-threat explanations in important cases, but it does not justify treating every weak post-deal outcome as proof that a buyer intended to kill an idea.

One symptom, several possible processes

A target can file fewer patents after acquisition for more than one reason. The visible symptom is the same—a lower patent flow—but the economic meaning differs. If the target continues research under secrecy, the patent measure understates innovation. If the parent turns the target’s technology into a product, the start-up may be doing less exploratory work because it has moved into implementation. If the merger disrupts research teams, innovation may fall even though neither side wanted that result. If the buyer owns a project that threatens an existing business, shelving it can be a rational way to protect current profits. A sound diagnosis has to keep these possibilities separate.

The first possibility is a change in intellectual-property strategy. Patent applications make an invention visible to competitors and create legal protection, but they also disclose information and impose costs. After acquisition, a parent may prefer secrecy, internal know-how, or another form of protection. Under that explanation, target patenting would fall without a corresponding fall in the underlying research. The parent’s own patenting might not rise either, because the work could remain secret. A patent decline alone can therefore never establish that invention stopped.

The OECD study tests this possibility indirectly by comparing targets according to the patenting intensity of their acquirors. If low-patenting parents were systematically more secretive and imposed that strategy on targets, the target decline should be strongest for acquisitions by those parents. Instead, the reduction is statistically significant for targets bought by high-patenting acquirors and not significant for targets bought by low-patenting acquirors. That pattern does not prove that secrecy never matters, but it weakens the simple explanation that the aggregate decline is mainly an IP-policy change imposed by low-patenting buyers.

The second possibility is a move from technological development into commercialisation or integration. A start-up may stop generating patents because its important work has already been done and its new owner is trying to manufacture, distribute, or embed the technology. This explanation makes additional observable predictions. Market-entry activity, such as trademark registration, might increase. Productivity might improve as the technology is connected to the parent’s assets. The parent’s innovation measures might rise if integration makes the acquired knowledge productive elsewhere.

The supporting indicators do not fit a clean version of that story. The OECD finds a statistically significant decline in the target’s trademark flow, approximately half of its standardized average. In the restricted sample with financial-statement information, total factor productivity falls by about 4% for targets and about 8% for acquirors. Those movements are difficult to reconcile with a simple claim that the target merely stopped patenting because it was successfully commercialising.