Available via license: CC BY 4.0
Content may be subject to copyright.
Mapping academic migration of German-affiliated researchers
across countries using 8 million Scopus publications from 1996
to 2020
Xinyi Zhao a∗
, Samin Aref a, Emilio Zagheni a, Guy Stecklov b
aLaboratory of Digital and Computational Demography, Max Planck Institute for Demographic Research, Germany,
Xinyi Zhao - zhao@demogr.mpg.de, Samin Aref - aref@demogr.mpg.de, Emilio Zagheni - zagheni@demogr.mpg.de
bDepartment of Sociology, University of British Columbia, Canada, Guy Stecklov - guy.stecklov@ubc.ca
* Corresponding author
Keywords: Academic migration, Density equalizing cartograms, Bibliometric data, Germany
Abstract:
Academic mobility plays a predominant role in the context of increasing the internationalization of science and research
production, thus creating a global market of qualified professionals (Franzoni et al., 2015, Schiller and Cordes, 2016,
Netz and Jaksztat, 2017). Germany has launched a series of programs that have the aim to attract researchers with high
citation performance and link German-affiliated scientists abroad with each other (Schiller and Cordes, 2016). Thus,
mapping the German academic in- and out-flow patterns across countries and disciplines can help clarify the position and
attraction of Germany in the global science system. Taking all fields of scholarship, this analysis speaks directly to policy
development. Using large-scale bibliometric data from over 8 million Scopus publications during the period 1996-2020,
we analyze and visualize the international migration to and from Germany among published researchers by taking into
account their citation information and semantically identified disciplines. This reveals the spatial hotspots of migration
and in- and out-flows for all fields of scholarship.
The large-scale bibliometric data in our study includes an exhaustive set of over one million researchers who have pub-
lished with an affiliation address in Germany at some point during the 1996-2020 period in Scopus-indexed publication
venues. The raw bibliometric data from over 8 million Scopus publications are firstly pre-processed to infer missing coun-
try information using a neural network algorithm, and to disambiguate authors using an unsupervised machine learning
algorithm. We assign the country of academic origin and the country of academic destination to each author according to
their most frequent (mode) country(ies) during the first year of publishing and that of the most recent year of publishing.
Accordingly, immigrant researchers to Germany (origin: not Germany; destination: Germany) and emigrant researchers
from Germany (origin: Germany; destination: not Germany) can be detected. We illustrate the academic migration in-
and out-flow pattern (figure 1, 2) using density equalizing cartograms (Hennig et al., 2012), which transforms the shape
of map polygons proportional to the total migration inflows (outflow) of researchers moving to (from) Germany. The
total citations for each researcher are determined by the total number of citations (as of April 2020) for all publications
authored by a researcher. We further calculate the average annual citation rate for each researcher by dividing their total
citations by their academic age (the number of years since first publication until 2020). The divisions of the average
annual citation rates of mobile researchers to Germany and from Germany are represented by different colors in the two
maps and explained in the legends.
A Latent Dirichlet allocation (LDA) model is used to determine the main discipline for each German-affiliated migrant
researcher, by identifying the semantic topic of bibliometric information composed of the publication titles, journal titles,
and keywords authored by a researcher. We further evaluate and visualize the migration preferences of mobile researchers
across different disciplines, and use kernel density estimation to detect the respective hotspots of the origins of migration
inflows to Germany and the destinations for migration outflows at different scales.
In the next steps of this work, we intend to gain further insights into the role of migration both into and out of Germany
in each field and discipline from a micro perspective. Methodologically, our computational approach of processing,
analyzing, and visualizing large-scale bibliometric data can be adopted as a framework of analysis for studying scholarly
migration in other countries.
Acknowledgements
This study has been funded by the German Academic Exchange Service with funds from the Federal Ministry of Ed-
ucation and Research. This study has received access to the bibliometric data through the project “Kompetenzzentrum
Bibliometrie”, and the authors acknowledge their funder BMBF (funding identification number 01PQ17001).
Abstracts of the International Cartographic Association, 3, 2021.
30th International Cartographic Conference (ICC 2021), 14–18 December 2021, Florence, Italy.
https://doi.org/10.5194/ica-abs-3-327-2021 | © Author(s) 2021. CC BY 4.0 License.
2 of 2
USA
FRA
ITA
GBR
USA
RUS
ESP
CHN
CHE
AUT
AUS
IND
BEL
SWE
BRA
CAN NLD
EGY
POL
IRL
IRN
CZE
JPN
FIN
MEX
HUN
TUR
UKR
GRL
NOR
SVN
ARG
HRV
PRT
ROU
GRC
ZAF
SVK
SRB
CAN
PAK
CHL
SAU
JOR
CAN
KOR
ISR
JPN
BGR
BIH
DNK
COL
BLR
NZL
CAN
THA
VEN
ETH
LBY
ISL
IDN
NZL
IDN
DZA KEN
SYR
SDN IDN
MAR
TZA
JPN
MDG
PER
LTU
NGA
IRQ
ALB
KAZ
VNM
LUX
MOZ
LVA
PNG
NAM
MMR
MYS
BGD
IDN
CUB
IDN
PRK
OMN
EST
MNG
MYS
NPL
ZMB
PHL
TUN
JPN
COG
NIC
BOL
MRT
CIV
YEM
AFG
DNK
PHL
GEO
ERI
URY
LKA
ECU
PAN
CRI
GTM
DNK
CAF
DNK
KHM
CHL
DOM
FJI
IDN
NCL
IDN
LIE
IDN
IDN
FLK
PHL
SWZ
PNG
ECU
BHS
VUT
REU
GLP
FRO
GUM
FSM
MCO
SYC
SYC
SHN
FSM
MUS
Ü
Legend
A
verage annual citation
rate of immigrants to
Germany
79.46 - 321.87
65.99 - 79.45
55.48 - 65.98
43.01 - 55.47
33.04 - 43.00
26.06 - 33.03
18.60 - 26.05
14.15 - 18.59
8.09 - 14.14
0.13 - 8.08
Figure 1. Migration inflows to Germany (proportional to the area) and average annual citation rates of immigrant re-
searchers to Germany.
USA
FRA
ITA
USA
GBR
CHE
CHN
CAN
AUT
ESP
AUS
IND
SWE
BEL
BRA
RUS
EGY
NLD
IRL
POL
CZE
JPN
FIN
MEX
NOR
GRL
SVN
HUN TUR
IRN
SAU
HRV
ARG
ZAF
PRT
ISRJOR
CAN
PAK
GRC
ROU
UKR
DNK
CAN
SVK
CHL
JPN
KOR
SRB
THA
NZL
BIH
BGR
COL LBY
CAN
ETH
IDN IDN
ISL
KEN
NZL
SDN
DZAVEN
IDN
SYR VNM
BLR
TZA
MAR
PER
AGO MDG
IRQ
NGA
JPN
KAZ
LUX
LTU
MOZ
MMR
PNG
MYS
OMN
LVA
IDN
ALB
AFG
IDN
MYS
MNG PRK
TUN
CUB
ARE
YEM
GHA
PHL
DNK
COG
NPL
JPN
MRT
ECU
BOL
GUY
NIC
PHL
ERI
AZE
LKA
URY
DNK
KHM
CRIPAN
DNK
GAB
CAF
HND
CHL
BWA
DOM
FJI
IDN
NCL
LIE
IDN
PHL PHL
PNG
ECU
REU
GLP
WSM
PYF
CPV
FRO
GUM
PYF
MCO
MUS
BMU
MDV
MDV
Ü
Legend
Average annual citation
rate of emigrants fro m
Germany
87.47 - 162.98
72.49 - 87.46
53.62 - 72.48
43.89 - 53.61
34.92 - 43.88
28.70 - 34.91
22.87 - 28.69
14.23 - 22.86
9.64 - 14.22
0.00 - 9.63
Figure 2. Migration outflows from Germany (proportional to the area) and average annual citation rates of emigrant
researchers from Germany.
References
Franzoni, C., Scellato, G. and Stephan, P., 2015. Chapter 2 - international mobility of research scientists: Lessons from
GlobSci. In: A. Geuna (ed.), Global Mobility of Research Scientists, Academic Press, pp. 35–65.
Hennig, B. D., Pritchard, J., Ramsden, M. and Dorling, D., 2012. Remapping the world’s population: Visualizing data
using cartograms. In: Fair Play: A Daniel Dorling reader on social justice, Policy Press.
Netz, N. and Jaksztat, S., 2017. Explaining scientists’ plans for international mobility from a life course perspective.
Research in Higher Education 58(5), pp. 497–519.
Schiller, D. and Cordes, A., 2016. Measuring researcher mobility. In: OECD Blue Sky Forum, p. 24.
Abstracts of the International Cartographic Association, 3, 2021.
30th International Cartographic Conference (ICC 2021), 14–18 December 2021, Florence, Italy.
https://doi.org/10.5194/ica-abs-3-327-2021 | © Author(s) 2021. CC BY 4.0 License.
