North and western Europe is second to North America overall in the index, and is ahead in terms of the number of physical sciences papers.

Germany leads the region by article count (AC) and weighted fractional count (WFC) – a measure that adjusts AC to reflect the relative contribution of each country or institution – and is also the country that spends the most of its gross domestic product (GDP) on research and development (R&D). Over the last decade, this percentage has increased from 2.5% in 2005, to just over 2.9% in 2012. Germany's Chancellor Angela Merkel, herself a physicist, came to power in 2005 and is widely credited with pushing these priorities. During this time, the government's spending on science has risen by 60% to €14.4 billion (US$18.5 billion), including annual increases for the large German science organizations such as the Max Planck Society (MPS).

Germany's spend as a fraction of GDP is above the average of 1.97% for the 28 European Union (EU) countries, including 2.26% for France and 1.72% for the United Kingdom, according to the Organisation for Economic Cooperation and Development (OECD). “Most EU countries have increased their R&D to GDP spending in the last 10 years, although at a slower rate since the 2008–09 economic crisis,” says Dominique Guellec, head of country studies and outlook at the OECD's Science, Technology and Innovation directorate. “The one exception is the United Kingdom, where the ratio has stagnated over the past decade.”

Lidia Borrell-Damián, Director for Research and Innovation at the European University Association (EUA) in Brussels, says that Europe's R&D budgets at a regional level may be on a par with North America and East Asia, but the disparity of investment between European countries creates a problem. “Some are investing a lot, and others are investing almost nothing,” she says.

The EU is trying to spur R&D spending with its Horizon 2020 Research and Innovation programme, which will provide nearly €80 billion (US$100 billion) in funding between 2014 and 2020. But this works out as a modest annual investment compared with the €260 billion the 28 EU countries spent on R&D in 2012, Guellec notes.

North and Western Europe's success comes partly from its ability to leverage its researcher base. In terms of efficiency (shown here as WFC per 1,000 researchers), many countries in this region are among the strongest in the world. Switzerland tops the region by an enormous margin, and is also the highest in the Nature Index (see 'Researcher efficiency'). It regularly tops global surveys of competitiveness and innovation, and its science spend as a proportion of GDP is above the EU average.

The top universities in Switzerland are the two parts of the Swiss Federal Institutes of Technology: the Federal Institute of Technology Zurich (ETH) and the Federal Polytechnic School of Lausanne (EPFL), which come sixth and eighth in the region respectively. Philippe Gillet, provost of EPFL, outlines factors that contribute to Switzerland's success: unlike France, for example, establishments in Switzerland have autonomy and total freedom over scientific and education policy; higher education, research and innovation are a priority for the country; and part of the polytechnics' mission is to help spur the economy.

Gillet moved from his native France to Switzerland specifically to experience the system “which works”. Many countries, such as France, proclaim that higher education and research are a priority, he says, but “they do not follow through with enough resources and structural reform”.

Scientists in North and Western Europe are less likely to collaborate with researchers from other regions than the global average (see 'Collaboration rate'). The breadth of languages (23 official ones) in the EU is not necessarily the culprit in lack of scientific collaborations, says Bernard Meunier, vice president of the French Academy of Sciences. He says that it is very much determined by discipline. “In particle physics, about 500 scientists collaborate in projects at the European Organization for Nuclear Research (CERN), whereas in biology, chemistry and pharmacology, research teams are in competition with each other,” he explains. “Astrophysicists work with colleagues from around the world, but mathematicians lead a solitary life.”

Another reason for the low level of collaboration in the EU could be the heavy administrative burden for researchers navigating different funding structures. The red tape can discourage researchers from looking abroad for collaborators, says the EUA's Borrell-Damián. At a recent university conference in Italy, one rector said the research projects in progress at his university involved 150 different funding regimes, which created huge management problems. “I am sure he is not alone in this, and that large universities would have even more than 150,” Borrell-Damián says. “The administrative requirements should be rationalized.”

Increasing collaboration is an attractive goal, given that benefits include enriching scientists' careers, raising the profile of universities and conducting research in the best conditions possible. International cooperation means that “universities are competitive worldwide and are at the forefront of their field and even driving it,” she says. That is “good for researchers, good for institutions and good for the advancement of science.”

Subjective spread

Of the countries in this region, the United Kingdom is most heavily skewed towards life sciences, which comprises a third of its output in the index. On its own, the United Kingdom contributes to 7% of all the life sciences research in the index (the largest after the United States). Of the seven European institutions in the top 50 globally for life sciences, three are in the United Kingdom, led by the University of Cambridge (see life sciences table, page S104).

Duncan Maskell, Head of the School of Biological Sciences at the University of Cambridge, says that the UK's strength in this field is the result of a supportive political and financial environment. “The government appears to understand the contribution that science – and in particular the life sciences – makes to people, and currently fosters an environment conducive to research. It's encouraging to see that it now even has a minister focused entirely on the life sciences, for example.”

Germany's strength lies in the physical sciences, where it makes up 10% of the Nature Index — on par with China. Its largest research institution, the MPS, is second only to the Chinese Academy of Sciences in WFC in this field, and MPS has overall contributed to more papers in the Nature Index.

Christoph Ettl, senior scientist in the Presidential Division of the MPS, says Germany has a very strong history of physics, with Max Planck, Albert Einstein and Werner Heisenberg among others having contributed to the bedrock of quantum mechanics. Until 1976, physicists born in Germany topped the list of Nobel laureates. And the discipline has many applications beyond science, says Ettl — who is also a physicist. “Physicists have the quantitative tools for simulation, applicable to almost any field from society to the stock market.”

Institutional comparison

In most European countries, the production of papers is spread between many institutions. In the United Kingdom, universities — or tertiary educational institutions — produce the bulk of the country's research in the index; in Germany, it tends to be research-only societies and associations that dominate, including MPS, and the Helmholtz and Leibniz Associations.

The story is very different in France, where the National Centre for Scientific Research (CNRS) towers above all other institutions with a WFC of 719 — similar to Germany's MPS at 726 (see France). Then there is a notable drop-off: the next French research institution, the Atomic Energy and Alternative Energies Commission, has less than a fifth of this score, and the highest placed university is the Pierre and Marie Curie University (UPMC) in Paris, with a WFC of 138 (see 'Institutional subject spread').

Physicist Paul Indelicato, President of Research and Innovation at UPMC, does not believe the enormity of the CNRS has a negative impact on his or other research institutions. “On the contrary, it is a strength,” he says.

Comparing like-for-like single institutions across the three leading Northwest European countries pits Germany's Ludwig Maximilian University of Munich (LMU) against UPMC and the University of Cambridge.

UPMC has strong output in the physical sciences (see 'Institutional subject spread'), particularly in space sciences: more than a third of its 1,319 papers in the Nature Index are in astrophysics journals. Indeed, astronomy is a historical legacy at UPMC through its longstanding links with the Paris Observatory. UPMC founded its Paris Institute of Astrophysics in the 1930s, which a few years ago became a joint lab with the CNRS. However, because of the down-weighting of astrophysics journals in the Nature Index (see 'Introducing the index', page S52), UPMC ends up with a WFC of only 95. UPMC's namesakes — Pierre and Marie Curie — are of course famous for research in the physical sciences. “The spirit in which they conducted research still has an important influence on how we work today,” says Indelicato. “We are extremely attached to Marie Curie's name and to her demand for quality research.” Since the 1960s, the UPMC's mixed labs have produced three Nobel laureates in quantum physics.

LMU, in contrast, has a more even distribution of publications across several journals, which translates into an even subject split between chemistry, physics and biology, each of which comprises roughly a third of its WFC (overlap between the subjects means that the total can be more than 100%).

Of the three, Cambridge is the strongest in the life sciences, with a WFC of 151, three-times that of LMU's. It also leans marginally more towards this field in its overall subject spread. A lot of this strength comes from neuroscience: the Journal of Neuroscience is one of the university's top journals by WFC. Cambridge's Maskell identifies the university's cross-disciplinary activities, including its Strategic Research Initiatives (one of which is in neuroscience), as helping to make it a leader in this field. “We have excellent opportunities to go from very basic research right through to translation, across the spectrum from areas like agricultural and conservation biology to clinical medicine.” Notably, Cambridge also has the most wholly-authored papers of the three: 138 compared to 53 from LMU and none from UPMC.

When it comes to the highest impact science, both Cambridge and LMU are strong in publications in Nature and Science, well above the global average (see 'Nature and Science split').

In raw numbers, Cambridge is more successful at publishing in these two journals, but in terms of proportion of its overall output there is little to choose between the UK and German institutions.

Less-traditional measures of the impact of research, called altmetrics, track a paper's visibility in the online world including social media. All three west European institutions have contributed to papers that have scored highly, according to altmetrics.com, but it is Cambridge that appears to be the most successful. Its top paper is from 2013, involving two authors from Cambridge's Psychometrics Centre, along with a researcher from the nearby Microsoft lab. The paper seems made for social media: it describes how a variety of personal information can be gleaned by reviewing someone's Facebook “likes” (see 'Cambridge's online visibility').

Altmetrics may be a new rating system, but Maskell sees its benefit for the university. “A high score in altmetrics reflects a high level of engagement with our research, within the sector and with the wider public. We cannot expect the public to support research if they don't know what we're doing.”