The Age of Research

7/28/2019 The Age of Research

http://slidepdf.com/reader/full/the-age-of-research 1/4

GENETICS Time for biologists toventure beyond the central

dogma p.565

SPACE Astronomer atRoyal Observatory talkscelestial images p.564

BIOGRAPHY A life of NikolaTesla — engineer, inventorand showman p.562

ENVIRONMENT Street lights thatuse less energy are better forpeople and animals p.560

The fourth age of researchJonathan Adams analyses papers from the past three decades and finds that

the best science comes from international collaboration.

R esearch has progressed through threeages: the individual, the institutionaland the national. Nations competed

to be at the cutting edge because this con-tributed to the wider economy throughknowledge, new processes and products.

Today, we are entering a fourth age of research, driven by international collabora-tions between elite research groups. This willchallenge the ability of nations to conservetheir scientific wealth either as intellectual

property or as research talent1–3. Tensions are

growing: between the knowledge a country needs to remain competitive and the assetsit can exclusively secure, and between thecollaborative and domestic parts of theresearch base. Institutions that do not forminternational collaborations risk progressivedisenfranchisement, and countries that donot nurture their talent will lose out entirely.

To explore the scale of this challenge,I analysed data on research articles andreviews from Thomson Reuters Web of

Science between 1981 and 2012. I included

papers in a country’s tally if one or moreauthor addresses included that country, butno weighting was applied for the number of authors or addresses on each paper. For the25 million papers I included in my analy-sis, I calculated the balance of internationaland domestic research collaboration forestablished economies (the United States,the United Kingdom, Germany, France,the Netherlands, Switzerland) and emerg-ing ones (China, India, South Korea, Brazil,

Poland). I looked particularly at all

I L L U S T R A T I O N B Y D A L E M U R R A Y

3 0 M A Y 2 0 1 3 | V O L 4 9 7 | N A T U R E | 5 5 7

COMMENT

© 2013 Macmillan Publishers Limited. All rights reserved



papers from the United Kingdom andthe United States that listed multiple authors.Then I probed the relative citation impact of domestic and international research in thetwo countries.

I found that, over more than three dec-ades, domestic output — papers that listonly authors from the home country — has

flatlined in the United States and in West-ern European countries. The rise in totalannual output for each country is due tointernational collaboration. As a result, thepercentage of papers that are entirely ‘homegrown’ is falling. In emerging economies, by contrast, domestic output is rapidly expand-ing (see ‘Strength in numbers’).

These shifts stand to change the dynamicsof how nations invest in science, who ownsits outputs and who best exploits them.

BORDER CROSSING

In Western Europe in 1981, domestic share of

output was about five papers in every six pub-lished by each country, but fell to half of totaloutput in 2011. The United States started asless collaborative — just 6% of US papers in1981 had a foreign co-author, but it too hashad a marked fall in domestic share of output:it is the most frequent partner for many othernations. Nonetheless, it remains less interna-tionally collaborative than Western Europe.

In emerging economies in 1981, thedomestic share of output was similar to that of the established economies. Today, about 75%of the research output of China, Brazil, Indiaand South Korea remains entirely domestic.The total volume of papers from these fourcountries has increased 20-fold — from fewerthan 15,000 papers annually in 1981 to morethan 300,000 papers now. Despite its Euro-pean location, Poland’s balance of domesticand collaborative papers is closer to this group(see Supplementary Information at go.nature.com/nszeck). Quantity is now being matchedby significant improvements in quality: thecitation impact of more than 10% of China’sdomestic research is already above twice theworld average.

For established economies, total nationalresearch output has more than doubled overthe past 30 years. Meanwhile, domestic out-

put has increased by only about 50% for allexcept the Netherlands (which more thandoubled). After the mid-1990s, the domes-tic research output of the United Kingdom(47,500 papers per year), Germany (45,000papers) and France (30,000 papers) levelledoff while international collaboration in thesecountries increased more than ten-fold.

What fuelled this growth in collabora-tion? Much of it has come from bilateralpartnerships rather than multinational pro-grammes (see Supplementary Information).For instance, less than 1% of UK papers from2003 to 2012 list co-authors from all three of

France, Germany and the United States, and

STRENGTH IN NUMBERSGrowth in international collaboration eclipses domestic output

in established economies, but not in emerging ones.

The country is less internationallycollaborative than those in Western Europe.

International collaboration has almostdoubled in the past decade.

More than three-quarters of researchoutput remains domestic.

Multinational programmes have led toabove-average growth and citations.

Even more rapid growth than China,driven by domestic research.

30

60

90

120

1981 1991 2001 2011

P a p e r s p e r

y e a r ( t h o u s a n d s )

D o m e s t i c % t o

t a l p a p e r s


t a l p a p e r s


t a l p a p e r s

D o m e s t i c

% t o

t a l p a p e r s

D o m e s t i c

% t o

t a l p a p e r s


t a l p a p e r s

P a p e r s p e r y e a r ( t h o u s a n d s )


P a p e r s p e

r y e a r ( t h o u s a n d s )



UNITED KINGDOM

Total output

Domestic papers

10

20

30

40

1981 1991 2001 2011

SWITZERLAND

25

50

75

100

100

0

0

0

0

0

0

25

50

75

100

0

25

50

75

100

0

0

25

50

75

100

0

25

50

75

100

0

25

50

75

100

0

200

300

400

1981 1991 2001 2011

UNITED STATES

50

100

150

200

1981 1991 2001 2011

CHINA

12.5

25.0

37.5

50.0

1981 1991 2001 2011

SOUTH KOREA

12.5

25.0

37.5

50.0

1981 1991 2001 2011

BRAZIL

Collaboration varies but domesticresearch is as predominant as in Asia.

5 5 8 | N A T U R E | V O L 4 9 7 | 3 0 M A Y 2 0 1 3

COMMENT

S O U R C E : T H O M S O N R E U T E R S W E B O F S C I E N C E .

A N A L Y S I S : E V I D E N C E




less than 0.1% of US output over the sameperiod has co-authors from the United King-dom, Germany and China together.

Switzerland is an interesting outlier. Fora small country, its research output has seenabove-average growth and a high average cita-tion impact. Today, more than two-thirds of its publications have a foreign co-author. A

large part of this comes from the country’shosting of multinational programmes, such asCERN (Europe’s particle-physics lab) and theWorld Health Organization, both in Geneva.

Citation impact is typically greater whenresearch groups collaborate, and the benefitstrengthens when co-authorship is interna-tional4–6. I calculated the citation impact of each UK and US paper published in 2001and then for 2011 by normalizing citationcount to take into account the time elapsedsince year of publication and the subject areaof the journal. For each of these years, I thencalculated an average normalized citation

impact (where 1.0 is the world average) forpapers that were purely domestic and forpapers with a foreign co-author.

For both countries, papers that are inter-nationally collaborative are cited relatively more often than purely domestic papers.And this ‘impact premium’ rose in bothcountries by around 20% between 2001 and2011 (see ‘Impact premium’).

A CASE IN POINT

A closer look at the United Kingdom revealsthe drivers of some of these changes. TheUK government’s Department for Business,Innovation & Skills (BIS) reports both rising volume and rising average citation impactfor the UK research base relative to the worldaverage. My analysis shows that this impact(up from 1.21 in 2001 to 1.47 by 2011) ismainly attributable to the near-doubling of its international collaboration (from 26,608to 50,423 papers with an average impact upfrom 1.48 to 1.72). By comparison, the UKdomestic research base, which BIS has neveranalysed separately, has changed much lessin quantity and quality. In 2001, the country published 46,671 domestic papers with anaverage impact of 1.05; in 2011 it published47,323 domestic papers with an average

impact of 1.21. And in 2010, the number of UK papers with no overseas co-author actu-ally fell below the count of internationally collaborative papers.

The benefits of collaboration are sys-tematic. I found that 130 UK universi-ties and colleges published more than 100research articles and reviews in journalsindexed on Thomson Reuters Web of Sci-ence in 2002–11. Around half of these universities have an international co-authoron less than 40% of their papers. Averageimpact in this group rarely gets much above1.1. By contrast, research staff in institu-

tions with greater than 50% international

co-authorship on their papers had an

average citation impact above 1.6 (seeSupplementary Information). This category includes the five or six UK universities thatare generally recognized as standing aheadof others, including the universities of Oxford and Cambridge.

This same handful of elite institutionsreceives about one-third of all availablecore research income from the UK higher-

education fundingcouncils, more thanone-third of the totalavailable researchgrant and contractincome and one-thirdof all research-councilstudentships. By con-

trast, the 65 or so UK universities that havean international co-author on less than 40% of their papers collectively account for less than10% of funding-council research money, lessthan 5% of research grants and contracts andless than 1% of research-council studentships(my analysis; using data from the UK HigherEducation Statistics Agency).

CONTEMPORARY GIANTS

First-rate researchers are, and always were,thinly distributed. In the past they could col-

laborate only intermittently. Contemporary giants call more readily across the desolateintervals of space (to paraphrase FriedrichNietzsche) and rapidly receive answers,because information technology has sharply reduced the costs of collaboration7.

Excellence seeks excellence, so elitenational universities are also leading inter-national collaborators. Exceptional researchgroups share ideas, resources and outcomes.For example, the most frequent internationalpartners of the University of Cambridge, UK,are the Max Planck institutes in Germany, theMassachusetts Institute of Technology and

Harvard University, both in Cambridge; the

California Institute of Technology in Pasa-

dena; the University of California, Berkeley;and the universities of Toronto, Heidelbergand Tokyo. Harvard’s frequent internationalpartners are Imperial College London, Uni-

versity College London, the Max Planck insti-tutes, the Karolinska Institute in Stockholmand the universities of Cambridge, Torontoand Geneva. Internationally co-authoredpapers are more highly cited because theauthors are more likely to be doing excellentresearch. Multinational programmes are stillmarginal as drivers of performance.

So, in this fourth age of research there isa growing divide between international anddomestic research. This will influence eachnation’s ability to draw on the global knowl-edge base, and could in turn compromisenational scientific wealth. For establishedeconomies, the risks are a gradual financialand intellectual separation between insti-tutions that are largely international andthose that are largely national. Such separa-tion could lead to the erosion of adequateregional competency for future researchtraining and collaboration and for knowl-edge flow to the national industrial base.

Government policy should focus on threeissues to address these risks. First, in this ageof big data that are internationally shared, the

question will be who has the skills to exploitknowledge assets fastest, not who owns them.Shared knowledge and discovery sidelinethe idea of securing intellectual property by commercial confidentiality and patenting.The growing scarcity of truly able research-ers could lead to a ‘global war for talent’8. Sogovernments must provide and sustain con-ditions that attract and support the best sci-entists, otherwise talent flows elsewhere andthere is no capacity to generate replacements.

Second, incentives must be put in placeto enable universities to participate in inter-national networks. For example, tangible

projects involving participation could be

IMPACT PREMIUMIn the United States (US) and the United Kingdom (UK), papers with at least one author fromanother country are cited more often than purely domestic work.

0

0.5

1.0

1.5

2.0

US 2001 US 2011 UK 2001 UK 2011

R e l a t i v e a v e r a g e c i t a t i o n i m p a c t

+0.22 +0.36

+0.42

+0.52

InternationalDomestic

“Elite national universities are also leading

international collaborators.”

3 0 M A Y 2 0 1 3 | V O L 4 9 7 | N A T U R E | 5 5 9

COMMENT

S O U R C E : T H O M S O N R E U T E R S W E B O F S C I E N C E .

A N A L Y S I S : E V I D E N C E




given explicit recognition and creditin systems for assessing research qual-ity (such as the UK Research ExcellenceFramework). Insight into the evolvingresearch of the emerging economies willbe limited without active engagementand collaboration. Relying simply onwhat is published in journal articles has

long been recognized as a poor mode of knowledge transfer9.Third, collaboration must include

European and US researchers in Asianand South American laboratories. Theolder economies can no longer rely onthe best foreign researchers to come and visit. Maintaining a talent pool is neces-sary, but it is not sufficient: if a nationreally wants to remain at the highest levelscientifically, it needs to get its talent outand about — and then bring it back again.One way of doing this is to create worth-while travel and exchange programmes

that include an assured post for thereturning researcher. The United King-dom has a shocking record on mobility in European research programmes andthe United States has the lowest level of international collaboration among theG7 countries. By contrast, the Nether-lands achieves excellence by enablingits researchers to be much more mobile,despite its lesser resources10.

Impact and innovation will flow from acoalition of the willing, not the straitjacketof international policy and coordination.Multinational programmes are not theanswer, nor are academic memoranda of understanding. To maintain the dividendthat governments garner from researchexcellence, they must ensure that universities and their researchers have theresources, facilities and incentives to createand sustain flourishing partnerships.■

Jonathan Adams was formerly director of research evaluation at ThomsonReuters and is now a research-management consultant in Leeds, UK.e-mail: [email protected]

1. May, R. M. Science 275, 793–796 (1997).

2. Adams, J. Nature 396, 615–618 (1998).3. Adams, J. Nature 490, 335–336 (2012).4. Van Raan, A. F. J. Scientometrics 42,423–428

(1998).5. Adams, J., Gurney, K. A. & Marshall, S.

Scientometrics 72,325–344 (2007).6. Schmoch, U. & Schubert, T. Scientometrics

74,361–377 (2008).7. Adams, J. D., Black, G. C., Clemmons, J. R. &

Stephan, P. E. Res. Pol. 34,259–285 (2005).8. Dobbs, R., Lund, S. & Madgavkar, A. ‘Talent

tensions ahead: a CEO briefing.’McKinsey Quarterly (November 2012); available athttp://go.nature.com/hz7ars.

9. Gibbons, M. et al. The New Production of Knowledge (Sage, 1994).

10. Gurney, K. & Adams, J. Tracking UK andInternational Researchers by an Analysis of Publication Data (Higher Education Policy

Institute, 2005).

A green lightfor efficiencyEfforts to improve street lights are providing a rare

opportunity to cut both financial and environmentalcosts, argues Kevin Gaston.

O

ver the past few years, local govern-ments in towns and cities across theworld have been dimming street

lights, turning them off for parts of the night,investing in more energy-efficient technolo-gies and gradually modernizing old lightingsystems. Hard times and tight budgets havehelped to drive this flurry of activity.

The combination of financial pressures, abroad array of new lighting technologies anda richer understanding of the importance of natural-light cycles to the health of many organisms presents an unusual opportunity.To take advantage, environmental scientistsmust work quickly to gather and build onexisting knowledge about the effects of artificial light on plants and animals, and

must collaborate with those responsible for

the development and implementation of public lighting schemes.

As well as disrupting the development,

physiology and behaviour of many organ-isms, grid-based electric lighting is esti-mated to produce 1.5 billion tonnes of carbon dioxide each year globally. Moreo- ver, artificial light disturbs people’s naturalcircadian rhythms. Various studies havelinked this disturbance to conditions suchas breast cancer1 and obesity. A notoriousimpediment to astronomical observations,artificial light also robs us of moonlit land-scapes and properly dark or star-filled skies.

Most existing street-lighting systems areinefficient and expensive. Much of the lightthey produce is wasted, contributing to sky-

glow often tens and possibly hundreds of

Wasted light from street lights, such as those in New York City, contributes to skyglow.

B E N O I T P E S L E / A N Z E N B E R G

E R / E Y E V I N E

5 6 0 | N A T U R E | V O L 4 9 7 | 3 0 M A Y 2 0 1 3

COMMENT

© 2013 Macmillan Publishers Limited All rights reserved

The Age of Research

Documents

Transcript of The Age of Research