Collaborative research can be measured through the international affiliations attached to scholarly articles. The figures below compare selected countries’ shares of U.S. international articles with a Chinese coauthor in five benchmark years: 2004, 2010, 2014, 2018, and 2024. The measurements are reported in NSB-SEP-2026 Table S18; they are shares, not counts of articles or estimates of the quality of collaboration.
Contents
- How to read these collaborative research statistics
- The five-year comparison
- China’s changing share
- European partners
- Canada and Australia
- India and South Korea
- What the 2024 snapshot shows
How to read these collaborative research statistics
Each percentage describes one country’s share of U.S. international articles with a Chinese coauthor at the stated measurement year. The denominator and the collaboration relationship therefore matter: these are not each country’s share of all global research, and they are not a direct measure of the number of coauthors.
The years are unevenly spaced. The series moves from 2004 to 2010, then to 2014, 2018, and 2024. A value for 2024 is a measurement for that year in the cited table, not a forecast. Likewise, the older observations are presented as historical measurements from the same table; they should not be treated as independently rechecked here.
The comparison is also geographically specific. It concerns U.S. international articles and identifies a Chinese coauthor. A country can therefore appear with a relatively high or low share in this particular collaboration-linked set without that percentage describing its entire research system.
Source label: NSB-SEP-2026 Table S18.
The five-year comparison
The compact table below puts all 10 reported countries beside one another. Keeping the years in columns makes it easier to distinguish a sustained direction from a short-term change.
| Country | 2004 | 2010 | 2014 | 2018 | 2024 |
|---|---|---|---|---|---|
| China | 7.02% | 14.20% | 19.64% | 25.50% | 22.53% |
| United Kingdom | 12.87% | 12.61% | 13.08% | 13.49% | 15.15% |
| Germany | 12.47% | 12.00% | 11.53% | 11.05% | 11.83% |
| Canada | 11.56% | 11.28% | 10.69% | 10.40% | 11.78% |
| Italy | 6.48% | 6.76% | 6.58% | 6.44% | 7.57% |
| France | 7.95% | 7.74% | 7.68% | 7.19% | 7.18% |
| Australia | 4.76% | 5.33% | 5.97% | 6.50% | 6.84% |
| India | 2.41% | 3.01% | 3.42% | 3.75% | 6.79% |
| South Korea | 4.53% | 5.36% | 5.22% | 4.54% | 5.34% |
| Netherlands | 4.09% | 4.45% | 4.75% | 4.73% | 5.32% |
Source label for every value: NSB-SEP-2026 Table S18.
The table shows several different shapes. China rises across the first four observations before a lower 2024 value. The United Kingdom ends at its highest reported value. Germany and Canada both move down through 2018 and then show higher 2024 values than in 2018. Australia increases at every listed observation. France trends lower across the series, while Italy finishes above its 2004 value after relatively narrow movement in the middle years. India records a marked increase in its final observation, and South Korea and the Netherlands remain within comparatively narrow ranges.
China’s changing share
China’s reported share was 7.02% in 2004. It was 14.20% in 2010 and 19.64% in 2014. The value reached 25.50% in 2018, the highest Chinese observation in the supplied series. In 2024, the share was 22.53%.
This sequence matters because it contains both a long rise and a later lower observation. The 2018 figure should not be described as the 2024 level: the two measurements are 25.50% and 22.53%, respectively. Nor should the 2024 figure be treated as a forecast of a future share. It is the reported 2024 measurement in NSB-SEP-2026 Table S18.
China’s five observations also provide the clearest example of why a single year can give an incomplete picture of collaborative research. Looking only at 2004 would understate the later values in this series; looking only at 2018 would miss the lower 2024 observation. The useful description is therefore a time-ordered one: 7.02% in 2004, 14.20% in 2010, 19.64% in 2014, 25.50% in 2018, and 22.53% in 2024.
European partners
The United Kingdom’s share was 12.87% in 2004 and 12.61% in 2010. It then measured 13.08% in 2014, 13.49% in 2018, and 15.15% in 2024. The final value is the United Kingdom’s highest observation in this five-year series.
Germany began at 12.47% in 2004. Its share was 12.00% in 2010, 11.53% in 2014, and 11.05% in 2018, before measuring 11.83% in 2024. The German series thus has a lower 2018 observation than its 2004 and 2024 observations.
France’s reported values were 7.95% in 2004, 7.74% in 2010, 7.68% in 2014, 7.19% in 2018, and 7.18% in 2024. The two latest measurements are nearly identical in the supplied data, with 7.19% followed by 7.18%.
Italy measured 6.48% in 2004, 6.76% in 2010, 6.58% in 2014, 6.44% in 2018, and 7.57% in 2024. Its middle observations stay close to the mid-six-percent range, while the 2024 value is higher than each of the three preceding Italian measurements.
The Netherlands recorded 4.09% in 2004 and 4.45% in 2010. The share reached 4.75% in 2014, was 4.73% in 2018, and measured 5.32% in 2024. This is a gradual upward sequence overall, with only a small difference between the 2014 and 2018 observations.
All five European series describe shares within the same U.S.-article and Chinese-coauthor context. They should therefore be read as a common measurement frame, not as standalone national rankings of research collaboration.
Canada and Australia
Canada’s share was 11.56% in 2004, followed by 11.28% in 2010 and 10.69% in 2014. It measured 10.40% in 2018, then 11.78% in 2024. The Canadian values are lower in 2010, 2014, and 2018 than in 2004, while the 2024 observation is higher than all three of those intervening values.
Australia shows a different pattern. Its share was 4.76% in 2004, 5.33% in 2010, 5.97% in 2014, 6.50% in 2018, and 6.84% in 2024. Every listed Australian observation is higher than the preceding one. Within this series, the 2024 measurement is the highest and the 2004 measurement is the lowest.
These two country profiles illustrate why the word “trend” needs a time frame. Canada’s sequence includes a decline through 2018 followed by a higher 2024 measurement. Australia’s sequence rises at every benchmark year. Both statements are limited to the percentages reported for this collaboration-linked group and do not describe causes.
India and South Korea
India’s share was 2.41% in 2004. It rose to 3.01% in 2010, 3.42% in 2014, and 3.75% in 2018. The 2024 value was 6.79%. The first four reported observations form a gradual upward sequence, while the final observation is substantially higher than the preceding Indian value in the table.
South Korea measured 4.53% in 2004 and 5.36% in 2010. Its share was 5.22% in 2014, 4.54% in 2018, and 5.34% in 2024. The series therefore includes a higher 2010 value, a lower 2018 value, and a higher 2024 observation than in 2018.
The contrast between India and South Korea is useful when reading international collaboration statistics. India’s listed values rise from one benchmark to the next, ending at 6.79%. South Korea’s values move within a narrower band, beginning at 4.53% and ending at 5.34%. These are descriptions of the reported percentages, not explanations for why the countries differ.
What the 2024 snapshot shows
The 2024 observations range from 5.32% for the Netherlands to 22.53% for China among the countries listed here. The United Kingdom is reported at 15.15%, Germany at 11.83%, and Canada at 11.78%. Italy and France are reported at 7.57% and 7.18%, respectively. Australia is at 6.84%, India at 6.79%, and South Korea at 5.34%.
Read alongside the earlier observations, the 2024 snapshot contains several distinct recent positions. China is below its 2018 value of 25.50%. The United Kingdom is above its four earlier observations. Germany and Canada are above their 2018 observations. Australia, India, and the Netherlands end at their highest reported values. France ends at 7.18%, nearly the same as its 2018 value of 7.19%, while South Korea ends above its 2018 value of 4.54%.
The figures support careful comparisons of measured shares across years and countries. They do not, by themselves, establish article counts, causal relationships, research quality, or the reasons behind any country’s movement. For those questions, this table should be treated as a starting point for further source-specific investigation.