Stratified Research Mobilization: Coupling Modes and Institutional Recomposition in China's Academic Patent Expansion

When a nation decides it needs to master a specific technology, it does not simply tell its scientists to work harder. It sets a direction, pours in resources, and expects the entire research system to turn toward that goal. For decades, experts have debated how this works. One school of thought sees the research world as a single machine: if you push the lever of national priority, the whole machine speeds up uniformly, producing more of everything in the same proportions. Another school argues that the research world is actually a collection of different kinds of organizations—universities, government labs, and specialized institutes—each with its own rules, funding sources, and ways of measuring success. These groups do not all hear the same signal in the same way. The question that matters is whether a national push simply makes everyone produce more, or whether it fundamentally reshapes who is doing the work and how they are doing it.

A team of researchers at Nanjing University of Science and Technology set out to answer this by looking at the massive wave of academic invention patents filed in China between 2018 and 2023. This was a period when the country made technology self-reliance a central policy goal, and the number of patent applications grew dramatically. The researchers did not just count the total number of patents to see if the system expanded. Instead, they treated the research system like a landscape with different elevations, or strata, and asked how the growth was distributed across them. They looked at nearly two million patent applications and over four million scientific articles, sorting them by the type of institution that filed them: elite universities, other universities, government research institutes, and units within the Chinese Academy of Sciences. They wanted to see if the system grew as a single block or if the pieces shifted around each other.

What they found was that the system did not just get bigger; it changed its shape. The expansion was not a uniform scaling up. Instead, the growth redistributed the work. Government research institutes, which had accounted for about 12 percent of all academic patent filings in 2018, saw their share rise to nearly 17.3 percent by 2023. At the same time, universities that were not part of the country's top-tier accreditation scheme saw their share drop from 42.9 percent to 35.0 percent. In just six years, the balance of power in the system shifted by more than thirteen percentage points between these two sectors. The system did not grow proportionally; it recomposed itself, with different types of organizations taking on different amounts of the inventive load.

The researchers also discovered that these organizations did not respond in the same way, even when they were all trying to meet the same national goals. If the system were simply reacting with different levels of intensity, you would expect every type of organization to move in the same direction on every measure. For example, if one group was "more mobilized" than another, it should have more patents, more industry partnerships, and more changes in its technology focus all at once. The data showed the opposite. The response was multidimensional and often contradictory.

Take the relationship with industry. Every group of universities increased its collaboration with companies, filing more joint patents with businesses. However, the government research institutes, which started with the highest rate of industry collaboration in the system, actually reduced theirs slightly. They did not stop working with industry; they simply grew so fast on their own internal resources that the percentage of their work involving companies went down. Meanwhile, the Academy units, which are part of the national science academy, changed their technology focus the least of all groups, yet they were the only ones whose patent growth outpaced their publication growth. They were the only group that produced more new patents relative to their new scientific papers than they had before.

Perhaps the most striking pattern appeared when the researchers looked closely at the universities themselves. They found that the relationship between patenting and publishing ran in opposite directions depending on the university's status. The most elite universities, those with the highest accreditation, increased their patent filings the most relative to their other peers. Yet, these same elite universities increased their scientific publications the least relative to their peers. The universities outside the top accreditation scheme did the opposite: they published the most new papers but filed the fewest new patents. This means that the elite universities were tilting their work more toward proprietary inventions, while the others were tilting more toward open scientific discovery. You cannot see this by looking at patents alone or papers alone; you only see it when you look at how the two move together.

The study then asked why these different groups moved in such different ways. The answer lay in the resources they used to grow. The researchers broke down the new patents to see which ones were filed with a company partner and which were filed alone. They found that the universities outside the top accreditation scheme grew almost entirely by coupling with companies. More than half of their net growth came from joint applications with businesses. In contrast, the government research institutes grew almost entirely on their own internal capacity, with very little of their new growth coming from industry partners. The elite universities sat somewhere in between, but the pattern was clear: the weaker a university's position in the official hierarchy, the more it relied on industry partnerships to expand. The research institutes, governed by administrative tasks rather than market competition, expanded by doing more of what they already did, internally.

This finding challenges the idea that a national mission simply makes everyone work harder in the same direction. Instead, it shows that a common goal is filtered through the specific structure of each organization. The universities, needing to prove their value and often lacking deep internal funding, turned to industry to fuel their growth. The research institutes, with different mandates and resources, expanded their internal output. The result was not a system that got uniformly stronger, but one that rearranged itself. The organizations that grew the fastest were not necessarily the ones that changed their behavior the most; they were the ones that found the specific path available to them.

The researchers also noted that not every part of the system behaved differently. When it came to the specific fields of technology, almost every group shifted toward computing, control, and instrumentation. This suggests that when a technological opportunity is broad and obvious, institutional differences matter less; everyone moves toward the same opportunity. But when it came to how they grew, who they partnered with, and how they balanced patents against papers, the differences were sharp and consistent.

This study does not claim to have identified the exact policy documents or budget lines that caused these shifts, but it provides a clear map of the outcome. It shows that when a nation pushes for technological advancement, the system does not just expand; it reorganizes. The growth is not a single wave rising everywhere at once. It is a complex redistribution where some sectors take on more of the load, some deepen ties with industry, and others rely on internal capacity. The result is a system that looks different at the end of the period than it did at the beginning, not just in size, but in its very composition. The lesson for policymakers is that a single national goal will not produce a single type of response. The outcome depends on the specific shape of the organizations that are asked to deliver it.