Antitrust authorities increasingly assess mergers through the lens of innovation, particularly in research-intensive sectors such as pharmaceuticals. In new research, Carmine Ornaghi and Lorenzo Cassi show how mergers disrupt human capital and reduce innovation in what they call manslaughter acquisitions.
Innovation has become central to merger enforcement, particularly in pharmaceuticals, where research and development drives new therapies and patient outcomes. Recent debates have focused on the risk of “killer acquisitions”: mergers and acquisitions that eliminate overlapping research pipelines and suppress future competition. But harm to innovation need not arise only from the intentional termination of projects. It can also stem from disruption to, and the loss of, the human capital that drives innovation. Highly productive inventors—who are more likely to receive external job offers—may decide to leave if integration generates dissatisfaction or instability. Such departures can disrupt established collaboration networks and negatively affect the productivity of scientists who remain within the organization. Other organizational changes from consolidating laboratories and research teams to reprioritizing projects can further disrupt research.
Historical examples illustrate these dynamics. After Glaxo acquired Wellcome in 1996, Wellcome’s main British research facility in Beckenham—home to roughly 1,500 scientists and staff — was closed. According to industry observers, GlaxoWellcome lost more research talent than anticipated. Similarly, following Pfizer’s acquisitions of Warner-Lambert and Pharmacia in the early 2000s, major R&D operations in Michigan and Illinois were shut down. Even when mergers aim to improve efficiency, the resulting organizational changes can substantially erode the human capital base that underpins innovation.
To examine this mechanism systematically, our recent study analyzes more than 500 pharmaceutical acquisitions worldwide from 1988 to 2015. Using detailed patent data from the European Patent Office, we track the activity of thousands of inventors before and after an M&A. This allowed us to observe whether scientists remain with the merged firm to continue R&D, move to other organizations to pursue R&D, or exit innovative activity altogether, though we cannot observe in an exit whether inventors retire, move into non-R&D roles, or leave the sector, as individual HR records are not available. When scientists do not exit innovative activity, we are also able to track how their productivity evolves following an acquisition. We find that acquisitions are associated with a marked increase in inventor exits, substantial productivity declines among those who move, and significant reductions in patent output even among scientists who remain with the consolidated firm. Because these patterns suggest an erosion of innovative human capital driven by integration rather than the deliberate suppression associated with killer acquisitions, we refer to them as “manslaughter acquisitions”— to highlight both the human nature of the loss and its non-premeditated character.
Evidence on human capital disruption after mergers
Our study provides new evidence on how acquisitions affect inventors’ careers and productivity in the pharmaceutical industry. We document three consistent patterns.
First, acquisitions are associated with a marked increase in exits from innovative activity. Depending on the specification, the probability that an inventor stops patenting rises between 7-17 percentage points following an acquisition. This effect is not confined to less productive scientists. In the absence of an acquisition, inventors in the top productivity deciles for number of patents are 18.2 percent less likely to exit innovative activity (stop patenting) five years after their first patent compared to their less productive peers. However, during an acquisition, these highly productive scientists experience an increase in exit probability of around 22 percent. As a result, their usual advantage in remaining active innovators largely disappears, and they become just as likely—or even more likely—to exit the innovation market.
Second, increased mobility does not offset these losses. Inventors employed by acquired firms are between 12-19 percentage points more likely to move to another firm following an acquisition than comparable inventors working for non-targeted firms. However, those who relocate experience a sharp decline in productivity. Our estimates indicate that moves produce roughly two fewer patents over the subsequent three years relative to what they would have achieved had they stayed at their firm in the absence of the acquisition—a counterfactual constructed using comparable inventors who did not move. This corresponds to a reduction of around 30 percent in patent output over the relevant post-acquisition period.
Third, innovation declines even among inventors who remain with the merged firm. Treated innovators who stay produce about 0.92 fewer patents than comparable inventors who stay at the same, non-targeted firm after three years, with the gap increasing to 0.97 patents after five years: an output reduction of roughly 11 percent. These findings indicate that post-merger integration can depress innovative performance even without physical separation, suggesting that organizational disruption affects not only those who leave but also those who stay.
Why might integration reduce innovation?
Our empirical results clearly document declines in inventor retention and productivity following acquisitions. While the precise organizational dynamics are difficult to observe directly, the evidence points to two mechanisms that are particularly consistent with the patterns we document. These mechanisms are suggestive rather than definitive, and further research is needed to disentangle them fully.
A first mechanism relates to how research teams are integrated after an acquisition. Consolidations are often expected to foster collaboration between the target’s scientists and those of the acquiring firm. Our data allow us to examine whether such collaboration mitigates productivity losses. Among inventors who remain with the acquired firm, 57 percent of post-acquisition patents involve joint work with colleagues from the acquiring firm, indicating that integration and cross-team collaboration do occur. However, collaboration is selective and uneven in its effects. Inventors who engage in cross-firm collaboration were already more productive before the acquisition, suggesting that integration builds on pre-existing strengths rather than creating them from scratch. More importantly, post-acquisition outcomes diverge sharply: non-collaborating inventors experience substantial productivity declines, whereas collaborating inventors see only temporary and smaller reductions. This indicates that synergies are not automatic and may be more difficult to realise than often anticipated.
A second mechanism concerns the disruption of inventors’ careers following acquisitions. Exit is not confined to marginal researchers. Even highly productive inventors are significantly more likely to leave innovative activity. This points to substantial instability within the acquired organization and suggests that the decline in patenting reflects more than internal project rationalization. Inventors who move to new employers also generate significantly fewer patents than comparable peers not affected by acquisitions, indicating that displacement carries persistent productivity costs.
Implications for merger policy
Our results do not imply that mergers are inherently harmful. Acquisitions can generate efficiencies and, in some cases, enhance innovative performance. For example, mergers may improve access to financial resources, combine complementary technologies, or reallocate capital across projects more effectively. However, innovation often depends on scientists’ tacit knowledge, and complementary research efforts. When post-merger integration disrupts these organizational foundations—through laboratory consolidation, restructuring, or unexpected inventor departures—the anticipated efficiencies may fail to materialize. Once lost, innovative capacity is difficult to rebuild.
Our findings suggest that merger analysis should pay closer attention to the human-capital dimension. Beyond assessing product overlaps, authorities may need to consider when integration risks weakening the scientific workforce that sustains innovation. Identifying such risks is challenging, just as predicting post-merger efficiency gains is inherently uncertain. Closer attention should be devoted to acquisitions in which innovative output is concentrated among a small number of highly productive inventors and to situations where research teams are tightly interconnected and therefore more vulnerable to disruption. When evaluating projected efficiencies, agencies might also examine the credibility of post-merger retention strategies and R&D staffing commitments, as well as plans involving extensive laboratory closures or the geographic relocation of R&D units.
Authors’ Disclosures: The authors report no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
