Figure 3 shows the trend toward teamwork in journal articles and patents. The mean
number of authors (papers) or inventors (patents) are plotted over time. The journal article data
incorporate 20 million publications since 1955 as indexed by the Web of Science. The patent
data consider 2.1 million patents issued by the U.S. Patent and Trademark Office from 1975-
1999.
We see general increases in teamwork across Science and Engineering journal
publications, Social Sciences journal publications, and patenting. Mean team sizes have risen
quickly, at rates of 15-20% per decade. The shift toward teamwork also appears in virtually all
subfields of research and invention, including 170 of 171 science and engineering subfields, 54
of 54 social science subfields, and 36 of 36 patenting subfields (see Wuchty, Jones, and Uzzi
2007 and Jones 2009). In percentage terms, over 80% of Science and Engineering publications,
over 50% of Social Sciences publications and over 60% of patents had multiple authors or
inventors by 2005, with the frequency of teamwork rising rapidly in all three areas. As with the
life-cycle patterns, we see exceptionally general trends.
Figure 4 presents additional analysis of individual specialization. Figure 4A considers
the probability that a solo inventor jumps to a new primary technological field across consecutive
patents. A declining tendency to switch fields suggests that individuals are more specialized.
Figure 4A shows that solo inventors appear more specialized with time when considering all
patenting fields; this tendency also appears in 34 of 36 technology subfields when analyzed
separately (see Jones 2009). By contrast, Figure 4B shows that, when operating in teams,
inventors move across technological fields with the same frequency over time - so that
teamwork appears to overcome the increasing narrowness of individuals.
The relationship between teamwork, specialization, and the depth of knowledge is further
supported when comparing fields at a point in time. Comparing across fields, Jones (2009)
shows that deeper areas of knowledge are associated with more specialization and more
teamwork. Thus, the time trends and the cross-sectional field differences can be interpreted in a
consistent and simple manner.
12