Modeling Spin-off Creation in University Technology
Transfer with System Dynamics
Mutiara Laksminingrum Sidharta’?, Takeshi Arai‘,
Utomo Sarjono Putro’, and Hidetsugu Morimoto!
1Department of Industrial Administration, Faculty of Science and
Technology
Tokyo University of Science, Japan
School of Business and Management, Institut Teknologi Bandung,
Indonesia
laksminingrum@sbm-itb.ac.id, tarai@rs.noda.tus.ac.jp,
utomo@sbim-ith.ac.id, mOrim0t0@rs.tus.ac.jp
ABSTRACT
University spin-offs become more and more preferred in recent years as
an option to transfer university technology to the market due to its social
economic advantage. As a result of Japanese government policy in 2001,
number of university spin-offs increased at the beginning yet declined
after five years. By focusing on university level, this paper explores
activity of technology transfer from research project as initial step of
technology transfer to university spin-off as one of technology transfer
channel. System dynamics is used to model the system. The proposed
model allows intervention to the activity in form of initial capital and
support on business model and practice to increase number of new
university spin-off. Simulation results show that both factors positively
correlates with number of new university spin-off. In addition, TLO
support is found to have higher influence than initial capital in longer
term.
Keywords: University technology transfer, Spin-off, System dynamics
1. INTRODUCTION
In recent years, more attention has been given to new firms founded by
university faculty and staffs to exploit intellectual property (university
spin-offs), as it is believed to have importance in regional economic
growth through knowledge spillover. University spin-offs also open wider
chance of employment, which cannot be met by established firms.
Further, with good management university spin-offs may grow into major
firms, naming Genentech in biotechnology, Cirrus Logic in
semiconductors, and Lycos in Internet search engines in the line (Di
Gregorio & Shane, 2003). Therefore, not only as technology transfer
channel, university spin-offs also contribute in helping the economy.
Compared to the United States and the United Kingdom, Japan is very lack
behind in number of university spin-offs (Figure 1) although it leads in
number of patent application per year (The World Bank, 2014). To
encourage the creation of university-launched ventures’, Japanese
government assigned ‘1000 university ventures plan’ (Hiranuma Plan) in
2001. This plan reached its goal within three years where the plan
successfully doubled the number of new companies (Figure 2) (Miki, 2012)
. However, unexpected declining trend followed the peak without any sign
of significant rebound until recent years. This phenomenon was also
recorded in a survey on university ventures with high sales volume as
shown in Figure 3 (Teikoku Data Bank, 2013).
In the latest plan to improve the economy, Japanese government had once
again put expectation toward university ventures as one strategy in
strengthening science and technological innovation (Teikoku Data Bank,
2013). Reflecting to past experience, appropriate efforts should be
considered to support creation of university ventures. Although
government provides large budget to be used, e.g. for capital, any
significant results would not be apparent if not supported by other
aspects, including business model, market research, and excellent
management team.
This paper tries to unravel the process leading to university spin-off
creation (type 1 in Japanese university venture categories). Focusing on
institution (university) level, this paper explores possible introduction to
increase number of university spin-offs. The policy would be in the form
of initial capital and Technology Licensing Office (TLO) support where
1in Japan, the term ‘university-launched venture’ or ‘university venture’ includes four categories,
ie. 1) firms that are founded by university faculty, researchers, postdocs, students or graduate
students based on patented technology; 2) other than 1), firms that are founded based on
non-patented technology; 3) firms that are founded by involving university faculty, researchers,
postdocs, students or graduate students; 4) firms that are founded due to TLO investment (Ogura,
2009). The term ‘university spin-off used in this paper refers to university venture type 1.
)
2
positive correlation is expected.
Founding university spin-off is a dynamic process happens in a highly
complex environment. It involves numerous interactions within
university and with external environment, which may be subtle enough
to be easily pointed out. Time lag may also occur between action and
result, adding complexity to the process, especially regarding
Ee
consequences of one policy.
Figure 1. Number of Japanese newly established university spin-off
in comparison with US and UK (Adapted from Ogura, 2009)
Pp
Figure 2. Number of newly established university ventures in Japan (Miki, 2012)
Figure 3. Number of university ventures with high sales volume according to year of
establishment
(Teikoku Data Bank, 2013)
System dynamics (SD) model is employed to model the problem. SD model
is characterized by feedback loops and stock and flows to model behavior
over time of a complex system (Wu et al., 2010). Model structure, time
delays, and amplification, which occurs through feedback affect SD model
(Thompson & Bank, 2010). There has been no study so far about
university spin-off using SD model. Through simulation, SD model can be
used to help decision makers in making policy for the problem in question
by testing different strategies.
The proposed model is tested by a case study of Tokyo University of
Science (TUS). By applying data of university spin-offs, results show that
introduction of initial capital and TLO support does influence number of
new university spin-off. It may give perspective for future policy
regarding both factors.
The rest of the paper is organized as follow. Section 2 consists of literature
about university spin-off. Section 3 explains current condition of
technology transfer in TUS. Section 4 presents a SD model of technology
transfer from university, which illustrate interactions among variables in
the system. Section 5 demonstrates the model through simulation using
the software Stella 9.1.3. Section 6 is conclusions.
2. TECHNOLOGY TRANSFER CHANNEL: UNIVERSITY SPIN-OFF
The enactment of Japanese Bayh-Dole Act in 1998 became the milestone in
the trend of entrepreneurial university in Japan. University is no longer
seen solely as an entity where teaching and research takes place; its role
expands to what is called as ‘third mission’, that is economic and social
development through commercial activities including patenting, licensing,
and creating company (Baycan & Stough, 2013; Etzkowitz, 2003). This
highlights the importance of technology transfer in the way to
commercialize knowledge produced in university.
University technology transfer process passes on scientific and technical
knowledge, in the form of invention or intellectual property, from one
party (university) to another (for-profit entity) to gain economic
advantage (commercialized) (Friedman & Silberman, 2003; Sheft, 2008).
University technology transfer allows scientific and technological
developments accessible to wider range of users who can then transform
it into products, processes, applications, materials or services (Mitasiunas,
2013). Not just accelerating the process of moving university-produced
knowledge to the market place, university technology transfer also
contributes largely in realizing knowledge-based economy of one country.
University technology transfer process consists of several steps as
explained by Friedman & Silberman (2003) (Figure 4). After disclosure of
invention by university faculty, TLO conducts necessary assessment to
measure potential marketability of the inventions. The office then submits
patent application for technology with good prospect in the market. The
right to exploit patent-protected technology may be transferred through
licensing agreement to an individual or company who is interested to
further develop the early-stage technology. University earns licensing
income once the technology is commercially produced and profitable.
Particularly for spin-off and start-up, direct impact on job and wealth
creation is suggested in the model.
To clearly define university spin-off which become the focus of this paper,
we follow Bradley, Hayter, & Link (2013, p.17) who distinct spin-off from
start-up. Spin-offs are defined as ‘new companies formed by individuals
(faculty members) related to the university or university research park to
develop a technology that was discovered in, and is transferred from, the
parent organization’, while start-up companies are “created by licensing
an embryonic invention to an independent entrepreneur (who is not
necessarily a university faculty member), with the goal of developing the
company around the growth and commercialization of the technology’. In
summary, university spin-offs involve two main points (Breznitz et al.,
uo
2008): 1) transfer of technology from university to new company; 2)
inventor academic(s) who may or may not be currently affiliated with
university are part of founding member(s).
Spin-offs are exceptionally important in the topic of academic
entrepreneurship. Spin-offs are more likely to develop basic research
technologies that are not favored by established companies due to its less
profitability or which market has not readily available (Swamidass &
Vulasa, 2009). Through spin-offs, the gap between university research and
industrial commercialization may be reduced. Furthermore as mentioned
earlier in this paper, spin-offs also bring social and economic advantages,
including employment creation, especially for high-educated graduates
and strengthening the local economy (Peng, 2006; Szetre et al., 2009). In
terms of early-stage technology development before ready to be launched
to market, spin-offs are benefited from the involvement of inventor(s) in
creating the company from the beginning due to the tacit knowledge of
the invention they possess.
Regarding university spin-offs creation, Di Gregorio & Shane (2003)
mentioned about micro- and macro-level factors influencing the decision
to create new company intended to exploit university invention. While
micro-level factors explain about individual-related determinants,
macro-level factors deal with wider scope of determinants, including
organizational, institutional, and external determinants (O’Shea et al.,
2008). Despite the importance of both factors, this paper only discusses
the effect of macro-level factors on university spin-off creation rate
corresponding to boundary of the system described in next section.
Intellectual eminence and university policy are two factors, which are
suggested to increase creation of new spin-offs (Di Gregorio & Shane,
2003). O’Shea et al. (2008) mentioned tangible and intangible factors
related to spin-off creation, such as university funding for R&D activities,
nature of research, faculty quality, university policy, and role of
Technology Transfer Office (TTO). Entrepreneurship climate is important
to support the creation of new companies, thus efforts to foster the
climate to grow are necessary, such as constructing reward system for
TTO staffs, increasing competencies of TTO staffs, designing flexible
university policies, allocating resources to TTO, and eliminating cultural
and informational barriers that impede technology transfer process
(Siegel et al., 2004).
In a study involving seven European universities, GOmez Gras, Galiana
Lapera, Mira Solves, Verdu Jover, & Sancho Azuar (2007) found that the
6
excellence of academic staff and financial support available in
universities were two main factors associated with spin-off number and
performance. The authors also mentioned about university’s support on
training and advice before and during first start-up stages, as well as
infrastructure support. In addition, the importance of qualified TTO staffs
with marketing, technical, and negotiation skills, more than just number
of personnel was further underlined. These skills may enable spin-off to
be more attractive to external financing.
Research Invention Patent Technology Technology License
expenditure i es application: licenses licenses income
—> > -— — -—
executed yielding
Ny Start-up or A viola
spin-off
Figure 4. University technology transfer process (Friedman & Silberman, 2003)
3. CASE STUDY: TOKYO UNIVERSITY OF SCIENCE SPIN-OFFS
Tokyo University of Science is one of Japanese private university, which
found its history back to 1881 foreran by Tokyo Butsurigaku Koshujo
(Tokyo Academic of Physics). Due to the enactment of Japanese TLO Law
in 1998 and Bayh-Dole Act in 1999, more attention was paid to transfer
university technology to industry. TUS TLO was established to facilitate
cooperation between academia and industry by creating, protecting,
managing and utilizing university-owned intellectual property rights. The
goal is to disseminate university accumulated knowledge to society (TUS
TLO, 2010a).
Figure 5. Tokyo University of Science technology transfer statistics
Increasing trend in number of invention disclosure and national patent
applications was historically recorded as shown in Figure 5 with highest
increase (almost three-fold) happened in 2004 (TUS TLO, 2010b). Slight
decrease in 2008 was due to priority shift from quantity to quality. This
suggests a preference toward concentrating TLO resources to inventions
with high commercial potential rather than submitting many patent
applications yet less marketable. Of total national patent applications, TUS
sole applications hold about approximately 48% in the last 5-years
average. In this paper, we do not incorporate joint applications since its
chance to be exploited through industrial licensing is higher than through
university spin-offs.
Responding to 2001 Hiranuma Plan, by 2004 fifteen new university
spin-offs were created in TUS (Figure 6) (TUS TLO, 2010c). However in the
succeeding years, most of the companies were out of business, leaving
four companies in 2012. Although one new company was established
afterward (will not be incorporated in the model simulation), low
establishment and survival rate suggests necessary effort should be taken
to improve current condition. At present, TLO support mainly consists of
free facilities and equipment for a certain period of time (later use will be
offered for a fee) and waived utility use fees for a limited time. Besides
that, new spin-off may use the name of university to indicate the
institution origin.
Figure 6. Number of Tokyo University of Science university spin-offs
4. CAUSAL LOOP DIAGRAM
A SD model is usually constructed based on a causal loop diagram (CLD).
CLD helps to visualize connections among variables in a system, which is
represented by arrows with positive or negative labels. Positive causal
link means when the cause (node where the link starts) increases (or
decreases), the effect (node where the link ends) also increases (or
decreases). The opposite direction is shown by negative causal link, in
which increase (or decrease) of cause resulted in decrease (or increase) of
effect.
CLD is developed to have comprehensive view of university technology
transfer process, particularly that leads to spin-off creation. Variables
incorporated in CLD, and in SD model as well, are extracted from direct
interview with TUS TLO staff and university faculty who is also a founder
of university spin-off and homepage of related institutions, ie. Japan
Patent Office, TUS, and TUS TLO, by referring to previous literatures on
university technology transfer. Through these variables, CLD helps to
determine the system boundary, which is set to university internal
environment.
As suggested by Siegel et al. (2003), the general flow of university
technology transfer process consists of variables such as scientific
discovery, patent, and license to firm (existing firm or start-up).
Consequently university may benefit from royalty or equity stake in a
new venture established around licensed technology. Variables related to
source of research funds give insight about how university research
projects are initiated Ustundag et al. (2011). Research projects might be
9
driven by industrial demand (resulted in industry sponsored research
contract) or personal interest of faculty members (funded by university
budget or national government fund) (Wayne, 2010). However, in this
paper we only include university budget and government fund, assuming
that industrial research fund would lead to joint patent application, thus
more likely to be licensed by corresponding industry. Other variables
include capacity of TLO e.g. size, labor division, PhD holder staffs,
experience, policy (Friedman & Silberman 2003; Ustundag et al. 2011), and
capacity of university, e.g. type, location, policy, regulations (Friedman &
Silberman, 2003; Wayne, 2010). Influence of external factors on university
technology transfer is represented by national government policy. This
variable will not be included in the SD model as it is out of the system.
Availability of initial capital and TLO support are two variables specific to
new business creation focused in this paper. Studies by, e.g. Ismail et al. (
2010), Peng (2006), and Seetre et al., (2009) emphasized the importance of
securing sufficient funding to develop new technology-based firms. In
terms of TLO support, Lockett & Wright (2005) found that TLO staffs with
business development capabilities are important determinants of a univer
sity’s success in creating spin-offs.
CLD of university technology transfer is illustrated in Figure 7. The
diagram is composed of core process of university technology transfer
(blue arrows), which is inseparably related to university and TLO (red
arrows) and external factors, in this case national government policy
(black arrow).
University technology transfer traces its process back to research projects,
which funds come from the university and government. New technologies
resulted from a particular research project add to the stock of invention
and may be patented then licensed by established firm. Licensee may gain
profit from the licensed technology, which includes percentage of royalty
to be given back to researcher. Royalty may further encourage researcher
to contribute to increase of stock of invention, thus completes the loop of
university technology transfer. Other than established firm, TLO may
decide to license new technology to university spin-off. However, higher
percentage of royalty from licensing activity to established industry may
demotivate researcher to create spin-off.
University and TLO influence the core process of university technology
transfer. Amount of yearly budget for research heavily depends on
university policy. University capacity, especially the one relates to faculty
quality, positively influences number of new spin-offs. University capacity
10
also indirectly influences number of new spin-offs through available
initial capital, which is provided through government grants. University
policy relates to the capacity of TLO in which it will support or deter
university technology transfer process. TLO capacity positively influences
number of patent application directly, while indirectly affect number of
new university spin-offs through its various support provided to
university faculty or staffs.
Figure 7. Causal loop diagram of university technology transfer
5. MODEL STRUCTURE
Based on CLD’s qualitative descriptions of university technology transfer
system, a SD model is then developed to have a quantitative analysis of
the system. Three sub-models represent research project model,
technology transfer model, and university-spin-off creation model (Figure
11
8).
5.1. Research project model
Research project is considered as the initial step of technology transfer
process as suggested by Friedman & Silberman (2003). Two source of
funds are highlighted in this model, i.e. university budget and government
fund. University yearly budget is largely determined by number of new
and current enrolled students, as it is the first contributor of university
yearly income. Researcher capacity (e.g. experience, achievements) may
increase the possibility of winning research fund for new project.
5.2. Technology transfer model
Subsequent to the enactment of TLO Law and Japanese Bayh-Dole Act,
TLO manages the filing of invention disclosures, which become the entry
of proceeding technology transfer stages. Nevertheless, faculty may not
disclose their inventions because of various reasons, such as not being
able to recognize the commercial potential of the invention or not willing
to spend time and effort in licensing or further development of the
technology (Jensen et al., 2003). These might be the reasons why only less
than half of the inventions with commercial potential are disclosed as
reported by the same authors.
On the other hand, not all invention disclosed are actually have
commercial value. TLO conducts necessary assessment in terms of
patentability and marketability before submitting patent application.
Patent application is not certainly examined; Japan Patent Office will
carry out an examination on one application only by request. Patent will
be granted if no reasons for refusal are found (Japan Patent Office, n.d.).
Patents may be licensed to attracted industry, yielding amount of license
income for university that is shared with faculty in form of royalty.
Royalty may further motivate faculty to disclose their invention and
involve in technology transfer activity (Goktepe-Hulten & Mahagaonkar,
2010).
Figure 8. University spin-off creation model
5.3. University spin-off creation model
Although licensing and university spin-off creation seems to not having
any close relationship, Di Gregorio & Shane (2003) found that percentage
of royalty share inversely related to spin-off creation per year. Higher
royalty distribution rate for faculty discourage faculty to choose spin-off
to exploit their invention. Other determinants include faculty quality,
which is measured by examining overall university ranking of Japanese
universities. Business birth rate and death rate are calculated by dividing
number of newly established or closed company in certain year by
number of active companies exist in the previous year.
Introduction to be observed in this SD model is the introduction of TLO
support and initial capital. The first introduction, TLO support,
emphasizes TLO roles in providing necessary aids to spin-off creation,
particularly because faculty lacks of business expertise and might prefer
to concentrate more on research. The importance of TLO in spin-off
creation is mentioned in previous literatures, e.g. Siegel et al. (2007),
13
Siegel et al. (2004). Here we define TLO support into constructing business
model, conducting market research, connecting to potential business
partner, assisting in legal issues. To provide these, TLO must have enough
resources, capabilities and knowledge in spin-off creation, which could be
met by recruiting qualified and experienced staffs.
Securing capital is an important matter in creating a business, including
spin-off. Given many constraints, university cannot afford to provide such
fund, thus external funding is crucial (Ismail et al., 2010). Nevertheless,
since the technology is still in early-stage, it might be difficult to compete
for venture capital. The funding may come in form of government grants
allocated to support commercialization of innovative, yet highly risky
university technologies (Peng, 2006; Seetre et al., 2009). Different with past
program where government only gave certain amount of money at once,
this time besides large amount of government grant given in the
beginning of policy introduction, fund was also available through a
continuous program of which university researchers could access any
time.
6. EMPIRICAL ANALYSIS
6.1. Preliminary model testing
Throughout model building process, several model tests were iteratively
done. Although it is impossible to validate or verify a model due to the
nature of model as limited and simplified representations of the real
world (Sterman, 2000), model testing is undoubtedly necessary. Model
testing took place after the initial model formulation was completed.
Three steps of model testing were suggested by Barlas (1996) of which
each has to be fulfilled before proceeding to the next step.
Direct structure tests are the first step, carried out by directly comparing
model structure with knowledge about real system structure. Simulation
is not yet involved. Besides empirically, direct structure tests can be
performed theoretically. Examples of test results are shown in Table 1
(see Barlas (1996) for more tests).
Table 1. Direct structure tests
Test name
Description
Example
a. Empirical
test
Structure-conf
irmation test
Parameter-con
firmation test
b. Theoretical
test
Direct
extreme-condi
tion test
Dimensional
consistency
test
Form of equations of the model is
compared with the relationships exist
in the real system.
Constant parameter is evaluated
against the corresponding element
exists in the real system, both
conceptual and numerical.
The likelihood of the resulting values is
assessed against the knowledge of
what would happen under similar
condition in real system.
The right-hand side of each equation is
checked for its dimensional
consistency with the left-hand side.
Model:
Obsolescence=(1-Approval of
patentability)*Stock of
invention
Real system:
Inventions that fail TLO
assessments (patentability
and marketability) may end
up obsolete, thus reduce the
stock of invention.
Model:
Approval of patentability=0.8
Real system:
Average percentage of TUS
national patent applications
is around 80% of number of
notification of invention in
the last five years (2008-2012)
(Figure 5).
Model:
Research project model
Real system:
If no new students enroll in
university, then university
budget for research must
decline. Number of new
project, ongoing project, and
finished project also must
decline.
Model:
New project (number/year
)=((University budget (yen
/year/number)/10000000 (yen
))+(Government fund (yen
/year/number)/50000000 (yen
))*Accelerating rate (unitless)
The second step is structure-oriented behavior tests, which indirectly
15
assess model structure by performing certain behavior tests on
model-generated behavior patterns. This time it involves simulation. In
extreme-condition test, extreme values are assigned to selected
parameters. The model-generated behavior is then compared to the
observed or anticipated behavior of the real system under the same
extreme condition. The result of extreme condition test toward
constructed model is shown in Figure 9. With a 1-year delay in ‘finished
project’ variable, behavior of ‘ongoing project’ is oscillatory. As explained
in previous direct extreme-condition test, if there is no new entrance, then
university budget must decline. New project and ongoing project must
decline as well. The same behavior happens when no government fund is
available.
[Pp
Figure 9. Results of indirect extreme-condition test
The third, also the last test, is behavior pattern test. The model is expected
to be able to accurately reproduce major behavior patterns exhibited by
the real system. We perform behavior pattern test using current data of
university spin-offs in TUS, by also considering actual amount of
university income raised from tuition fees and spent as university budget
for research, government fund for research acquired by TUS, number of
invention notification per year (calculated as notification rate), number of
patent application per year (calculated as application rate), total number
of patent grant, and number of university spin-offs (unpublished work)
(equations are presented in Appendix A). When run with available data,
the result shows similar trend with actual data (Table 2).
Table 2. Comparison of actual data and model testing result on number of university
16
spin-offs
Year Actual data Model testing
2002 12” 12
2003 15” 14.8
2004 15 14.8
2005 15 14.8
2006 15 14.8
2007 10 10.8
2008 10 9.8
2009 9 8.8
2010 8 7.9
2011 5 49
2012 4 3.9
“Initial number is based on assumption due to unavailable data.
6.2. Simulation
Simulation with the constructed SD model is run for three scenarios. In
Scenario 1, new spin-offs receive adequate support from TLO. We set
support on business model and practice as logistic curve (value: 1.06-3.76)
assuming that TLO would need time to carry out internal transformation,
especially in preparing qualified staffs with knowledge, capability, and
experience in spin-off creation. Scenario 2 reflected the availability of
government grant as initial capital (value: 1.5). Following large amount
given in the beginning, government grant is assumed to be available in
continuous manner.
To figure out which parameter contributes more in spin-off creation, we
run one more scenario. In Scenario 3, both TLO support and initial capital
are introduced at the same time. Although it was easy to assume that in
this scenario, number of university spin-offs must be higher than
previous scenarios, the idea was to see how large the two factors could
affect university spin-off creation if applied at the same time.
Because this study aims to, one of them, search for policy that can
promote spin-off creation, policy introduction would likely, and logically,
increase business birth rate. Corrected birth rate started at 0.67 (same as
test model), assumed as a response to Japanese government grant to
stimulate university spin-off creation. Yet, it gradually declined over time
17
and was stable at around 0.06. Initial simulation conditions are set as
follows: initial time=0, final time=10 (year), time step=1. All other
parameters are equal. Figure 10 shows simulation results compared to
base simulation.
[Pp
Figure 10. Number of university spin-offs after introduction by initial capital and TLO
support
As expected, introduction of TLO support and initial capital would
stimulate creation of university spin-off. Results for Scenario 1 suggest the
importance of providing services related to business creation, especially
because usually university researchers have insufficient business
knowledge and experience to start new company on their own. They
might also still prefer to spend their time and effort in research than in
commercializing their technology. Quality of services is highly influenced
by quality of TLO staffs. Therefore, it is worth the effort to recruit
excellent staffs with business experience, knowledge, and skills.
Introduction of initial capital in Scenario 2 would increase number of
spin-off in TUS. However, at year nine the number decreases. This
indicates that even when initial capital is available continuously, it was
not enough to keep number of spin-off in longer term. Other factors might
be essential to prevent business collapse.
18
Interestingly, when introduced separately, effect of initial capital exceeds
effect of TLO support for the first six years. It suggests that availability of
initial capital would encourage creation of university spin-off in a more
instant way, yet it would not last in a longer period. In reverse, since TLO
would need some time to improve their support and reach optimum level,
following increase in earlier years, number of spin-off would be more
stable, with tendency to increase in longer term.
When TLO support and initial capital introduced at the same time,
number of spin-off in year ten is 2x compared to effect of initial capital
alone and 1.5x compared to effect of TLO support alone. Furthermore,
growth trend of spin-off followed the trend for TLO support policy alone.
Consequently, when speaking about degree of contribution in the
combined policy, TLO support is likely to have higher contribution in
increasing university spin-off number than initial capital.
7. CONCLUSIONS
This paper described the use of system dynamics to model university
technology transfer, especially spin-off creation. The constructed model
simulated relationship between research project funding as the initial
step of technology transfer, main process of technology transfer, and
spin-off creation as one channel of technology transfer. Case study of TUS
was conducted to validate the proposed model.
Simulation results provided general view of how introduction of TLO
support and initial capital may enhance number of university spin-offs.
TLO support contributed more on the effect, especially for longer term.
This suggests that even in the absence of initial capital, it is important for
university to improve quality of TLO support, first by recruiting qualified
staffs. They may open new doors of opportunity, including finding
potential financing and connecting to business partners by using their
network. Nevertheless, it is important to be noted that ups and down can
still be seen in the later period. Appropriate policy should be made to
increase survival rate of university spin-offs, while fostering new
companies at the same time.
REFERENCES
19
Barlas, Y. (1996). Formal aspects of model validity and validation in
system dynamics. System Dynamics Review, 12(3), 183-210.
Baycan, T., & Stough, R. R. (2013). Bridging knowledge to
commercialization: the good, the bad, and the challenging. The Annals of
Regional Science, 50(2), 367-405.
Bradley, S. R., Hayter, C. S., & Link, A. N. (2013). Models and Methods of
University Technology Transfer (pp. 1-73).
Breznitz, S. M. O’Shea, R. P., & Allen, T. J. (2008). University
Commercialization Strategies in the Development of Regional Bioclusters.
Journal of Product Innovation Management, 25(2), 129-142.
Di Gregorio, D., & Shane, S. (2003). Why do some universities generate
more start-ups than others? Research Policy, 32, 209-227.
Etzkowitz, H. (2003). Research groups as “quasi-firms”: the invention of
the entrepreneurial university. Research Policy, 32, 109-121.
Friedman, J., & Silberman, J. (2003). University Technology Transfer: Do
Incentives, Management, and Location Matter? Journal of Technology
Transfer, 28, 17-30.
Goktepe-Hulten, D., & Mahagaonkar, P. (2010). Inventing and patenting
activities of scientists: in the expectation of money or reputation? J
Technol Transf, 35, 401-423.
Gomez Gras, J. M., Galiana Lapera, D. R., Mira Solves, I., Verdu Jover, A. J.,
& Sancho Azuar, J. (2007). An empirical approach to the organisational
determinants of spin-off creation in European universities. International
Entrepreneurship and Management Journal, 4(2), 187-198.
Ismail, K., Mason, C., Cooper, S., Wan Zaidi, W. O., & Majid, I. A. (2010).
University Spin-off Formations: How decision making process has been
made? International Journal of Business and Social Science, 1(2), 103-123.
Japan Patent Office. n.d. Procedures for Obtaining a Patent Right.
Retrieved March 17, 2014, from
https://www.jpo.go.jp/cgi/linke.cgi?url=/tetuzuki_e/t_gaiyo_e/pa_right.htm.
Jensen, R. A., Thursby, J. G., & Thursby, M. C. (2003). The Disclosure and
Licensing of University Inventions (pp. 1-28).
Lockett, A.. & Wright, M. (2005). Resources, capabilities, risk capital and
the creation of university spin-out companies. Research Policy, 34(7),
1043-1057.
Miki T. 2012. RABICSITSAFHBAYF v -—HROBHYCSROABME
20
(Progress and future directions of university ventures policy in Japan). ~
FRAERAWMAIOY cr STARRY YRY IL, AAR
4 INAV AY - TAYATAOBRICAIT (Creation of new
university-launched industries based project (START) Symposium:
Towards the construction of a Japanese innovation ecosystem). National
Center for Industrial Property Information and Training, Japan.
Mitasiunas, J. (2013). Innovation and Technology Transfer (pp. 1-9).
Ogura M. 2009. Academic Start-ups Survey 2007-08. Survey Material
No.173. 3"¢ Policy-Oriented Research Group, National Institute of Science
and Technology Policy (NISTEP), Ministry of Education, Culture, Sports,
Science and Technology (MEXT), Japan.
O’Shea, R. P., Chugh, H., & Allen, T. J. (2008). Determinants and
consequences of university spinoff activity: a conceptual framework. The
Journal of Technology Transfer, 33, 653-666.
Peng, X. (2006). University spin-offs: Opportunity or challenge? Nature
Materials, 5(December), 923-926.
Rasmussen, E., & Borch, O. J. (2010). University capabilities in facilitating
entrepreneurship: A longitudinal study of spin-off ventures at mid-range
universities. Research Policy, 39(5), 602-612.
Setre, A. S., Wiggins, J., Atkinson, O. T., & Atkinson, B. K. E. (2009).
University Spin-Offs as Technology Transfer: A Comparative Study among
Norway, the United States, and Sweden. Comparative Technology Transfer
and Society, 7(2), 115-145.
Sheft, J. (2008). Technology transfer and idea commercialization. Nature
Biotechnology, 26(6), 711-713.
Siegel, D. S., Veugelers, R., & Wright, M. (2007). Technology transfer offices
and commercialization of university intellectual property: performance
and policy implications. Oxford Review of Economic Policy, 23(4), 640-660.
Siegel, D. S., Waldman, D. a, Atwater, L. E, & Link, A. N. (2003).
Commercial knowledge transfers from universities to firms: improving
the effectiveness of university-industry collaboration. The Journal of High
Technology Management Research, 14(1), 111-133.
Siegel, D. S., Waldman, D. A., Atwater, L. E., & Link, A. N. (2004). Toward a
Model of the Effective Transfer of Scientific Knowledge from
Academicians to Practitioners: Qualitative Evidence from the
Commercialization of University Technologies. Journal of Engineering and
21
Technology Management, 21(1-2), 115-142.
Sterman, John D. (2000). Business Dynamics: System thinking and
Modeling for a Complex World. The McGraw-Hill Companies, Inc. US.
Swamidass, P. M., & Vulasa, V. (2009). Why university inventions rarely
produce income? Bottlenecks in university technology transfer. J Technol
Transf, 34, 343-363.
Teikoku Databank. 2013. AZRAVF v—, MtRMOSFRS ~HRRWE
. 1000 #atis 4‘2¢5~ (University venture, majority profitable: Liberal
Democratic Party government, "1000 companies Plan" contribution). #35!
RE: AFHAVFv—-BRORRAA (Special Event: Survey of
university-launched venture company). Teikoku Databank, Ltd.
The World Bank. 2014. Patent Applications, Residents. Retrieved March 10,
2014 from http://data.worldbank.org/indicator/IP.PAT.RESD?order
=whapi_data_value_2011+whbapi_data_value+wbapi_data_value-last&sort=
desc.
Thompson, B. P., & Bank, L. C. (2010). Use of system dynamics as a
decision-making tool in building design and operation. Building and
Environment, 45, 1006-1015.
TUS TLO. 2010. Greetings from the Director. Retrieved March 15, 2014,
from http://www.tus.ac.jp/tlo/english/tlo/ greeting.-html.
. 2010. Historical Results. Retrieved March 15, 2014, from
http://www.tus.ac.jp/tlo/english/service/result-html.
. 2010. AFI Y F v —(University ventures). Retrieved March
15, 2014, from http://www.tus.ac.jp/tlo/contents/venture.html.
Ustundag, A., Ugurlu, S., & Kilinc, M. S. (2011). Evaluating the performance
of technology transfer offices. Journal of Enterprise Information
Management, 24(4), 322-337.
Wayne, K. (2010). Determinants of Commercial Innovation for University
Technology Transfer. Journal of Behavioral Studies in Business, 2, 1-23.
Wu, D. D., Kefan, X., Hua, L., Shi, Z., & Olson, D. L. (2010). Modeling
technological innovation risks of an entrepreneurial team using system
dynamics: An agent-based perspective. Technological Forecasting & Social
Change, 77, 857-869.
N
i)
APPENDIX A. Description of parameters
Parameters Equation Value Unit
Research project
model
Ongoing project 300 Number
New project Z1=10000000, Z2=5000000; Y= Number/year
(University budget/Z1
)+(Government fund/Z2
))*Accelerating rate
Finished project Y=DELAY(Ongoing project,1) Number/year
New entrance 5700 Person
Body of students 20000 Person
University budget Z1=1500000/person, Z2 Yen
=1200000/person; Z3=0.15; Y= /year/number
((New entrance*Z,)+(Body of
students*Z2))*Z3
Government fund Graph Yen
/year/number
Researcher capacity 1 Unitless
Accelerating rate Graph Unitless
Technology transfer model
Stock of invention 30 Number
Patent application Y=Application rate Number
Patent grant 5 Number
Licensed patent Y=Licensing rate Number
Notification Y=Finished Number/year
project*Willingness to notify
Obsolescence Y=(1-Approval of Number/year
patentability)*Stock of
invention
Application rate Graph Number/year
Inspection Z=0.4; Y=Patent Number/year
application*Sole patent
granted*Z
Licensing rate
Willingness to notify
Approval of
patentability
Sole application
Sole patent granted
License contract
Royalty
Royalty share
Graph
Graph
Graph
Y=Royalty share*License
contract
University spin-off creation model
University spinoff
Creation
Out of business
Business birth rate
Business death rate
Intellectual
eminence
Capital
Support on business
model and practice
Y=Patent grant*Birth rate of
business*Intellectual
eminence*Royalty share*
Capital*Support on business
model and practice
Y=University spinoff*business
death rate
Graph
Graph
Graph
0.8
0.5
0.4
0.5
12
0.97
15
Number/year
Unitless
Unitless
Unitless
Unitless
Yen
Yen
Unitless
Number
Number/year
Number/year
Unitless
Unitless
Unitless
Unitless
Unitless
24
