Can Digital Spillovers Help Industrial Latecomers Become Innovators?

When industry peers adopt digital technologies, they can change the knowledge, competitive pressures and investment signals surrounding other businesses. The resulting opportunity extends beyond copying a rival's software: it may help firms pursue technologies they previously lacked the confidence or resources to develop.

Research titled "How Does the Peer Effect of Digital Transformation Promote Breakthrough Innovation?", by Nana Yang, Qiming Liu, Ziqing Xing and Zujun Zhu, examines this relationship among Chinese manufacturers. Published in Systems, the study links greater digital transformation among industry peers with stronger invention-patent activity, suggesting that the returns from industrial digitalisation can extend beyond the companies making the initial investment.

The findings raise a consequential question for industrial policy: how much innovation potential is lost when digital programmes support individual firms without helping others learn from them? The evidence points towards benefits from wider diffusion, particularly in less developed regions. It also demands restraint: patent growth does not automatically establish technological breakthroughs, and the study's financing explanation accounts for only a small portion of the reported relationship.

Digital Investment Can Travel Beyond the Company That Pays for It

Digital transformation consumes money, management attention and technical expertise. For manufacturers, adopting new systems can compete with funding for experimental research, especially when both require substantial spending before producing returns. Observing how other companies implement technologies may reduce some of that uncertainty, making the wider industry a potentially valuable source of practical knowledge.

The study examines 1,527 Chinese A-share listed manufacturing companies between 2008 and 2023, covering 17,436 firm-year observations. The authors combine financial and patent records with annual-report analysis, measuring digital transformation through references to digital technologies. They define peers as other companies operating in the same manufacturing sub-industry and calculate their average digital transformation score.

They find a positive association between peers' digital transformation and a company's invention-patent activity. The relationship remains statistically significant after accounting for stable differences between companies, common annual influences and selected financial and governance characteristics. Further checks using alternative measures and time periods broadly support the direction of the result.

The authors interpret this pattern as evidence that peer influence can encourage innovation beyond simple imitation. Learning how another manufacturer deploys digital tools could reduce avoidable implementation errors, allowing a company to direct more effort towards its own technological objectives. Competitive pressure may also push managers to explore new applications rather than merely match existing practices.

Finance Opens One Door, but Leaves Most of the Story Unexplained

Ambitious innovation requires firms to sustain expenditure through uncertain results. External financiers may struggle to distinguish a credible research project from an expensive gamble, particularly when unfamiliar technologies are involved. The authors propose that digital progress across an industry can improve confidence and reduce information problems, easing financing constraints for individual manufacturers.

The study reports evidence consistent with this channel. Greater peer digital transformation is associated with lower financing constraints, while tighter financing constraints are associated with weaker innovation outcomes. A separate resampling analysis supports a statistically significant indirect relationship running through financing, giving the proposed mechanism some empirical backing.

Its size, however, changes the interpretation. The authors report an indirect effect of 0.0028 through financing and a direct effect of 0.0548. Calculated from these estimates, financing represents approximately 4.9% of their combined effect in that specification. The result supports financing as one contributing mechanism; it does not establish financing relief as the principal explanation.

Most of the estimated relationship thus remains outside the mechanism examined. Knowledge exchange, organisational learning and reduced uncertainty are plausible explanations discussed by the authors, but their individual contributions are not established. Nor does the research directly observe lenders changing interest rates, extending credit or approving projects because other manufacturers have become more digital.

For banks and development institutions, the practical implication is to investigate how learning support and finance might work together. Shared technical programmes could help businesses develop more credible investment proposals, while suitable financing could enable implementation. Such arrangements require evaluation, with lending decisions grounded in operational capabilities and project quality rather than digital terminology in corporate disclosures.

The Gains Look Stronger Where Catch-Up Has More Room

The study's regional findings suggest that digital spillovers may have particular relevance outside established industrial centres. The reported relationship between peer transformation and innovation is stronger among companies in China's central and western regions than among eastern firms. The authors' statistical comparison supports a meaningful difference between the regional estimates.

Their interpretation is that firms with less developed digital foundations can gain more from adopting practices already tested elsewhere. Access to proven applications may reduce the costs of trial and error and help businesses overcome knowledge gaps. Regional development strategies could therefore benefit from connecting manufacturers with usable expertise alongside investment in infrastructure.

Ownership differences offer a less definitive signal. The association is stronger among non-state-owned enterprises than among state-owned firms, but the statistical comparison falls just outside the conventional 5% significance threshold. The authors suggest differences in flexibility and incentives may explain the pattern, although the evidence does not conclusively establish those organisational explanations.

Competitive conditions also appear relevant. The study finds that greater market competition strengthens the peer–innovation relationship, although this result has comparatively weak statistical support. Higher organisational slack, measured through the ratio of current assets to current liabilities, weakens it. The authors interpret these findings as evidence that pressure to respond can influence whether firms act on peers' digital progress.

Neither finding justifies stripping companies of financial reserves or intensifying competition without limit. Liquidity can protect long-term research when revenues weaken, and excessive rivalry can undermine investment capacity. The narrower implication is that exposure alone may be insufficient: companies also need incentives and internal processes that turn external learning into purposeful experimentation.

Industrial Policy Needs Evidence of Invention, Not Just Digital Activity

The strongest qualification concerns what the study calls "breakthrough innovation". Its principal measure counts invention patents granted, adjusted logarithmically. Patents provide observable evidence of inventive activity, but they do not establish that an advance transforms a market, enters a fundamentally new technological field or produces substantial commercial value.

The digital measure has similar boundaries. References in annual reports can reflect genuine implementation, strategic aspirations or changing disclosure language. Companies classified within the same industry may never exchange knowledge directly, while important learning relationships may cross industry boundaries. These limitations make the findings more persuasive as evidence of an association than as a complete account of how innovation spreads.

The authors use several methods to address alternative explanations, but causality remains uncertain. Their instrumental-variable approach relies on historical telephone infrastructure combined with a company's earlier digital transformation measure. Whether these factors affect innovation exclusively through peer digital exposure remains debatable. Reporting inconsistencies in some estimates and sample counts also warrant clarification and discourage overconfident numerical claims.

For developing countries, the research nevertheless identifies a useful policy question: can digital investment build capabilities across a manufacturing base rather than remain concentrated among leading companies? Shared demonstration facilities, industry learning networks and technical advisory services are plausible responses. Their success should be assessed through additional research, commercially useful technologies and productivity improvements, with explicit attention to smaller firms' participation.

The connection to sustainable development is strongest through industrial upgrading and innovation under SDG 9. Employment gains, environmental improvements and reduced inequality are possible policy objectives, but the study does not measure them. Extending its conclusions to those outcomes would require evidence about what firms invent, who benefits and how technologies are deployed.

A digital leader can provide an example without creating an innovation ecosystem. The larger development opportunity emerges when other businesses can understand its experience, finance their own experiments and move beyond replication. Industrial policy should build and evaluate those connections, treating the spread of productive capability as an objective alongside the adoption of technology.